Don Lancaster's What's New? 2011 Archive Blog

Pick your blog year...
1997	1998	1999	2000	2001
2002	2003	2004	2005	2006
2007	2008	2009	2010	2011
2012	2013	2014	2015	2016
2017	2018	-----	-----	-----

Seems to be the time of year for WAG predictions.
Here's my current assortment...

While pv solar still remains a gasoline destroying
net energy sink that is in no manner renewable
nor sustainable, it is in the process of turning
around and ultimately becoming viable. In
part because of CIGS and the new quantum dot
developments, improvements in high frequency
efficiency, and ( thankfully ) to the elimination of
obscenely counterproductive subsidies.

Goldilocks and other exoplanets are now being reported
at a one per day rate and soon will become common
as dirt. We will be in deep trouble the ( reasonably soon )
day that one of them receives a lucid ten second clip
of Roller Derby as the sum total of our culture. The
alternatives, of course, are Captain Video or Kukla,
Fran, and Ollie.

HVAC ( heating, ventillating, air conditioning ) is
obviously due for a major efficiency upgrade.
Brought about by MEMS and nanotechnology.
One obvious product crying to become reasonably
priced: hot tub heat pumps.

Your website is unlikely to just keep growing in
popularity like it did in the past. Brought about by
the "eyeball siphoning" by the social networking sites,
the upsurge in online video, and the norm now being
full HDTV video as a web experience.

Things that no longer make any sense whatsoever are
the US Post Office, "Big 3" network Television, over-
the-air tv transmission, public libraries ( unless totally
and immediately reporposed ), or any tax funded
Public Broadcasting. Also utterly useless is any
tv set that does not include unstricted web access
or does not provide at least full HDTV resolution .

Santa Claus machines are about to drop dramatically
in price and greatly expand in availablity. Brought
about in part by low end open source examples
such as this one, this one, or this one.

Word will finally get out over exergy being the reason that
the hydrogen economy flat out ain't gonna happen. In
particular, electrolysis from high value sources such as
grid, pv, or alternator is pretty much the same as 1;1
exchanging US dollars for Mexican Pesos. A kilowatt
hour of most electricity is ridiculously more valuable
than a kilowatt hour of unstored hydrogen gas.

Printed industrial trade journals can be expected to vanish
enritrely. Their useful content has been long since gone
and their economic model has become laughingly pointless.

Drug legalization should soon dramatically turn around as
governments and municipalities will no longer be able
to ignore it as a major income source.

Dot matrix printers can be ridiculously improved by going
to a pair of duplexing full width printheads. The mechanisms
become ridiculously simpler, the paper path totally direct,
and speeds insanely faster. Reliability and the cost learning
curve will be the key issues here.

The latest of light emitting diodes now include warnings
not to look at them because they are too bright. Efficiencies
are getting totally offscale, with 150 Lumens per watt
production and 250 in the lab.

Closed out the 2011 Archive and started this 2012 one.

Per the serial arson flap in Los Angeles: At least here in Arizona,
torching a car is no big deal.

But torching a car in a carport is arson of an occupied structure.
And has the same penalties and enthusiastic conviction as Murder One.

Finally got some distortion correction in our Architect's Perspective
that looks like it is complete and accurate. This sample shows
vertical, horizontal, and major diagonals straight. Intermediate
angles are also likely fully corrected, although this is not yet
proven.

The algorithm goes something like this: After "untilting" a
vertical subject edge, that line's scaling is remembered and
a correction gets mad by elevating row content, sort of in a
somewhat parabolic manner. This leaves verticals and horizontals
straight, and "unbends" the diagonals.

Starting with the top row, the scale factor needed for the initial
correction is found. This is applied to the rightmost diagonal
and its vertical position error is found. The line is then repositioned
upwards as needed to compensate.

An occasional missed line is filled in by a double repositioning.

I'll try to work up some further code and testing whenever.

Had a caller ask about finding geranium transistors for
guitar stomp boxes. Naturally, in this day and age, if you
want to add noise, hum, or distortion, you use a decent
DSP algorithm rather than vacuum tubes, primitive solid
state devices, or other performance limiting and highly
unpredictable components.

But to answer the specific question: One good info source for
older semiconductor secret sources in the sci.electronics.design
newsgroup. And eBay has bunches of older semis readily
available. Both on our own auction store and elsewhere.

A surprisingly high number of older components still remain available
at major distributors. A very good place to check is OEM's Trade.
A second often useful resource is PLC Center.

Meanwhile, there are firms that specialize in obsolete smis. And, if
all else fails, they will even custom build new ones for you. Two
examples include Rochester Electronics and 4 Star Electronics.

4 Star seems be down to their very last 28,525 NKT275's. Which
remain enormously popular with the stomp box revivers.

The neat thing about germaniums was that they were so leaky that
you didn't really have to bias them. Provided, of course, that you only
ran them at one temperature.

The math behind the distortion in Architect's Perspective
is a tad obtuse, but easily resolved using a first principle
of "make it worse" and seeing what a simple polynomial tells us.

Consider this extreme case of a fifty percent correction. We apply
the usual "progressively shorten each line" for this intermediate result.
Where we see that our vertical lines are now straight and vertical, and
our horizontal lines remain straight and horizontal. But, sadly, our principle
diagonals are badly distorted and bent. As will be much of the rest of the
image.

It looks like a repair can be made by raising the intermediate rows. Perhaps by
one sixth in the middle. Normalize the right diagonal so it goes from 0,0 to 1,1.
Assume this bent curve can at least be approximated by a polynominal of
y = ax^2 + bx + c. Since y=0 at x=0, c will also be 0 and can be ignored.

y = 1 at x = 1, so we have our first equation of a + b = 1. Assume from measurement
of the figure that y = 1/3 at x = 1/2. Which gives a second equation of a = 2b .
And a tentative solution of a = 2/3 and b = 1/3.

The desired "curve" should be the linear y = x, so our error function will be
x - 2/3*x^2 - 1/3*x    Or 2/3x*(1-x).

Per this plot, the approximation correction appears "close enough" But not
a true correction. Looks like a further small sinusoidal adjustment may help. Or
finding the actual "real" math.

More testing soon. And code eventually.

Some amazingly clean "nearest neighbor" interpolation code
seems to have evolved from our distortion correction
explorations.

Our usual array-of-strings will contain a pixel line in the
typical Bitmap   .....bgrbgrbgr...  sequence. A desired left
shift or right shift could end us anywhere in the sequence.

Which clearly is ungood, because blue has to stay blue
and the starting point of each pixel triad must be an
integer as well.

This repositioning does it all...

1.5 add cvi 3 idiv 3 mul

More on sneaky PostScript stuff here.

Piher has an interesting N-15 position sensor available
through Mouser at as little as a dime each. This is basically
a continuously rotating potentiometer without stops that
is electrically active over at least 340 degrees. It has
a mechanical rating of at least 100,000 cycles.

What could get cute in a hurry is if you stacked two of
these offset by 90 degrees. With some reasonable
microcomputer smarts, you should be able to directly
output either full angles or sine/cosine quadratures.

It can be real easy to lose track of the fact that storage
units should be emptied at the same rate they are filled.

Otherwise you end up needing more and more space with
worse and worse junk hiding the stuff you really need.

Otherwise known as "shit floats to the top".

There is an optimal rate at which stuff should end up on
eBay. For us, this rate is often a twenty eight day cashout
and a fifteen month hang time.

Thus, ideally, stuff should be listed within a very few days
of acquisition and never should need stored more than two
years. Most espeically if not listed.

Sooooo, we are long overdue in making a dramatic cutback in
our stored inventory items. These will be flushed to a
third party and should become available at ridiculously
bargain prices. By the truckload.

You can email me for further details.

More on eBay tips and techniques here.

A new video of San Carlos falls can be found here.
And other unusual local dayhikes here and here.

A reminder that our original Introduction to PostScript
video has been totally reworked and split into two
You Tube pieces. Find them here and here.

A reader asked about .PDF files for some handheld geneology
apps. I was not sure if what level of response was needed.

For many uses, Adobe Acrobat .PDF is an ideal way of presenting
technical information. Especially when the exact layout format is
important and when high quality integrated graphics is a must.

There are three flavors of .PDF. The Acrobat reader is free but largely
prevents you from creating yoxur own material. Commercial or educational
versions of Acrobat can accept a wide variety of input content for conversion.

And alternate versions such as GhostScript give you open source options.

Acrobat .PDF is based on the PostScript language, of which you'll find lots more
here. But because of the extreme power of PostScript, Adobe has restricted its
ability to do most of its more useful tasks involving unlimited disk access. The
SX workaround is to run "acrodist-F" from the SX command line for full use.

But Adobe has qotten quite secretive about this access option recently, especially
on other or newer computer systems.

I remain an overly enameled fan of PostScript and .PDF. But with newer
small screen or low resolution devices, there are emerging alternatives that
reasonably should be explored.

In particular, check out the Android 3 version of Java. With some help from
Linux and C+. Microchip has a new Android Development Kit as well.

Something that could be between an interesting development
and a major pv panel solar breakthrough can be found here.

At present, the efficiency of a conventional solar cell is abysmal.
You would be extremely fortunate to get nine percent efficiency at
the terminals of your synchronous inverter.

There is a fundamental technical reason for the bad efficiency.

Any wavelength of light has an energy level associated with it.
For e equals h nu and all that.

If you have a conventional single workfunction photoconverter,
any light whose wavelength is too long can not knock any electrons
loose, and is lost as heat.

If the light's wavelength ( and thus its energy ) is just right, you can
knock one electron loose with quite high efficiency.

If the light's wavelength is too high, all of the "spare change" energy
is also lost as heat. Thus, only one magic wavelength of light can
be efficiently converted. Too high or too low, and you burn up the
leftovers.

Previous proposed workarounds have included cells with several
workfunctions, quantum dots, or light frequency downconverters
involving phosphors and similar stunts. None have proven economically
or engineeringly viable to date.

The proposed new method apparently can knock two or more electrons
loose when fed light of a high enough frequency. And does initially
seem eminently workable.

The Science paper can be found here.

Some pv panel economics can be found here, an industry trade journal
here, a pv solar tutorial here, and general energy stuff here.

Expanded and improved our Gila Day Hikes library page. Besides
adding the Unusual Hikes Slide Show, we are now only 22 entries
shy of our goal of offering 365 unique things to do. Enough to keep
you busy for the next year. Please email me with any locations that
I may have missed.

Updated and expanded our Auction Help library page and our
Arizona Auction Resources links. Your own custom regional auction
finder can be created specifically for you per these details.

Our Incredible Secret Money Machine library page remains sorely in
need of an update and rework. But I did manage to add a link to our
newly available free ISMM eBook. And we also have a new
ISMM fast nav directory.

Uh, whoops. Part of the distortion observed in our Architects
Perspective tilt eliminator was simple operator error.  

On any subject with slanty dominent lines, it is crucial that the
neutral tiltaxis be "at the base of the vee". If needed, the entire
image can be rotated to make sure the base of the vee is at
a neutral vertical axis requiring zero correction..

Distortion can be further minimized by shooting at as low an
elevation as feasiible and keeping the dominant slanty line as
near horizontal as reasonable.

I was also surprised that the Nikon Coopix lens distortion was not
totally negligible. At times, it can approach one fifth to one quarter of
the total distortion.

At any rate, with reasonable care and a possible second pass or two,
the distortion can end up pretty much negligible. Especially for eBay.

It can be further minimized by a distortion corrector I am working on
that provides programmable vertical shifts in the low pixel range.
Centering the horizontal tilt axis may also end up worthwhile, but has
issues yet to be resolved.

Many years ago, I was attending a folk concert. The opening act
was a single and an unknown flute player, performing in front of the
closed stage curtains. His job was to warm up the audience for the
high priced help that was soon to follow.

He was good. Very good.

But as he went along, the music started getting strange and finally
downright weird. He was playing chords on his flute, along with notes
with unbelievably strong tonal structures. Eventually, the music turned
into bunches of impossible sounding and god-awful squawks.

Almost all of the audience got bored and restless as the music
seemed to deteriorate. Just then, I happened to notice a friend beside
me who had played in and had taught concert band. He was literally
on the edge of his chair with his mouth open.

He briefly turned to me and said very slowly, ''You can't do that with
a flute. It is not possible."

Of the thousands of people in the audience, at the most only five
realized they were now witnessing a once-in-a-lifetlme performance
involving the absolute mastery of a very difficult musical instrument. To
nearly everyone else, It sounded like a bunch of god-awful squawks.

Always play for those five.

Turns out there still is some minor distortion in our new Architects
Perspective tilt eliminator. The effect shows up as very minor barrel
distortion on subjects with exceptionally long diagonal elements.

The effect is negligibly minor on eBay image style resolutions and may
in fact be partially caused by the camera lens itself. You can view the
effect here by coming up in Paint and addiding red rules to the
diagonal lines.

The underlying problems are that both X and Y pixel shifts would be needed
for a more exact correction. And that much better pixel interpolation
techniques would be required. And that speed is likely to become an issue.
And that near perfect corrrection might demand a true ray tracing or
end up highly subject dependent.  

The simplest cure, of course, is to ignore it. Or to avoid compositions
with huge diagonal elements. Alternately, the code could be rewritten
so its tilt is vertically centered ( or even programmable ). That way the
least alteration would be made to the centralmost pixels.

Alternately, the center pixels could be raised in a slightly parabolic manner.

It does turn out that you can buy lenses that fake those wonderful view
camera swings and tilts. For a mere $2200 each. I'd tend to prefer using free
software and post processing tools instead.

Just received the one zillionth email of "I am an inventor.
How do I keep others from stealing my ideas?"

Sorry, but the reality is as follows:

1. If your idea is any good, it WILL be stolen.
 
2. Prior art almost certainly exists.

3. Chances are you don't have the faintest clue
     over the device marketing realities.

For the overwhelming majority of individuals and small scale
startups, patents are ABSOLUTELY CERTAIN to end up
as a net loss of time, energy, money, and sanity.

Thus, the mindset involved in the paranoia of "protecting my idea"
is not even wrong. Time and effort should instead be focused
on developing and improving the actual product.

Much more here and here.

And, of course, never, NEVER, NEVER call yourself an inventor!

Turns out the "reverse tilt" feature of our new Architect's
Perspective tilt eliminator works just fine.

Just enter a negative value for howmuchtilt. This will
compress the bottom, leaving the top the way it is.

A nonlinear process is applied to properly correct lines
intended to be vertical. Should a wildly wrong image be
input, the distortion can appear to become worse. This is
simply the math doing what it is intended to do.

Modifying the code to tilt center rather than tilt top or bottom
might minimize some of these effects. But this could raise
issues with needing a second array of strings or with needing
better interpolation algorithms.

I was apalled by the casualness with which an assistant of
ours dismissed the "only 4 Gigs" of memory in their latest
handheld toy.

For I was present that glorious day at Goodyear Aerospace
when the price of solid state memory finally dropped to an
astonishing and utterly unheard-of low of a nickel a bit.

Which was sort of cost parity with magnetic core.

At long last, it could be theoretically possible to do solid state
Fourier dechirping tricks to eliminate the horrendous optics of
side looking radar correllation . And have intelligent video displays
that could -- gasp -- recalculate what they are showing!

A nickel a bit would be a mere piddling forty cents per 8 bit word!

And someday in the dark distant future, it might even be possible to
build a suitcase sized device that could multiply as well as add or subtract.

Why, 4 Gigs of memory would only cost a trifling 1.6 billion dollars.
Say seven or eight billion with a B allowing for inflation.

Sigh.

Some preliminary code on our new Architect's Perspective
tilt eliminator can be found here.

This completely eliminates the distortion problem of the previous
versions. It is also quite fast, doing a 2K x 2K bitmap correctiion in
seven seconds.

There is still a bug in the reverse tilt option. Operation is presently
from bottom rather than center of screen. Several passes of
speedup have been completed.

Please email me with any problems, enhancements, or suggestions.

Managed to visit yet another mile of prehistoric hanging
canal yesterday. This one was a "more of the same"
western feeder to the Twin Boobs ponding area. And
had been previously surveyed by others.

Included was a fairly distinctive "S" turn to get across a
wash. Very similar to several others. The present score
is something like eighteen canals for a total of over
sixty miles.

Field mice and funding are definitely needed to give these
new and stunning world class rediscoveries the attention they
surely deserve.

Its source could be the three way switch on Deadman
Mesa via Lower Deadman tank. But this remains to
be proven.

Much more here.

Expanded and updated our Gila Valley Day Hikes page.

Newly included is a link to our Unusual Dayhikes Slide Show.
Whose sourcecode can be found here.

Science Magazine for December 2, 2011 describes some new
catalysts for hydrogen  electrolysis. On pages 1256 to 1260.

But (1) all a catalyist can do is speed up or slow down an existing
process, and (2) the system can only reduce excessive losses in
alkaline systems.

As I seem to have mentioned a time or two before, there is a
profound and fundamental thermodynamic principle that
ABSOLUTELY GUARANTEES that bulk hydrogen energy
electrolysis from high value sources ( such as grid, pv, or
alternator ) flat out ain't gonna happen.

Ever.

Because of exergy, a kilowatt hour of electricity is ridiculously
more valuable than a kilowatt hour of unstored hydrogen gas.

The process of electrolysis is thus similar to 1:1 exchanging
US Dollars for Mexican Pesos.

Much more on related topics here.

While I remain extremely impressed with the new Garmin
eTrex-30 GPS handheld with barometric altimeter, I have
not yet been able to resolve some slope direction issues
in our prehistoric hanging canals.

Yes, the display gooes up on foot on a one foot rise and
down one foot on a one foot drop from belt level. But
some of the longer term drift seems disconcerting.

Altitude performance can be greatly improved by setting
the unit stationary and flat in the shade and waithing a minute or
so for a reading. And by noting any pre- or post- longer term
drift caused by barometric variations.

Returning to the initial point should give you a reliability
check on the readings. Averaging can be useful, but only if
your time out equals your time back.

Picked up some interesting mirror style mounted beam splitters
and some "raw" 1x1x3 CM Lithium Niobate electro-optic single
crystals that we have placed on our eBay auction site.

We think our pricing on these aerospace surplus items is
exceptionally low, but it it tricky to find either data or pricing
on either of these intriguing items.

PostScript can be more than fast enough for reasonable
large bitmap image post processing. Provided some obvious
guidelines are strictly followed.

An array of strings seems to be a useful data structure, with
each string representing one horizontal line of the original
bitmap. As much as possible, tasks should be "along the
grain" involving only one string, rather than "across the grain"
in which many strings have to be seequentially accessed.

PostScript garbage collection can take an inordinate amount of
time, especially with Acrobat Distiller. A -2 vmreclaim can make
an amazing speedup in certain circumstances.

In any large loop within a loop, the inner loop will overwhelmingly
dominate your system timing. Thus, minimizing the total number of
commands in the inner loop is of crucial importance.

Conditional tests should be eliminated when and wherever possible.

Ferinstance, if overflows or underflows might happen, it is far
better to precalculate and prevent, rather than retest every time.

Stack manipulations are generally much faster than external lookups.
But a 7 index is slower than a zorch lookup, owing to the second
needed command. In general, multiplies, trig, and rand will be rather
slow and should be avoided. A sneaky rand speedup can be found
here. A small random pattern array can be ridiculously faster than some
"on the fly" techniques.

The minimum acceptable pixel interpolation technique should be used.
"Nearest neighbor" is considerably faster than quadratic or bicubic
interpolation. And may be all that is needed if the image size is
being reduced. Especially if routine bicubic reductions can be done
later by more conventional programs.

Timing measurements can be done using the stopwatch routines in our
Gonzo Utilities. In general, code should be gotten to work properly
first, and then sped up later. Several speedup passes may be needed
for optimal results. But avoid obsessive rework.

The bind command is often good for a modest speedup.

Two of our latest PostScript bitmap manipulation examples are found
here and here. Plus the whole gang here and here.

Consulting services available.

Yorg.

I've long been a fan of Architect's 3D perspective in which any
vertical lines appear truly vertical. Especially for our eBay images.

Our Swings and Tilts routine has long had a bug in it where corrections
above 15 percent would end up noticably curved. Up to the point of being
unusable above 20 percent. It turns out some super subtle math is
involved when you are trying to emulate the distortion free swings and
tilts of a view camera.

Consider an extreme case of a 50 percent correction and being 100 pixels
to the right of a neutral axis at screen bottom and 200 pixels at the top.
No correction and unity gain are needed at the bottom. A 100 pixel
correction and 0.5 gain are needed at the top. So far, so good.

You might ( as I did ) wrongly guess that an 0.75 gain would be needed
mid screen. Not so. Mid screen, our row would be 150 pixels wide. The
correct gain would be the inverse of 150/100 or 0.667!

A huge error!

The math, while simple, is amazingly subtle. Fortunately, it can all be
done only once per line, so it can actually be faster than the original.

Given a pc percentage correction and a sv line fraction, the fsf final
row scale factor is ...

1 dup pc sub div 1 sub sv mul 1 add 1 exch div

... and plots something like this.

I'll try to post the revised code in a day or two.

It pretty much pays to use a "last in, first out" strategy with
auction acquisitions for eBay resale. Otherwise, unlisted junk
tends to accumulate forever.

And made worse by the immutable fundamental storage principle of
"shit floats to the top".

Reasonable goals are a 30:1 SBR sell-buy ratio, a 21 day cashout
time and a 15 month hang time. Anything left unprocessed after two
years should clearly be flushed.

Much more here and here.

An update to pv panel pricing can be found here.

At $2.43 panel per peak watt, this is a mere TEN TIMES the
cost demanded for displacable renewability or sustainability.

But the gap is rapidly closing.

It can be useful to once again reveal where the holy grail
of "a dollar per peak watt" total installed cost comes from.

Assume a 1000 watt peak panel in a workable location. It
typically will generate 5 kilowatt hours of electricity per day with
a typical utility value of ten cents a kilowatt hour or 50 cents
total. Fifty cents a day is fifteen dollars a month.

Per this amortiztion schedule, an investment of $1135 for ten
years at 10 percent amortizes to fifteen dollars per month. Or
slightly over one dollar per peak watt.

But note several gotchas: First, that is the complete system cost.
Panel cost is typically 35 to 40 percent of the fully burdened
installed system cost. Second, subsidies are a scam that should
be treated as a 3:1 liability rather than as an asset, when all of
their true collection and distribution costs are fully and properly
accounted for.

Third, matching utility rates does not give much incentive for
displacing conventional energy sources.  All you have is a
"paint it green" feelgood going. A reasonable expectation for
everybody to jump on the pv solar bandwagon would thus be
twenty five cents per peak panel watt. With there being a
reasonable changeover profit incentive.

Finally, once a quarter per peak panel watt is achieved, things will
get a lot worse, owing to the zillions of new bucks ( at an exchange
rate of ten kilowatt hours per dollar ) being thrown at it for future
gain.

Today, pv panels remain in no manner renewable nor sustainable.

A reasonable expectation would be for true energy breakeven eight
to ten years after the peak panel price drops under twenty five cents
per watt.

Additional resources and analysis here.

GuruGram #121 is newly released on Some Little Known Gila
Valley Day Hikes.

I'll also be doing a talk on this Saturday December 10th in
Safford's EAC Discovery Park at 6:30 PM.

Additional GuruGrams can be found here and here.

Let me make sure I've got this straight: The US Post Office's
response to losing customers is going to be to reduce the quality
and timeliness of their service.

Loss of customers is certain to accellerate. Netflix wants to bail
in the worst sort of way. Outrageously expensive printed catalogs
make no sense whatsoever in the era of instant 24/7 sales and
situational pricing. Newspapers are clearly the walking dead.
And magazines lack timeliness, customizability, and searchibility.
Not to mention inconvenient and ungainly storage.

Do we really need a bureaucratic quasi-government solution at all?
And for what? How does this differ from slide rule manufacturers or
purveyors of printed circuit tape and dots?

To me, any solution would come about from a fundamental "reset to zero"
rethinking. And the essential key core to any possible solution would
be to immediatly repeal the Private Express Statutes.

 

The "terabyte era" is utterly and totally pointless. Why
don't we just skip it entirely and go directly on to petabytes?

In which you would have one thumb drive called "movies" with
all of them on it. And one thumb drive called "songs" with
all available music on it. And, of course, similar drives called
"magazines", "tech papers", "news", and such.

All attractively gift boxed for, say $19.85 for the set.

This would also once and for all completely resolve DRM issues.

It turns out that rectocranial inversion cen end up both
chronic and acute at the same time.

We have a winner!

The first Goldilocks exoplanet has recently been found and
verified by the Keck Observatory.

Other exoplanets are apparently being cataloged at nearly a
one a day rate. And it seems that many terms in the Drake Equation
were initially grossly underestimated.

I feel the number of Goldilocks candidates is essentially unlimited.
And that many thousands will be known in a remarkably short time.

When they marionate shrimp, how do they tie all those
little strings on?

Stanford University has apparently decided to join MIT in
offering free online courses.

I'm presently going through their Fourier Series course. I also
recently found MIT's Solid State Circuits course of interest.
Especially their Pythagorian Circuit which builds up an analog
square-root-of-the-sums of squares out of seven transistors
arranged as a pair of current sources and a differential amplifier.

Additional free online course links can be found here.

A caller aasked me about dies for a cardboard recycling app they
were exploring.

Some limited dies are available through scrapbooking sources,
while Di-Acro ( now Acrotech ) and Roper-Whitney are leading
sources for low end industrial punch and die sets. For years and
years, I've been particularly partial to the Roper XX punch.

But there's an entire converting industry based on modifying raw
paper and cardboard products and converting them into such things
as labels, tags, tickets, badges, bumperstickers, point-of-purchase
displays and such. Three of their remaining free trade journals include
Paper, Film, and Foil Converter, Converting Today, and
Converting Quarterly.

The usual way of dealing with the caller's problem is called a
steel rule die. These are far cheaper than industrial die sets and
extremely flexible. Typically, a better grade piece of plywood is
laser cut with a desired pattern and a steel strip similar to
strapping material is pressed into it. Rubber is usually added as an
eject mechanism.

One of the hardest and most crucial product development
development rules is simply "know when to stop".

I have an associate who, decades ago, started developing a
bleeding edge hand crafted product of extreme ruggedness
and reliability. Each time they were about to go into
production, they saw a "new" and "better" approach and
kept redesigning and redeveloping, over and over again.

Eventually, China caught up with them, and the competition now has
these wall to wall in Wal-Mart for unprofitable pennies. Quite
simply, the opportunity window slammed shut.

More to the point, I spent the last week trying to improve and
speed up our automatic background mottling utility. Despite its
taking two seconds to replace over an hour's manual image post
processing time. And any further speedups being utterly and
totally pointless.

I sure was proud of my new ultra subtle PostScript code of...

false 2 redmap curvpos get
{ 255 eq { exch pop true exch }
{1 index { exit } if } ifelse 3 add } forall

to scan a bitmap line from its border to just beyond its last red
( or white ) white outline pixel. Until I found out it was actually
slower than the earlier simpler and more obvious "for" loops.
Turns out that anything that gets executed millions of times is
extremely sensitive to the number of procs involved. Further,
a newly needed "prep proc" added to the speed penalty.

In my own case, ultra obsession with continuing development can
usually be traced to avoiding other tasks that are far more urgent and
definitely need done now. Such as our eBay listing backlog of nearly
a thousand items.

Speaking of which, our first "real" demottling app appears here.

I guess I have become overly enameled with the concept of
"throw another million calculations at it". Simple math repeated
over and over again a ridiculous number of times now becomes
a routine new ( but previously unthinkable ) solution tool.

We have lots of examples of this in our Magic Sinewaves, and
especially in our new ultra fast Magic Sinewave Calculator.

More recently, the present vingetting corners in our auto
bagkground mottler seemed to be to be usuful, but very
slightly rough looking.

So, I decided to treat them as a full electromagnetic field problem
and use a zillion calculations to improve their smoothness. And
all done within a very few seconds at most.

The general concepts appear here and here. You exactly define
the edges of your rounded corner. Directly for the straight parts
and using the magical PostScript infill operator for the curved.
The rest of the corner map array gets filled with any old number,
preferably midrange.

You then use a classic field technique called relaxation. But made trivially
easy here. Each internal luminescence point gets replaced with an average
of its four nearest neighbors. After many dozens to a few hundred repeats,
the accuracy gets good enough that any errors are unnoticably small.

I'll try to post the updated code shortly.

Added a greatly improved optional vignetting capability
to our new auto background mottling code. Here's a
typical example. Both the autobackgrounder and the
vignette together take less than four seconds total on a
1000x1600 bitmap! With further speedups possible.

Besides giving a "richer" background" and obviously
sharpening subject edges, mottling also dramatically
reduces JPEG edge artifacts. A subject bitmap gets white
or red fully outlined and the filenames and directories are
changed in the mottbk1 .psl utility. The utility is then sent
to Acrobat Distiller. being sure to use acrodist-F from the
command line to permit disk work. And using double reverse
slashes in any Windows filenames.

I chose to use a "redscreen" rather than the more comon
bluescreen. Either will also respond to 255-255-255 true white.
A PostScript array-of-strings makes a very useful data structure
that is both compact and powerful. Each string row is kept in a
bgrbgrbgr... bottom up format rather than separating the color
planes.

The mottling algorithm begins by starting at the west margin and
replacing all pixels up to and including one or more of a
continguous series of red 255 pixels. Replacement consists of
a pseudorandom start of a pseudorandom mottling sequence.

The algorithm then similarly continues from the east margin but
rejecting any full lines already mottled for speedup. A data array
is then created that keeps track of the start and end of each
processed line. This dramatically reduces the number of "across
the grain" pixel manipulations required by the north and south
sequences. Again for a significant speedup.

South to north column and then north to south processing follows,
working only with essential remaining pixels. After the background
knockout is complete, a separate module scans for any remaining
true red or white pixels as might be present from preprocessed
undercuts, bolt holes, or subject gaps. These are similarly replaced
with a suitable pseudorandom mottled background.

The optional vignetting begins with a corner map of appropriate size,
generated manually, by scaling the others, or by using our unique
field mapping techniques. Each map element sets the luminance scaling
for its underlying rgb pixels. The four corners are placed, followed by
the remaining vertical or horizontal edge elements.

Additional image post processing info here.
Consulting services available

There's a bizarre collection of "castellated" CCC dam projects
in a seldom visited area an eighth of a mile northwest of the
Discovery Park turnoff. Such as this example.

There's at least six of these. Typically, they consist of an
earthen dam about six feet high and dozens to hundreds of
feet long. Each end will often have a hand laid stone "tower"
protecting the dam ends.

From either end, a cable runs many hundreds of feet. Tied
into short chunks of vertically driven railroad track. And
supporting a mostly buried two foot high steel fence.

Apparently the goal was to spread and slow flood waters.
For no apparent rhyme or reason.

To me, these appear almost ( but not quite ) as monumentally
stupid as more modern economic stimulus spending. At least
some Texan young men got their first pair of shoes out
of the deal.

Here's some of the improvements I'd like to see in eBay...

First and foremost, drop any and all restrictions to having
email addresses in listings or discouraging direct contact
between buyers and sellers. The perceived occasional lost
listing fee ( which they probably are not entitled to anyway )
is largely mythical. And utterly overwhelmed by the bad vibes
of discouraging and complicating personal contact.

Second, eliminate any hassles over placing "do not bid" or
words to that effect in any listing. Many items ( especially
industrial ones ) might have severe safety issues or require
special expertise to even understand what they are, let alone
intelligently use them. In these cases, a warning to the clueless
is essential to head off the bad vibes of returns or refunds.
Or worse.

Third, add a new optional twenty five cent autosnipe feature.
After auction close, all the autosnipe bids would be viewed,
and the highest that exceeds the auction bids wins by an
increment. This converts eBay into much more of a
Vickrey Auction and very much benefits all of the buyers.
Especially since the buyer now has to reveal nothing to his
competitors and the value of their time ( or having to be in
a certain place at a certain time ) is much more fully
respected. As is the ability to deal with multiple closures.

Fourth, the abuse potential of eBay Bucks is off the wall.
A number of scamsters are sending out eBay looking bulk
mailings that suggest the receiver has several hundred dollars
of cash benefits that must be spent immediately. Fix it or
flush it.

More here.

It is usually difficult to get excited about any new math
book. While there may be a great cast of characters, the
plot development is usually rather thin and the story arc
is often somewhat predictable.

Not so with Mathematics 1001. Which I literally could not
put down and promptly ran out and bought two copies of
after the second time the library copy was overdue.

What you have here is basically a summary of everything
mathematics. All in concise and bite sized chunks.

There's also a Science 1001 which I have yet to look at.

Beware of eBay Bucks scamsters.

The original and legit eBay Bucks program seems to be a copy
of the usual VISA/MC rebates The scam comes about when you
receive a bogus email saying you have a few hundred dollars of
credit that you must immediately spend

Your first clue is usually that you would have had to have spent
$3000 online for each $100 credit. In our case in two separate
scams, we were not remotely near that.

A second clue is that your credit is about to immediately expire,
rather than giving you the month or so for a valid claim.

I'm not sure exactly how this becomes a scam, but it almost
certainly involves your submitting sensitive or credit card info
when you respond. .

Be sure to send a copy to spoof@ebay.com if you get hit.

This seems "too good to be true". But it appears legit.

I seem to have found a way to speed up all of our image
postprocessing utilities ( especially motbk1.psl ) by a factor
of five! Automottling a 1500x1000 bitmap background now
takes three seconds, rather than fifteen!

Apparently Acrobat Distiller will spend as much as 80 percent
of its processing time reclaiming vm if your program has great
heaping bunches of stores and lots of array or string reuse.

This is likely done to minimize memory leaks and to place as few
demands on the host system as possible. And "they" do warn you
to disable vm if you are timing a proc.

At any rate, a -2 vmreclaim will temporarily disable any garbage
collection.  A 0 vmreclaim will restore it, and a 2 vmreclaim will
force an immediate use.

You do want to have lots and lots of available system RAM and
few competing active programs if you try this stunt.

Our new mottled autobackground utility now seems good
enough to release. You can find the preliminary code here
and a sample input bitmap here.

The code replaces hours of manual bitmap retouching with a
few seconds or less of fully automatic operation. It accepts a bitmap
that initially had no unwanted true red 255 or true white pixels
but has otherwise been carefully outlined in red 255 or true white.

Optionally, any interior areas can also be true red or white filled
if they have undercuts or need boltholes or whatever.

Any unwanted background or desired punchthru is then replaced
with an attractive mottled background of your choice. Besides
looking quite good, mottling also dramatically eliminates any
later JPEG distribution artifacts.

Many mottled background examples appear here.

The bitmap is then placed in an accessible location and motbk1.psl
is renamed and edited to have the proper filename and filename prefix.
And then saved as an ordinary textfile, albeit with a .psl posfix

A chosen color is then entered in a [123] ( for 216 web friendly PostScript
or [36 179 205] ( for all available bitmap colors ) format. motbk1.psl
is then sent to Acrobat Distiller, preferably from Acrobat X or higher.

It is super important to run Distiller from your command line, using
an acrobat-F command. This is required to permit the disk access essential
to the background mottling operation. More on this detail here.

It is also super important that there are absolutely no breaks in the true
red or white outline. The slightest gap will cause row background lines to
appear in the portions of the image you wish to preserve.

Mottling "depth" or variability can be adjusted with a /mv variable. A mv
of 25 gives a ten percent color range from a scale of 255. Note that
mottling decreases slightly on any later image downsizing.

More on PostScript here. And our older image post processing utilities here.
Consulting services available.

An early heads up...
================================================
"Gila Valley Day Hikes" subject of Saturday December 10th Free
Discovery Park Talk
================================================
Local author and researcher Don Lancaster returns to Discovery
Park this Saturday December 10th at 6:30 PM in the Jupiter Room.
As part of director Paul Anger's ongoing fall lecture series.

The presentation will zero in on some of the "secret" little known
insider exciting things to do within or near the Greater Bonita-
Eden-Sanchez metropolitan area. Such as the 60 miles (!) of newly
rediscovered prehistoric hanging mountain stream fed canals. Or
our own El Capitan slot canyon. Or the impressive cliffs of
Fisherman's Point. Or the phyically challenging San Carlos Falls.

Or the UFO "fish fillets" of Taylor Canyon.

Other topics might include the Montez and Weech toll roads, the
loops and tunnels of the former Morenci Southern Railway, the
tens of thousands of our prehistoric agricultural grids, seldom
visited Oak Grove Canyon, extremely remote Hanna Hot Spring,
and, of course, the Mount Graham Aerial Tramway.

Many of these dayhikes can now be easily previewed at Don's
< https://www.tinaja.com/gilahike.shtml > web page.  More than
365 intriguing destinations are listed, so there should be enough
here to keep you busy for at least the next year. Additional details
can be found at  < https://www.tinaja.com/tinsamp1.shtml >

There is no charge and families with children are welcome.
Discovery Park is located at the corner of Safford's Discovery
Park Boulevard ( formerly 32nd street ) and 20th avenue.
You can get more info by calling Paul or Jackie at (928)428-6260
or visiting their < discoverypark@eac.edu > website

Per this Extrasolar Planets Encyclopedia, the number of known
and verified extrasolar planets now number 704. While the
present reports are averaging two per week, these are shortly
expected to start arriving daily or faster.

This suggests that the unavoidable and inevitable discovery
of life elsewhere in the universe is likely to occur much, much
sooner, than many people suspect.

A reminder that several years ago, I published an incredibly
heretic and ridiculously simple approach to solving electromagnetic
and similar fields. One that completely eliminates virtually all
of the messy math.

Per this tutorial and this one. And this animated demo.

Uh, you may have to download the demo before you can run it.
There may be issues with the new pseudo PDF readers in some
browsers.

Expanded and updated our Gila Valley Dayhikes library page.

I remain fully convinced that the E-cat fiasco will remain
forever just that. The longer a "looks like a duck, quacks
like a duck" ludicrosity continues without strong and
independent verification, the greater the probability of
it ending up something between a scam, a hoax, or plain old
rotten labwork.

Much more on spotting pseudoscience here.

To me, finding a low cost unlimited source of free energy
would be one of the most unimaginibly henious crimes
against humanity. Making Hitler look like Mother Teresa.

For it would convert the planet into a cinder. Global warming
would likely be measured in days, rather than centuries.

It seems to me that the overwhelming benefit of solar derived
energy ( such as upcoming anticipated pv developments ) is
that it minimizes upsetting the heat flux balance of the planet.

Much more here.

Know when to walk away, know when to run.

I frustratingly decided to pass up an intriguing auction
opportunity involving a fascinating item. A local auction
had hundreds of automatic lubricators scunging away in
obviously long stored but original packages. Which
likely could have been stolen for pocket change.

These were basically a "slow airbag" that pressurizes a
fist sized grease container in response to a programmable
current off of two AA cells. The current activates an
electrolytic azide cell to generate nitrogen gas which pushes
against a bellows, delivering long term grease to a bearing
or whatever.

Problem #1 was that despite web data, nobody but nobody
would tell me how much these cost. Apparently you have to
prespecify what grease you want in the unit at time of
manufacture. The electronics involved was rather trivial,
consisting of a dollar's worth of DIP switch, transistors,
and diodes. There was no obvious repurposing potential
that I could see. And it wasn't clear whether our usual rule
of "one sixth list price" would be worthwhile.

Problem #2 was there is a specified shelf life of one year.
Several units had obviously corroded batteries and these
were obviously ancient. A key question is whether there
were lifetime factors for the electrolytic azide cell as well as
the batteries. Thus, even with new batteries, the units may
or may not work. With no guarantee for how long.

Problem #3 are the hazmat warnings. Activated units can
have and retain up to 50 psi of nitrogen pressure. Which
is comparable to puncturing a paint spray can. Opening
the top to replace the batteries could inadvertently spray
grease all over hell and gone.

Problem #4 is that the demands of automatic lubrication are
rather customer specific. Bottom line is that there ain't much
demand, and what demand there is would be unlikely to buy
from anything but a known reliable and proven source.

And, of course, Problem #5 would be that a customer might end
up mightily pissed if their $50,000 machine froze because its
bearing lubricator failed to work properly.

Sigh.

More on auction stuff here.

Expanded and updated our Auction Help Library.

Your own custom auction finder can be creasted for you
per these details.

Curious factoid of the day: The Canary Islands were
named after a large dog.

One of the most maddingly infuriating happenings in all of
PostScript occurs when the contents of old strings mysteriously
change later on in unexpected ways.

The usual problem is failing to understand that PostScipt does
NOT place strings into arrays or other data structures! All it
places are POINTERS to those strings!

Thus, if you reuse a string ( such as a disk reading workstring),
much if not all of your previous string storage can get corrupted.

Note further that PostScript string contents are NOT preserved
or protected by saves and restores!

The subtle but simple cure for these hassles is called string
dereferencing. Any time you are about to use a string that may
change later in your program, you take a "snapshot" that
creates a new and unique string that will not change…

(oldstring) dup length string cvs --> (safestring)

Pointers to your dereferenced "safe string" can now be placed
in an array or whatever. (safestring) will stay exactly the way you
created it. And your (oldstring) is free to go on to other tasks.

There are speed and vm penalties for string deferencing, but
these are usually fairly modest.

Othere sneaky tricks involving PostScript strings, arrays, integers,
procs, and dictionaries can be found here.

Beware of clean new original boxes at any auction.

Many times, these are just a convenient stash for a
upgrade or a broken item that got replaced. One
immediate clue is some or all of the original packaging
is missing.

A second is any disparity between the box description and
what is stashed inside.

In an older example, many boxes of slot machine replacement
wheel assemblies were older units needing refurb and repair.

Or in yesterday's auction, bunches of squeaky clean new
monochrome laser printer boxes held older and worn
color units from an obvious upgrade. Units with very little
current demand.

More secret auction insiders stuff here.

I'm trying to work up an automatic mottled background
filler-inner for our eBay photos.Vaguely similar to our
white knockout routine.

The algorithms end up nontrivial because of speed restrictions.
The goal is to be reasonably fast on a 1300x1300 bitmap and to
fill in at least 98 percent  that is outside of a red or white
unbroken outline.

The PostScript array-of-strings structure seems to be a good
starting point. It is a very compact way of representing an
array of 0-255 integers. And quite good and fast with "along
the grain row putintervals and such. But "across the grain"
column modifications ( where, say, column 43 of 1300 different
array strings needs work ) should clearly be avoided.

Generating a random mottle for each pixel is clearly ungood, so
a small 5000 entry mottle array string can be used instead. Access
for getintervals can be randomized to eliminate any unwanted
patterns. There are two reasons for mottling. First, it looks good,
and second, it dramatically reduces JPEG edge artifacts.

An EWNS strategy can be used. First, each row is scanned
from the left until a red ( or white) pixel is reached. Scan continues
until no continiguous red or white pixels are found. A suitably
selected subrow of mottle can then be substituted all at once.

Then the same line is scanned from the right until either a full red
or a white pixel is incountered If no hits are found, the scan stops
at the limit already handled by the EW scan. This can eliminate
entirely as many as one half of the EW scans.

Now for a tricky detail. A two array map of the scan widths is saved
as well as the maximum full east and the minimum full west column values.
These can be used to dramatically restrict the needed column manipulations.
Perhaps by a 100:1 reduction in the number of column pixels involved.

If an easternmost or westernmost column is already filled top to bottom, it is
not NS scanned at all. A suitable starting point for a NS or a SN scan is found
based on values derived from the two map arrays.

I'll try to post some code shortly. Other image postproc tools here.

There are incredible new opportunities for amateur astronomers
to make significant world class contributions. A number of services
now offer "instant" web based time rentals on quality instruments with
charges starting near the hundred dollar per night range.

Examples include pioneer Tenagra Observatories, Plus newer
LightBuckets, Global Rent a Scope.

If you are local and have more casual astronomy interests, winter
tours of the Mount Grahm International Observatory are offered
by Discovery Park. While the Desert Skygazers can give you personal
and unlimited free access to a 24 inch professional telescope. You can
even become a docent or group tour telescope operator by taking an
easy EAC course.

Or check out the smaller hands on instruments available at
Casitas De Gila. Their real time planetarium simulator above
the hot tub is especially impressive.

Other local or nearby things to do here.

In another major step towards freeing up older technical
information from the outrageously overpriced and restricted
Scientific Journal Mafia, The British Royal Society has
freely posted 60,000 older papers, making them all
permanantly free online.

Now, if the IEEE would only take the hint...

The one that really pisses me off is the unavailability of
Sallen and Key's seminal 1955 active filter paper.

Three words when it comes to dealing with auction
pallets, "contents of cabinent" or "contents of room":

Triage
Triage
Triage

A fascinating free reprint of the older Navy Electronics
Training Series can be found online.

Much of the content is dated beyond the point of being quaint. But
interesting reading nonetheless.

Did yet another revision and update to our Bitmap Typewriter.
The new code can be found here and the old tutorial here.

The new commands appear in the program listing.

New is a much faster random number generator for the
mottled background options. Plus a combined color
selector that lets you pick rgb or 216 color inputs.

A convenient array listing of all available fonts is
included, but you will have to revise this for your own
machine. Getting a combined list of all PostScript and
other fonts seems tricky, and I am still seeking a good
solution or utility for this.  

How many mathematicians does it take to change a
light bulb?

Only one. Who hands the bulb to five Californians, thus
reducing the problem to a previously solved riddle.

A second local restaurant just totally trashed their buffet,
eliminating the salad bar part entirely. While doing an
ultra spartan ( and obviously MacDonaldsesque ) stark
remodel that totally ruined their overall vibes.

About as sharp as five pounds of raw liver.

Here's a revised and updated summary of our ongoing Safford
Basin prehistoric hanging mountain stream fed canal research to date.

A total of 17 canals are reasonably known or believed to exist
with an estimated  total distance exceeding 60 miles! With
engineering clearly beyond world class.

From east to west...

0. - P Ranch - remains speculative with only a short
V section of canal, favorable terrain, and persistent
rumors. Not included in totals.

1 - Ledford - total length of 3 miles. Difficult access.
Initial hanging portion from Smith Dam. Needs
further exploration. Studied and reported on by Neely.

2 Henry's - Recently visited portions end in modern
stock tank. Connection to Marijilda unproven. The
known length is 2 miles, likely length is 4 miles.

3 Lebanon - A modern canal that likely closely follows a
prehistoric origin. Marijilda fed east from Lebanon
Reservoir #2. Possible length is 2 miles. Prehistoric
origins not yet conclusively proven.

4. Marijilda - Well researched with spectacular hanging
portion. Historic rework flows to this day. 5 miles long.
Includes "aquaduct" portion a meter high by a hundred long.
Studied and reported on by Neely.

5a. Twin Boobs West- Fields area appears to be fed by two
distinct canals. Western "clockwise" feeder may be derived
from Marijilda via Rincon Canyon. 3 miles estimated length.
Parallel feeders newly discovered in the Twin Boobs area.
Extensive rock alignments, grids, mulch rings, field houses,
and similar ag development.

5b. Twin Boobs East- Apparently sources in the Lebanon
Resorvoir #2 area, then climbs "up" and "under" the Labanon
Cemetary leading to a terrain following spectacular "S" turn.
Strong hanging portion seems to end up feeding the Twin
Boobs area from the east in a "counterclockwise" direction.
Length at least 2 miles. Slope remains to be proven.

6. Tranquility- Amazingly close to 5b, approaching within
30 feet easterly and vertically. Might be artesian sourced.
Somewhat over one mile long. Apparently routes water to
fields in the Cooks Reservoir area. Portions rebuilt to
puddled high aggregate concrete similar to the Marijilda
rebuild. A strong candidate for proving "stole the plans".
Significant hanging portions.

7a . Deadman - Initial hanging portion speculatively covered
by modern pipeline. Possible 3-way switch mid channel.
At least 5 miles with possible extensions into Longview area.
Original canal flows to this day.

7b Longview - Hints of canal structures in the Freeman Flat
area below Longview Ruin. Which could have been Deadman
fed via switching. Highly speculative. Not included in totals.

8. Robinson - Reworked to feed cattle tanks by anglo
pioneers. Excellent topo traces, well hung. Destination
not currently known, but likely Robinson Flat.. At least
3 miles long.

9. Allen - Sources from Spring canyon, runs under Allen
resevoir to an unknown destination. Several key but
small segments unproven and unexplored. Exceptionally
large "Culebra" cut two meters deep and a hundred long.
At least 6 miles long.

10. Mud Springs - Sources in upper Ash Creek and impressively
climbs "up" the Mud Springs bajada. Some impressive hanging
portions and a companion structure. Several pieces missing.
Believed to source the Jernigan Canal. Final destination
unknown. At least 7 miles long. Ten inch diameter mesquite
tree midstream and overrun by SCS dam strongly supports
prehistoric origins.

11. Jernigan - Apparent branch of Mud Springs seems to terminate
in the Central Dump area. Has a triple "U" turn to maintain
slope. Small but crucial portions missing and unexplored.
Possibly 3 miles long. Possible branch with short aquaduct.

12 Shingle Mill - Newest "rediscovery" appears to source
in Shingle Mill canyon and deliver to the Cluff Ponds area.
While only a few hundred feet have been visited to date,
projected length is 3 miles. Still under active exploration.
Modern rebuild as the Minor Webster Ditch System. Prehistoric
origins remain unproven, but in a compelling location for total
Mount Graham stream water recovery.

13. Lefthand - Studied and reported on by Neely. Includes
plant nurseries and other intensive ag. Projected length is
2 miles. Possible associated canal feeding Lamb Tank.

14. Carter/Nuttall - Possible continuum with other
hanging canals remains largely unproven and unexplored.
Includes Taylor Terraces which likely are CCC
but may have prehistoric origins

More hanging canal info here.

Expanded and updated our Gila Valley Day Hikes web
page.

We are now only 28 listings short of our goal of having
365 major entries and over 500 things to do.

Please email me with anything I missed or any
corrections or suggestions.

Water electrolysis for bulk energy apps seems to be once again
rearing its ugly head in the energy and electronics newsgroups.

One more time: There is a fundamental thermodynamic reason
why electrolysis for bulk hydrogen is absolutely guaranteed to
ain't gonna happen.

Because of a profound limitation called "exergy", a kilowatt hour
of electricity is ridiculously more valuable than a kilowatt hour of
unstored hydrogen gas.

Thus, the process of conventional electrolysis is pretty much
equivalent to 1:1 exchanging US Dollars for Mexican Pesos.
Nobody could possibly be that stupid.

Much more here and here.

A reasonably unbiased analysis of the e-Cat fiasco
can be found here. My own view, of course, is that
this is utterly and totally and ludicrously bogus.

Based on "Looks like a duck, quacks like a duct" with
the same lameB-movie wannabe actors appearing over
and over again.

ANY free energy scheme demands that extraordinary
evidence is required for extraordinary claims.

AND that the claimants should be spending exceptional
effort to prove themselves wrong.

But on the basis of "once bitten, twice shy" proof
of cold fusion related anything wiould absolutely
have to go orders of magnitude beyond utterly,
unambiguously, unquestionable, and totally convincing.

Anything less than Walmart Aisle #13 being sold out
guarantees a lack of credibility.

More on bashing pseudoscience here.

Rediscovered yet another of our prehistoric mountain stream
fed hanging canals. Depending upon who is keeping score, this
makes Canal #17 for a total length over sixty miles!

We might call it the Tranquility Canal. It may be artesian sourced
and likely delivered water to prehistoric fields in the Cook's
Resevoir area. It seems a little over a mile long. Portions have
apparently been "steal the plans" rebuilt with a puddled high
aggregate concrete lining remarkably similar to ( and likely
contemporaneous with ) the Marijilda rebuild.

There are hints of an old canal beside the rebuild, and this
instance may be a useful route to proving that nearly all of
the pioneer mountain canals were "steal the plans" rebuilt by
digging out an old ditch. Such evidence is sorely needed in
the Minor Webster Ditch Systen refind.

Curiously, the canal comes within thirty feet (!) easterly and
vertically to the Twin Boobs eastern feeder canal. But the
two appear totally unrelated. There are several other instances
where nearby canals only differ by a few feet elevation, yet
may take miles to build up the needed elevation offset.

A major and apparntly new and free source of user and
repair manuals can be found here.

And some HP specific sites that seems to be filling the void of
"discontinued but not ancient" info can be found here
and here.  And here. And here.

A number of years ago, both HP and Tektronix released all of
their older manuals to the public domain. You are thus free
to scan and publish them.

How can you prove if an old canal is prehistoric?

My own belief is that virtually all of the pioneer hanging canal
activities simply "stole the plans" and just "dug out an old ditch."

Especially if only a fraction of the canal was "repurposed" by
feeding a new cattle tank or irrigation reservoir or becoming
part of a city water system. And since evidence supports the
premise that the prehistoric populations was much, much higher
than the few hundred pioneers in the area.

It certainly helps if there are large cacti or ten inch mesquite
trees repeatedly growning mid channel. Or that 1930's dams
and similar projects ran roughshod over the canal routes without
any accomodation whatsoever. Same goes for obviously older
roads or fences.

Desert varnish, patina, caliche, and lichens also reveal distinctly different
patterns with their age of disturbance.

Under certain circumstances, fingerprints can last a rather long time.
But certainly not multiple centuries. Could new forensics find
fingerprints under rocks that the CCC could have left but clearly
could not have been prehistoric?

The goals and purposes of the canals also seem to me to much
better support prehistoric goals of small dispersed fields, rather than
the modern ones of cattle wattering or large linear ag areas. And
with thusands of Gila River acre feet safely available to modern users,
there did not seem to be all that much point in exploiting every drop
of water from minor mountain streams. Unless the canal was already
there and nearly ready to reuse.

Even the earliest anglo pioneers had access to concrete, iron, wheelbarrows,
winches, prybars, mules, and, of course, picks and shovels. Many portions of
many canal routes do not seem to support even the slightest evidence
that such tools were being used in any manner.

But I very strongly feel that the canal architecture itself can strongly
support a prehistoric origin. The canals are often hung on mesa sides
to make their slope independent of terrain. There is a conspicuous
absence of long linear segments.

Most of all, the prehistoric population had no horses or other means
of conveying rock or dirt moderate distances with a reasonable
energy expense. And certainly no wheelbarrows, scrapers, or
Gradealls. Many of the canals also have a conspicuous absence
of any companion service road or maintainence trail.

Thus, most construction activity went ACROSS the canal rather than
along it. With cuts and fills avoided when and where possible.
And an obvious major goal being maximizing the energy efficiency
of construction.

The canals thus tended to follow an "optimal terrain" and "minimal
energy" routing rather than a "shortest distance" one.

Even if it meant adding a mile of canal simply to gain or lose a few feet
of elevation. There are also several instances of very elaborte
"U turn" or even "S turn" routes that could easily have been avoided
with more modern ( but far more energy intensive ) tools and techniques.

Found a curious happenstance in my new Garmin eTrex 30.
Not sure if it is a bug or a feature.

On a supposedly turned off unit, the display flashes for a
fed milliseconds every ten minutes or so. Very disconcerting
the first few times it happens.

This could be an emergency locator beacon for a dropped or
lost unit. But only works if it fell rightside up and is not
covered.

Here's a cute variation on the random command in our
Gonzo Utilities that is much faster and simpler but has a
curious glitch that may need dealt with.

It is well suited for such tasks as mottling the background
of a bitmap...

/wantedrandnumber 15 store

/randfact 2 31 exp 1 sub wantedrandomnumber div
floor cvi store

/fastrandom {rand randfact idiv} store

The glitch is this: Doing a 15 fastrandom will return fifteen
equiprobable digits in the 0 to 14 range. But it also may
exceptionally rarely and highly unlikely instead deliver a 15.

Perhaps something around seven times in 2^31. Do not
use this speedup anywhere this could create problems.

Added another Dr. Neely paper to our Tinaja Questing
Library.

This one is on the Gila Valley lowland canals.

The Garmin eTrex-30 can in fact real time register a one foot
altitude display change for a one foot raising or lowering of the
device.

You do have to preset the Barometer Mode to Variable
elevation and be in the Trip Computer mode.

I am still not sure whether this will be good enough to
resolve some crucial prehistoric canal slope issues.

Its user manual can be found here.

How much resolution is "enough" for an eBay image photo?

We pride ourselves in having the best eBay images bar none, so
here are some of the guidelines that we use...

I stongly feel that at least an hour should be spent in image
postproc, and that the majority of your creative effort should
be spent after the image is captured. Naturally, everything
up to your final out-the-door image should remain in the
bitmap format.

If a scanner is appropriate, we scan to 200 DPI for larger items
or to 600 DPI for integrated circuits. The latter gives us about a
6X magnification and leads to superb lettering during posproc.
With or without use of our Bitmap Typewriter.

With a camera source, we use multi-step resolution that depends
on what the current postproc step needs. We start with a ten or a
twelve megapixel image. This allows significant cropping while
keeping useful results.

After a preliminary and somewhat oversize crop an image in the
2400 pixel wide range results. If multiple exposures ( such as to adjust
side lighting or improve a nameplate appearance ) are needed,
they are done at this time. The reason being that any distortion
correction ( and its consequent resizing ) is best done only once.

We are very big on correction to Architects 2-1/2 D perspective in
which every line intended to be vertical ends up exactly that way.
This takes considerable resolution. We use my custom swings and
tilts routine and Imageviewer32.

After tilt correction, we reduce the image to something around 1300
pixels wide. Which is a good compromise between good background
knockout and taking forever. While we have several automated
background procs, we usually end up doing this step manually,
starting with our knockout.bmp patterns.

After a high quality bitmap is exactly the way we want it, we
recrop then reduce it to 700 pixels wide, adjust the gamma and brightness,
and possibly do a color correction and a very slight sharpening.
And then convert to JPEG for smaller file size distribution.

We have a superb vignetting option, but have not used it much
lately.

eBay will then resize to 200 pixels or so but still allow magnification to
the 700 pixels needed for most industrial products. Such things as
test equipment font panel details are instead made much larger.

Consulting, seminars, and custom services available.

Adobe's Acrobat X offers a fairly easy way to "tour" all of
the fonts available on your machine. Move any PDF file with
a large line of text into Acrobat X, select Tools > Edit Text
& Objects > Edit Document Text.

Then highlight the large text line and right click on properties.
As you change the font name, the displayed line should change.

The only little problem is that I have not found any fast and
easy way to cut and paste the font selections. Hand keying or
screen grabbing gets old if hundreds of fonts are involved.

I'm working on a font sampler utility and a way to include
it in our Bitmap Typewriter.

A reasonable question is that "If I am having so much
difficulty measuring the slope of a prehistoric canal,
how did they do it?"

No obvious instruments seem to survive, but it seems to
me that there isa possibility that the canal itself couold
form a water level.

Assume a one percent slope is desirable. Cut or find a
stick 25 inch equivalents long. Make a mark or notch.
at 3 inch equivalents. Use this as your canal standard.

Assume a properly sloped canal segment is being extended.
Measure off twelve stick lengths and dig the new 25 foot
channel. Chances are a very good slope guess can be made
based on the previously completed 25 feet through simple
sighting.

Just to be sure, run just enough water into the new portion
so it just barely overflows at the far end. The water level
at the start end of the new portion should be three inches
lower in 25 feet. Or one notch short of overflow.

If needed, adjust the slope accordingly.

A fascinating and highly eclectic collection of magazine
article reprint listings can be found here. Bunches of ham
radio and hobby electronics stuff is hidden in the listings.

It would be even better if there was a "directory or everything"
for anything that ever appeared in print. Full text, fully searchible
Acrobat PDF of course.

Especially in hobby electronics, any remaining copyrights are
ambiguous if they still exist at all.

The "directory of everything" might print their intent to list
something 30 days in advance, allowing any legitimate
copyright holder to block publication. That same holder
could also withdraw the reprint at any time.

Alternately, a micropayment scheme could be set up
for any legitimate copyright holders.

I remain apalled that legitimate electronic papers cannot be
found on the web with reasonable access.

We seem to be still getting our email return address hijacked
by some Romanians. If you are in this loop, we are sorry but
it definitely ain't us. Nor our ISP.

Whatever system is doing the actual spam generation apparently
shuts down on Saturdays and Sundays. And their lead list seems
super worthless, judging by the hundreds to thousands of those
returned undeliverables we have to muck out of our email.

It is unusual for a hassle of this sort to continue for this long.

The "Tesla Turbine" nonsense is infesting the newsgroups again.

One more time: Bladeless turbines and pumps DEMAND a
thermodynamic irreversibility for their FUNDAMENTAL operation
of maintaining a lossy shear force in a viscous medium. And are
thus useless when high efficiency is a factor.

They, of course, do not work at all if the fluid has zero viscosity.
I'd expect them to be far less efficient in vapor rather than liquid.

These devices ~are~ superb as shit pumps or when pumping frozen
chickens or live fish. As the now defunct leading supplier
tells us, blades have to be added if higher efficiency is needed.

Most of these problems would go away if they were simply called
"shit pumps" rather than naming them after the patron saint of the
church of the latter day crackpots. More in our Tech Musings library.

Particularly here and here.

We are trying to dramatically reduce our inventory to make
room for some upcoming SMT auctions. As a result, we throw
away hundreds to thousands of dollars worth of quality
electronics each week that do not meet our eBay goals.

Thoroughly mixed in, of course, with utterly worthless grunge.

Please let me know if you want any of this at no charge. The
only rule is "take any, take all." Typically a small pickup or
medium SUV worth each week.

You can email me for details.

I'm saddened by the best buffet restaurant for 80 miles
around folding because of a franchise flap. And got
replaced by yet another "me too" overpriced steak
house. Whose ambience not quite qualifies as
mesmerizingly awful.

Meanwhile, Pima now has a new and wildly improbable Copper
Town Bistro. Not sure if the town can support a seventh
food service outlet.

There seems to be a major eBay flap over free shipping.

Naturally, there is no such thing as free shipping. The
customer always pays, one way or another.

While we always attempt to combine a customer's shipping
for the best possible rate, we feel that there are major
problems with free shipping.

The overwhelming majority of ebay items sell for less than $24.
Which means that shipping is often a significant fraction of
the total.

If you offer free shipping, your opening price will thus
have to be much higher than otherwise, giving you a
competitive disadvantage.

Further, free shipping usually penalize your best customers
in that there is no obvious or reasonable way to reduce the
per-unit charges on multiple simultaneous orders.

If the shipping is distance related, free shipping also subsidizes
those further from you at the penalty of nearby customers
who are more likely to become repeat customers.

More eBay related stuff here.

One of the most important things I learned in college was
that the key secret to writing thick lab reports was to use
thick paper.

More on similar topics here and here.

Finally got my new Garmin eTrex 30 today. It is
smaller and significantly better performing than the
original eTrex. I am generally very impressed.

The barometric elevation display seems fairly stable short
term and displays to one foot resolution.

But it apparently can not resolve the difference in a three
foot vertical shift. Which I was sorely ( and likely
unreasonably ) hoping for to resolve some prehistoric
canal issues.

Back to the Brunton.

A reminder that my favorite bed and breakfast of all
time remains the Black Range Lodge, cleverly hidden
in the part of New Mexico that you cannot get to.

Their Kingston Frisbee Festival runs from January 1st
to December 31st this year. And the Percha Creek
Salmon run remains as spectacular as ever.

Two other places of interest are Casitas De Gila outside
of Cliff ( Check out their real time planetarium simulator
above the hot tub. And the art gallery. )

And the Snowy Mountain Inn, which seems to have fully
survived the Wallow fire disaster. But was recently sold
at auction.

There's another possible reason why newer microprocessors with
enormous amounts of Flash Memory might prove very useful for
variable speed or variable frequency Magic Sinewaves: The
filtering can possibly be dramatically simplified if larger values of
"n" are used for lower output frequencies.

Because the first unwanted harmonic frequency could thus be
held pretty much the same.

Yeah, higher "n" would be less efficient owing to more switching
transients. But lower frequencies in a motor ap are likely to be
associated with significantly lower load energy for a net gain.

More on this whenever.

Just came back from a New Mexico trip and discovered a major
new wind energy facility in a dead calm area with zero net generation.

The wind all the rest of the way home was blowing just fine.

Thinking this somewhat strange, I dug up this New Mexico wind potential
October 4, 2011nal, and Poor. This site seems to me to weigh in at Fair.

Thus suggesting that 100 percent of all Superb, Outstanding, Excellent,
and Good New Mexico sites are already fully in use. News to me.

I'm wondering if some county commisioners got greedy, or if
this is an outright scam, or a ploy to steal federal founds. At the very least,
somebody seems to be confusing the nameplate capability of a windmill
with its actual long term net amortitized energy potential.

More on similar topics here.

We've been so busy with current stuff that several back
burner long term projects look like they may never get
cleared. You are welcome to make offers on the following...

Five fairly recent "as new" Dell Servers with many
internal drives that need performance checked
before they can be sold.

A heavy duty industrial programmed wire cutter
that has been 90 percent carefully refurbed but
still needs an on-hand stepper ic replaced.

Great heaping bunches of premium quality alarm
panels that have yet to be sorted and tested.

A pair of eminently restorable 1908 (!) Eastman
commercial 35 mm movie projectors. These
are presently disassembled and UPS shippable

The last remaining developable property on the
north boundary of Gold Hill, Oregon. 20 acres.
Easments and power now in place.

An absolutely spectacular survivalist property on
the East Fork of the Gila River adjacent to
a NM wilderness boundary. 5 acres with access
highly restricted. Taxes are $2.70 per year.

Plese email me with your interests.

At least two of our prehistoric hanging canals ( Twin Boobs
and Jernigan ) now clearly have multiple sharp "U" turns
in them forming major "S" shaped routes.

The apparent goal was to be able to cross a wash or drainage
at grade. The "backwards" and "uphill" parts are certainly
counterintuitive and have beeen very hard to find.

Both the engineering and the societal infrastructure involved
are clearly world class and utterly mind boggling.

Several viewers have requested additional photos
of the professional dimmers we have up on eBay. The new
photo can be found both here and in our eBay offer.

We pride ourselves in having the best photos on eBay,
bar none. This particular image was shot in an area with
skylights, CCFL fluorescents and additional key lighting.

We firmly believe in heavy image post processing, ending
up in a combination photo and characture that accurately and
favorably represents the item being offered.

Some of the steps involved include distortion correction to
Architect's perspective. knockout to a mottled background
with minimum JPEG artifacts, and internal enhancement or
clarification of any murky areas.

Many of the secret insider details appear in our GuruGram
and eBay libraries. And especially in this tutorial.

Princeton has just placed the scholarly pub scamsters on
notice that copyrights will no longer be assigned on any
government funded work. Other schools are likely to
quickly follow suit.

It is waay past time that this happy horseshit has ceased.

Any and all research pubs older than three years should
be freely available to anyone who wants or needs them
at zero or negligible cost.

Wesrch is An example of how things should be.

Many of my own freely accessible papers appear here.

Somebody in Rumania has apparently hijacked our return
email address and has been sending tons of spam to worthless
addresses. We seem to be getting many hundreds of undeliverable
messages per day and the problem is not going away.

We were reasonably certain it ain't us, and an overnight shutdown
of all our computers did not change the volume in the least.

These problems usually go away in a day or so, but this one
seems unusually persistent.

We are sorry if this is causing any problems, but there seems to be
little we can do at this end. Our ISP is aware of the problem.

Several viewers have asked about making foreign sales to
New Mexico, Besides the language barrier and the hassles
at customs, another traditional problem was that their truck
tires were a different size and spacing, requiring everything
to be reloaded at the border crossings .

Fortunately, new reversible truck tires have been developed that
can simply be insided out at the inspection stations.

Additional details here. And here.

Made a subtle and difficult find of more feeder for the Twin
Boobs hanging canal area. In a crucial area right where it
had to be. Subtle but almost certainly definite. In an unexpected
"S turn route" reminiscent of the Jernigan Canal. Hidden in
dense cactus and similar grunge.

One present premise is that the Boobs themselves are a water
use area, rather than part of a "move it elsewhere" water
delivery system.

Evidence seems to me to point to a pair of "normal" two walled
hanging canals feeding the area from both the southeast and
the southwest. Quite a world class find.

The Boobs area itself seems to have far too little slope for
long distance water delivery. There also appears to be
parallel canal channels, and the single walled structures
more suggest "ponding" rather than delivery areas.

No truely single walled structures are known along any
of the other long distance mountain stream delivery canals.
In addition, the Boobs area has grids, rock alignments,
and mulch rings common to use areas but exceptionally
rare in delivery areas.

Curiouser and curiouser.

Salvex seems to have just listed 23 truckloads of Napalm.
Seems somebody at Home Depot discontinued stocking the
product after that they discovered that the stuff burns.  

But a maximum output frequency limit on Magic Sinewaves
definitely still exists. The techniques are best suited for
solutions in the 60 Hertz area, and even 400 Hertz might
be pushing it.

The reason being that, in order to zero out any arbitrarily
chosen number of low harmonics, a clock to output frequency
ratio of 44,000:1 or higher is required. In a PIC system, a
clock frequency of 10 MHz can give you a 60 Hertz output.

Thus, magic sinewaves are superb for power apps but unsuited
as a general purpose higher frequency approach to effieient
digital sinewave synthesis.

A website visitor asked what the frequency resolution of our
Magic Sinewaves can be. The glib answer is "anything you
want it to be".

It depends on whether you are using a fixed or a variable rate
clock. In the earlier designs, a PIC or similar microprocessor
had its clock frequency varied to set the output frequency.

This got done with a VCO or whatever and easily could hit
an 0.01 Hertz resolution. But took a "two stage" solution
where the frequency circuit was separate from the amplitude
circuit.

In newer microprocessors with huge amounts of flash memory, the
frequency and amplitude can be internalized into one larger data
file. For instance, if one hundred frequencies and one hundred
amplitudes are needed with eight data bytes per Magic Sinewave,
about 80K of flash data memory would be needed for table
lookup storage.

Did another revision and update to our Bitmap Typewriter.
You'll find the new code here, the older tutorial here
and general details on using PostScript routines here .

The code additions let you have a mottled background.
This looks generally better and "richer". It also further
minimizes any eventual JPEG edge artifacts.

The new commands are...

/mv - set mottle variability ( 25 recommended )
/mw - set mottle width ( 256 for speedup )
/ma - mottle existing background "A"
/mb - mottle existing background "B"
/mc - mottle existing background "C"
/md - mottle existing background "D"

Mottling the entire background can add about ten seconds
to the processing time. You can reduce this 4X by using only
A, B, C, or D. You can further reduce this 4X by setting the
mottle width to 256 rather than the full 1024. Speed penalty
becomes negligible with these combined speedups.

Much more here and here.

Further updated and expanded our Gila Valley Day Hikes.
We are now only 28 entries short of our goal of 365 unique
recommendations. 

Please email me with anything I may have missed, or with
suggestions for further expansion.

Here's an update on the prehistoric hanging canal status:

Some spectacular recent finds have newly identified a
stunning group of mountain stream fed hanging canals
in the Gila Valley. Over a dozen examples are now
known whose total length is approaching fifty miles!

Funding and ( especially ) field mice are needed to diccover,
map, and document this unique world class discovery of
utterly mind boggling prehistoric engineering. Whose
goal appeared to be to literally exploit every available
drop of Mount Graham water.

A present problem is picking up some precision elevation
measurement capability. Besides a good old Brunton
compass, we now have an automatic level that needs
a tripod or adaptor, and a Garmin eTrex-30 which has
yet to be delivered.

Some current issues...

TWIN BOOBS CANAL - This now appears to be a use rather
than a delivery area, owing to parallel channels and rock
alignments. Several exact key locations remain unresolved,
although the area appears to be fed both from the east and
the west.

ALLEN CANAL - Several reaches remain unexplored, as
does its ultimate termination. The mesa to dam segment remains
particularly frustrating. The "Culebra Cut" is spectacular.

JERNIGAN CANAL - Two large holes are conspicuous
in the probable route. In addition, there is a not yet interpreted
possible aquaduct and a possible prehistoric headgate that
still need extensively studied.

MUD SPRINGS CANAL - A key early crossover portion has been
located after several false starts. The exact Ash Creek supply
point remains to be found, and an obvious mile remains
unexplored. There is also the 3000 foot "Black Hole of Central"
and determination of just where this whole work is headed.

SHINGLE MILL CANAL - Extensive historical rework has
pretty much been established as fact. The issue is whether
the plans were stolen from a prehistoric original. The
location, slope and overall plan strongly suggest so.

Please email me if you want any part of this.

Another fireground issue that needs a modern high tech
solution: Accurately identifying individual fire hoses.

A typical department might hand stamp a numeric id
into each hose and then hand read and hand log them
during repacking. Instead, the system cries for rfid.

It sure would be nice to push a button on a hand held
remote that gives a full inventory of an entire packed
hose bed at once.

The id tags, of course, have to be exceptionally rugged
and withstand being drug down the street without
coming loose or failing to respond. After years of abuse.

The market should be huge if an economic and
reliable solution can be found.

I'm wondering if some opportunities exist in updating fire
sprinkler heads from nineteenth century technology.

Two problems are the long time delay betweeen the time the
sprinkler trips and the time a system pressure alarm can
respond. Plus the flood damage during the Keystone Kops
activities between the time the sprinkler needs shut
down and the time the entire system can be drained of water.

Yeah, there are wedges, and wedges on poles. But they are
quite awkward to use, especially fifteen feet or more in the air.

Instead, a new sprinkler head might immediately report its
being tripped. And then might either automatically shut down after
a time delay. Or else have a convenient lever or whatever style
shutoff.

Reliability and cost and approval issues might be severe, but
some high tech solutions seem long overdue to me.

We just picked up a bunch of power utility surplus
instrumentation and similarquality items and should shortly
be listing them as eBay auction or stores.

Much more on eBay here.

Something vaguely akin to dollar cost averaging can prove
useful at an industrial or other auction.

Say you see a dozen pallets you want that you value just above
a currency resistance threshold at $55 each. Say futther that
you end up stealing the first two pallets for $5 each.

This frees up $100 in funds you can use to more aggressively bid
on the other pallets. Or can use to lowball bid on other pallets
you normally would ignore. Just make sure your expanded bidding
does not trip any shipping or loading issues.

Much more here.

I am mightily displeased with the Salvex auction site
whose policies seem to overwhelmingly favor sellers.
Seems I "won" an auction with a fair and reasonable
bid, only to have the seller try to slap a totally unconscionable
and utterly ridiculous $15,000 reserve on me after the sale
was complete.

Salvex does clearly state that the seller can accept or reject
any or all offers. Or can add or change a reserve at any time.

All bids are usually kept invisible and your bids are absolute
rather than proxy. But bids that are under any stated reserve
are accepted anyway and get ranked.

Here's my current thinking on a useful bidder strategy on this
site: First and foremost, spend little time with them and do
not expect to win.  Make sure you know where the items are
located and verify a local shipping service ahead of time.

And, of course, never bid foreign.

Should you still find something of interest, wait exactly
until thirteen minutes before auction closing, and then bid
60 percent of your maximum. If you become the high
bidder, sit back and watch for any vile and despicable
10 minute autoextensions. If you do not become the high
bidder, rebid immediately at 80 percent and then at 100 percent.

If someone else still outbids you, extend the auction
yourself in the last minute, first with an 80 percent maximum
bid, (if needed) and then with your 100 percent maximum.

But these turkeys are best avoided entirely.

More on useful auction strategy here. Your own custom auction
finder can be created for you per these details.

This site seems to offer custom rechargable battery
package solutions for reasonable prices that typically
range from $25 to $50.

While intended for robotics and R/C model enthuasiasts,
they can nicely solve repairing the unrepairable in the
way of old or obsolete electronic test equipment.

In general, their packs will usually extend the initial
device life, owing to improved modern cells.

Secret "hobo marks" to the rescue.

Industrial auctions may not have lot numbers on a
confusing number of pallets. Or they may be a choice
auction where you pick what you want. Keeping track
of what is what in the middle of hectic bidding can be
a problem.

Assuming the packaging on the pallet is of little value,
it may pay to make "secret marks" on them to seperate the
gotta-haves from the sort-of-oks from the poisioned-lots.

Possibilities might be to use a square for something you want to
bid on, a triangle for something you want to pass unless you can
steal the item, or a circle for something to avoid.

Alternates might be to use small color coded stickers or
to press color coded push pins into the pallet bases.

Naturally, you should not get caught doing any of this
and you should not degrade the value of the lot in any manner.
Chances are it is good to use two marks just in case
somebody else stirs the pallet contents.

It is, of course, essential that the meaning of your
markings are not obvious to anyone else.

More auction tips here.

Let me be sure I've got this straight: The primary
function of the US post office today is to delay the
delivery of information or content. And their solution
to their present financial problems is to further delay
the delivery of information or content .

While making the process more inconvenient.

My solution instead is wildly counterintuitive: Immediately
repeal the private express statues. And the problems should
automatically take care of themselves.

In hindsight, the random convenience operator in our
Gonzo Utilities seems rather clunky. In that it involves
two divisions, a multiplication, and conversions from
integer to floating point back to integer.

Which is not quite as bad as it sounds, because the
underlying PostScipt rand operatior is rather slow. In
its present form, random gives something less than a
2:1 slowdown.

Here is another approach that runs at nearly the full rand
speed. It requires a precalculation and thus is usable mostly
when you want to generate identical random integers many
times in a row. Such as mottling a bitmap background...

/randint 7 store
/randdiv 2 31 exp 1 sub randint div store
/newrandom {rand randdiv round cvi} store

Three highly useful sources of industrial pricing info include
McMaster Carr, PLC Center, and OEM's Trade.

Additional pricing research tips here.

Many integrated circuit manufacturers have or shortly will
cancel their ceramic packaged products intended for Military
883 and similar protocols.

These packages typically sold for ten times or more the
price of plastic or industrial packages. Their prices are
now certain to skyrocket. When and if they can be
found at all.

I've just posted a bunch of these to our eBay auction and
store sites. Mostly the more popular items and usually
priced much less than their commercial package pricing.

All guaranteed functional and available for immediate delivery.

Except for the Cottage Bakery, the local bakeshop scene
is, to say the least, rather grim. But two outstanding locations
elsewhere are more than worth  the effort to travel the long
distances to them.

Especially as part of a daytrip.

Check out the Roundup Bakery ( formerly the Bakery
House ) in Elfrida for just about anything. Or the
Breadbasket in Sierra Vista for superb and unique
pretzel rolls. Which usually sell out by 8 am.

I consider automatic online auction extensions to be both
vile and despicable. They grossly undervalue the bidder's
time, reward the terminally stupid, force you to reveal your
presence, and make realistic bidding horribly inconvenient.

Public Surplus is a typical example.

My strategy in dealing with extendable auctions goes something
like this: First you avoid them because they are a gross insult
to the value of your time. Secondly, if you must bid, select a
price that is the absolute maximum you are willing to pay. I've
found that a price "a step above a step above" a currency
resistance threshold works well for me.

Such as $126.53.

Finally, you enter your price once and do so one minute before
the trip point of any extensions. Then you sit back and watch,
responding in no manner to the other bidders.

The best possible type of auction for a buyer is called a Vickrey
Auction. And, indirectly, also is the best possible type of
auction for the sellers. eBay could become much more of
a Vickrey auction by offering a twenty five cent autosnipe
service. After the auction is complete, your invisible proxy
is entered and you then win or lose.

I've just done yet another update and expansion of our
Gila Valley Day Hikes. The goal is to end up with
365 positive entries, and we seem to be at 337.

Which really represent over 500 unusual or interesting
things to do. Because of multiple topics in some entries.

I seriously doubt there is anywhere else in the world where
there is so much to do so close in in so many ways. All
centered on the greater Bonita-Eden-Sanchez metropolitan
area.

Did you know about the Renarium? Or where the
Shingle Mill really was? Or that the largest local
catfish weighed in at a "fair to middlin" 71 pounds?
Have you seen the UFO fish filet at the Taylor Terrace?

Finding valuable new stuff is getting tricky. Please
email me over anything I may have missed.

Additional similar topics here.

Win a few, loose a few. The apparent reason the Shingle Mill
Canal lo11inor Webster Ditch System. Many thanks to Cluff Ponds
historian George Hayes for clarifying this for me.

It still remains a fascinating entry for our Gila Valley Day Hikes.

A second anglo project called the Robinson Ditch very clearly had
solid prehistoric origins. Judging by its location, its "hanging canal"
orientation, and its stunning similarities to adjacent known prehistoric
water projects. Clearly, they "stole the plans".

Does the Minor Webster system also have prehistoric origins?
It certainly is in an expected place and not having a canal
here would leave a glaring and obvious gap if "their" goal was
to exploit every drop of Mt. Graham water.

There are a few potsherds there and there and an apparent
mono fragment or two present, but these can be reasonably
expected most anywhere in the Gila Valley.

Proving this one way or another might be tricky. Due to the
extensive rework and heavy brush.

I've got a whole pile of my classic books and articles that
I soon want to add to our free ebook archive and free classic
reprints archive.

One of the problems I've been having is finding updatable copies of
my Modern Electronics stories. Especially those involving Marcia
Swampfelder.

Please email me if you can get me copies of any of these or have
the original magazines for sale.

I've still got a pair of genuine Eastman 1908 commercial 35 mm
movie projectors for sale that I would like to clear out.

Their provenance was the original Cliffton, AZ movie theater. They
are presently disassembled and can be shippid in multiple UPS
boxes. They are eminently restorable and fairly complete.

Please email me if you have any interest in these extremely rare
collectibles.

For years, I've been creating what, for a better name, we might
call Lancasterisms. These are intentional but apparent
topographical errors intended to reveal a higher or greater truth.

Such as a groundswill of popular demand. Or what those French
Veternarians call a "four paw". Or being overly enameled on some
idea. Or ending up a few bricks shy of a full deck. Frosting the lily or
guilding the cake. Or not being able to hit the barn side of a broad.
Or the mythinterpretiation of something.

Or sources close to an associate of the barber of a usually reliable
spokesperson. New uses for Chebycheff Polynomials would take
the Cheby to the Leby. Many of the web perpetual motion schemes
and those electrolysis fantasies seem to involve electrocity.

All in one swell foop. Provided there's no oint in the flyment
An unauthorized autobiography. A jerk of all trades. The
local hysterical society.  

Letting the cows come home to roost. So long as they are
elected by acrimination. That little dip between the winter slump
and the spring slack period. Sort of the qualm before the scorn.
Geranium transistors.

Plays a mean eclectic guitar. Pioneers new methods of animal
husbandry. Speaks Esperanto like a native. Bruno's attitude
relateralization facillitation. Or the long lost oriental martial art
of Tai Wun Oun. Will be persecuted to the fullest extent of the law.
Thus reaching a new millstone.

Geologists, of course, classify rocks as sedentary, ingeneous, or
metaphoric. And New Mexico hikers might call an emergency
rain shelter a Poncho Villa. The illegal aliens in the Alabama
Grits Harvest, will, of course, be used for flavor only.

These are somehow related to the Yogi Berra's of others, such as
"Nobody goes there because it is too crowded", "Deja Vu all over
again", or "Let's keep the Status Quo right where it is. Or "When
you come to a fork in the road, take it".

Or Ed Abbey's classic "Androgynous Ammonia". Which might
even involve an engendered species.

I have a hollow feeling I've lost some of the better ones of these
somewhere along the way. As you go through some of my older
books and stories, please report any that may be missing in
action.

Because Opporknockity tunes but once.

GPS Receivers have gotten amazingly compact and low in cost.
Such as this latest example.

A good GPS free trade journal can be found here.

Once again, I've updated and expanded our Arizona Auction
Resources. An older survey of what is where appears here,
and bunches of additional support here. And a directory here.

Your own custom regional auction finder service can be
created for you per these details. These can be prime sources
to base your eBay business on. More here.

I was very surprised to see that barometric pressure for
many specific regions has strong diurnal components whose
slope and shape tend to repeat from day to day, even
over months or years.

In the basin and range southwest are earth tides and
canyon winds. The classic wind patterns are upcanyon
during the day and downcanyon at night.

This suggests measuring the average slopes and then
making your altitude measurements at appropriate
times. Locally, this seems to translate to a zero slope
between 8 and 9 am.

This site can give your local variations, their slope, and
their history. Just change your local airport call letters
These will be the luggage tag with a K in front.

( Continuing yesterday's entry on how good a barometric
altimeter can become... )

If we take the canals at a nominal 3500 feet, we see from
this reference that a tenth of an inch of mercury barometric
variation translates to a hundred feet of elevation.

The nearest barometric history reference I could easily
find was from Safford. Where we find that rapid variations
seem rather less common here than this Phoenix barometric
history.

Phoenix fairly routinely will show a tenth of an inch per
hour variation, although the norm is much less. Safford
only rarely has a slope steeper than a tenth of an inch
in two hours.

Thus, our normal expected local barometric altitude drift
would be at worst something like fifty feet per hour.

However, some correction techniques are possible.
The simplest is to close the loop and return to the
initial point, averaging the results. Next simplest is
to repeat the measurement closed loop a dozen times
and average the results.

Making the measurements on a dead run to minimize
any time delays would also surely help.

Curoiusly, the Safford barometric pressure seems to
be flattest betweeen 8 and 10 am. This appears to be
reasonably consistent from day to day.

Going true differential would likely be needed for more
accuracy. In which one eTrex-30 would sit on the initial
point and a second would be carried between the
measurement points. And continual correction being
made by the measured drift.

All of which predicts that special effort will likely be needed
to get anything remotely near a one foot resolution .

The obvious next step is to receive one or more
barometric altimeters and record their drift rate and
their differential drift rate. And their internal noise.

The radio control model folks surely must be working
on this as well.

We definitely need some accurate differential altimetry to
resolve several hanging canal problems. GPS vertical resolution
is pretty much worthless as it can be off by 400 feet or more. A
one foot differntial resolution sure would be nice.

Some painful older solutions would be to use a Brunton Compass,
an automatic optical level, a laser level, or full survey gear.
But these can prove expensive, bulky, inconvenient, or painful.

I just ordered a Garmin eTrex-30 which has a barometric based
altimeter. The obvious question is how accurate can barometric
altimetry be?

New mems and piezo low cost sensors are now available, and it
is reasonable that Garmin would use something similar.
Some useful tutorials and conversion sites appear here,
here, and here. The actual sensors can be found here.
These devices can in theory offer a nine inch resolution.

Factors which affect resolution are the micro and macro weather,
the noise of the sensor, the number of measurements, and
whether the loop is closed on itself to allow averaging.

Barometric pressure sometimes can vary half an inch of
mercury in a day, which translates to an absolute 500 feet
at an elevation of 3500. Which might be 50 feet per hour.
And might be significantly reduced by closing the loop and
averaging. Or with repeated faster measurements.

But average barometric variations should be much, much lower
than this. Especially on the upper Sonoran Desert on a cloud
free and wind free day.

More on this when and if the eTrex-30 shows up.

Once again expanded our Gila Valley Day Hikes library.

Revised and improved our Bitmap Typewriter with
the tenative new code here, the older tutorial here
and general details on using PostScript routines here .

The bitmap typewriter lets you add or improve small
typography lettering on a bitmap with the highest possible
resolution and quality allowable without going subpixel.
It is especially well suited for eBay photography.

In general, the new routines allow a larger transfer map with
greatly simplified entry, provides for new minor color
adjustments, and adds a 216 color capability. The latter
is somewhat coarse to be always useful, though.

The new space is 1000x1000 split into four regions.
Here is the new simplified command set....

ba - set backgroumd A (bottom quarter)
bb - set background B (next highest)
bc - set background C (next highest)
bd - set background D (highest quarter)

gk - set global kern

so - set color old RGB way (as in 127 26 47 so )
sc - set color new 216 way (as in 32 sc)
sf - set PostScript font family
ss - set bm size
st - setgraystring
sy - set y increment

tr - set color trim change

b+ - more blue
b- - less blue
g+ - more green
g- - less green
r+ - more red
r- - less red
w+ - more white
w- - less white

Please report any bugs or suggestions for further improvement.

Some epsilon minuses that were pissing around with Brown's
Gas just managed to blow Southern California off the map.

Hydrogen is exceptionally dangerous to play around with.
It gets ridiculously more so when you store it premixed with
stochiometric volumes of oxygen.

An analysis of the stupiditiies of both deadly and totally
worthless Brown's Gas appears here. And the utter ludicrosity
of the "hydrogen economy" here

Thanks to Keeylynet for this info.

When selling integrated circuits on eBay, it is super important
to have photos with identifiable part numbers. Subtle postfix
differences can make a radical difference in package, speed,
or temperature range. 

Such as this, this, this, or this image. The dilemma is to get
useful and decent pictures with a minimum of time and effort,
as ic's have to have a low unit price and you may not have
bunches of them to divide front end tasks.

Scanning at ultra high resolution is a must. I've found 575 DPI
to be a useful compromise. Since the bitmap will be on the
huge side of enormous, you'll want to crop and move into
Imageviewer 32 as quickly as you can. The usual first task
here after cropping is a slight rotation to get the chip
perfectly horizontal.

It is always important to scan at the same high resolution
so that any future cuts and pastes will end up the correct
size without rework.

At this point, you have a choice of reusing an existing
background and package and pasting the new lettering
over it. Or of starting over from scratch. To start from
scratch, pick a mottled background such as one of these.
The reason for a slight mottling is to eliminate JPEG
edge artifacts.

Pick the nicest looking edge pin and stretch it into a
credible shape. Then bump the new background up to
it. Then step and repeat the pins to cover one whole
edge. Always try to keep your horizontal and
vertical lines oriented to one pixel accuracy.

Then work out a credible corner and next do the
vertical edge on the right end that does not have the
notch in it. Then cut, flip, and paste to take care of
the other horizontal pin edge. Next, do the left
vertical edge, taking care of the notch.

Note that you can temporarily put a vertical guideline
well out in the background to get your edges perfectly
aligned.

The remaining inside areas are redone with credible
patterns, again preferably with a slight mottling.

Ideally, you should not mess with the lettering beyond
raising its contrast and sharpness. Any rework that
introduces typos is clearly a no-no. Sometimes simple
patching of minor lettering dropouts will work, as will
carefully outlining each letter with the suitable background.

If you must do stem rework, start with an H and an O and cut,
clip, paste, and rotate as needed to generate the rest of the
letters. Generally, you'll want to work at least double size
and preferably larger. Your final reduction should cover a
multitude of sins.

If all else fails, you can use our revised Bitmap Typewriter. But this
takes time and could easily lead to typos.

We are in the process of adding and expanding many
of the items in our eBay auctions and store.

First and foremost, are more than a dozen variations on
genuine ITT Cannon round CA mil-style threaded connectors in
many of the most popular sizes and pinouts. Still sealed in
original factory baggies.

These are at a tiny fraction of distributor cost and are
available for immediate shipment. A request: Please
be sure you understand the part numbering! There can
be subtle differences in shell size, number of pins,
plug versus socket, rotation, mounting style, and such.

Second are great heaping bunches of classic integrated
circuits, many of them mil spec or otherwise premium. And
many of them impossibly difficult to get with prompt shipping.

And fourth are four flavors of classic all-metal XLR style Cannon
and Amphenol 4-pin connectors. Male and female, Square,
rectangular, and round.

Here's a summary of our ongoing Safford Basin prehistoric canal
research to date. A total of 12 canals are reasonably known with
an estimated total distance of 51 miles! Projections suggest at
least 16 hanging canals with a total distance of 60 miles.

From east to west...

0. - P Ranch - remains speculative with only a short
V section of canal, favorable terrain, and persistent
rumors. Not included in totals.

1 - Ledford - total length of 3 miles. Difficult access.
Initial hanging portion from Smith Dam. Needs
further exploration. Studied and reported on by Neely.

2 Henry's - Recently visited portions end in modern
stock tank. Connection to Marijilda unproven. The
known length is 2 miles, likely length is 4 miles.

3. Marijilda - Well researched with spectacular hanging
portion. Historic rework flows to this day. 5 miles long.
Includes "aquaduct" portion a meter high by a hundred long.
Studied and reported on by Neely.

4. Twin Boobs - Many mysteries remain and possible
Marijilda branch origin remains speculative and unproven.
Quite obvious on Aerial photos. Significant hanging portions.
Projected possible length is 5 miles. Includes apparent field area.

5 Deadman - Initial hanging portion speculatively covered
by modern pipeline. Possible 3-way switch mid channel.
At least 5 miles with possible extensions into Longview area.
Original canal flows to this day.

5A Longview - Hints of canal structures in the Freeman Flat
area below Longview Ruin. Which could have been Deadman
fed via switching. Not included in totals.

6 Robinson - Reworked to feed cattle tanks by anglo
pioneers. Excellent topo traces, well hung. Destination
not currently known, but likely Robinson Flat.. At least
3 miles long.

7 Allen - Sources from Spring canyon, runs under Allen
resevoir to an unknown destination. Several key but
small segments unproven and unexplored. Exceptionally
large "Culebra" cut two meters deep and a hundred long.
At least 6 miles long.

8 Mud Springs - Sources in upper Ash Creek and impressively
climbs "up" the Mud Springs bajada. Some hanging
portions and a companion structure. Several pieces missing.
Believed to source the Jernigan Canal. Final destination
unknown. At least 7 miles long. Ten inch diameter mesquite
tree midstream and overrun by SCS dam strongly supports
prehistoric origins.

9 Jernigan - Apparent branch of Mud Springs seems to terminate
in the Central Dump area. Has a triple "U" turn to maintain
slope. Small but crucial portions missing and unexplored.
Possibly 3 miles long. Possible branch with short aquaduct.

10 Shingle Mill - Newest "rediscovery" appears to source
in Shingle Mill canyon and deliver to the Cluff Ponds area.
While only a few hundred feet have been visited to date,
projected length is 3 miles. Still under active exploration.

11. Lefthand - Studied and reported on by Neely. Includes
plant nurseries and other intensive ag. Projected length is
2 miles.

12. Carter/Nuttall - Possible continiuum with other
hanging canals remains unproven and unexplored.
Includes Taylor Terraces which likely are CCC
but may have prehistoric origins.

Some free planetarium software for PC, Mac, or Linux
can be gotten from the folks at Stellarium.

We still get asked all the time why we have no foreign sales.
The short answer is that they are not remotely worth the
grief and hassles that inevitably result.

I guess I am still pissed over the Canadian post office refusing
to accept 45,00 pieces of my first class mail several decades
back. Separately, several of our suppliers contractally demand
that we permit no foreign sales.

The bottom line is this: If you are a success with your domestic
sales, you have no need for foreign ones. And if you are not a
success domestically, foreign sales only compound your problems.

More on eBay and other small scale high tech marketing here.

Updated and expanded our Gila Valley Day Hikes library.

Also added a link to mine and mineral data.

One of the foremost insider "secret" rules for eBay success
is simply no pissing around.

Paypal only. Absolutely no checks or money orders. No self
hosted Visa or MC unless you have tremendous volume and
ridiculously low rates. Absolutely no foreign sales.

Absolute minimum average order of $19.73. Minimum line
item of $9.73, hoping for multiple sales per customer. No
free samples, no exceptions on minimum quantity per order.

Discounts only on larger volumes, perhaps 5 percent at
$79, 10 percent at $199.

Nothing offered for sale that you cannot hold extended at
arm's length. Nothing for sale in obvious known problem
categories.

Minimum of 30:1 SBR. Absolutely no consignment sales, anytime,
ever. Always seek a 21 day payback and a 15 month hang time.

Above all, should a customer be unhappy for any reason,
refund them. Usually without a return.Then, of course, block them.

Your 30:1 SBR protects you from any occasional misfires.

Photography, scanning, and listing as efficient and hassle free as
possible. Shipping at cost. Combined where and when possible.

Much more here. Consulting services and lectures available.

One sure way to tell a microcomputer pioneer is by all
of the arrows in their back.

At first glance, these Taylor Terraces seem like a UFO
fish filet recipe. But a more credible explanation is that they
were likely a CCC project.

Onsite inspection shows many lichen covered rocks where the
lichen edge is non-horizontal and not in ground contact. Which
strongly suggests recent construction or disturbance.

There are also similarities to other CCC worthless boondoggle
projects, but none I know of that are quite this monumental.

Similarities to recent stimulus spending are stunning.

Google Maps just upgraded the newness and the resolution of
much of Southeastern Arizona. Acme Mapper also uses and offers
this service, among many others.

Here's a split between the old and new resolutions. Note that you have
one more click of resolution, but the old stuff goes green on you.

You can easily recognize telephone poles and wires. But it is still
hard to tell if the hawk on the crossarm is missing any tail feathers.

New coverage seems to range from Artesia to Portal and from
the new mexico line to the San Pedro River.

Garmin just announced an eTrex 30 handheld GPS unit
that includes a barometric altitude sensor. Presumably
something similar to to the Bossch BMP085 or its Sparkfun
SEN09694 breakout.

I have yet to find its specified resolution or when it will be
available. But it clearly should be ridiculously better than
GPS elevation readings. Whether it will be useful for finding
missing portions of our prehistoric canals remains to
be seen.

The theory goes something like this: In absence of
sheet flooding or other terrain mods, a canal can
reasonably be expected to continue at its previous
slope. The area in question can be no higher than the
entry point nor any lower than the exit point. And the
midpoint elevation should be somewhere near half the
difference between the entry and exit heights.

When these constraints are placed on the current
topography, candidate areas should often be
dramatically limited.

Much more here.

What improvements could be made in eBay? I'd start
with these...

First and foremost, eliminate any and ALL restrictions on URL listings
in offers or direct off-eBay contacts between buyers and sellers.
Ebay's paranoid obsession with losing an imagined offsite sale
is completely overwhelmed by the damage and inconvenience this
policy is causing both buyers and sellers alike.

Second, allow sellers to recover ancient listing history, Even if a modest
fee is necessary. It is common for a seller to realize they forgot to relist
something many weeks or months after the fact.

Third, institute a twenty five cent autosnipe service. In which bids would
be invisibly placed just after the normal closing. This would convert eBay
into a true Vickrey Auction with tremendous benefits to sellers and buyers
alike. Sellers should have the option of enabling or disabling this service.

Much more on eBay strategy and tactics here.

It is fast becoming apparent that a prehistoric mountain stream fed
canal system existed in the Safford valley that seemed to attempt
to exploit every drop of Mount Graham water. At least twelve
canals are known with a total length in excess of forty miles!

A distinguishing features of these world class engineered canals is
that they are often "hung" on the edges of steep sided mesas.
Giving precise control of slope independent of terrain!

A reasonable question might be why these canals and their significance
to Southwestern Archaeology have not been previsiously discovered.

First and foremost, despite a few superb papers such as this one and
this one, the Safford basin has largely been a backwater of Southwestern
Archaeology. Owing to the powers that be lying elsewhere. While the
local community college has an entry level archaeology program, they
are not in any manner a major recognized research institution.

Secondly, there seems to have been a local cultural bias against the
abilities of the stone age inhabitents. Suggestions that anglo pioneers
"stole the plans"or "dug out an old ditch" are often met with anathema.

Thirdly, most of the canal routes lie on state lands of difficult access
and sharply limited current interest or development. Physical
evidence of many canal reaches are subtle or missing entirely. Yet
others flow to this day.

Fourth, the very concept of the magnitude and scope of this canal
system demands a population and a political system vastly in excess of
what was previously believed to exist.

Fifth was the difficulty in recognizing continuums. A hanging canal on
a 200 foot high mesa is obvious, while one on a 20 foot mesa might not be
Nor is a hanging canal that was long buried by an anglo pipeline.

Sixth was the dearth of recognizable artifacts beyond the often filled in
canal reaches themselves.

Emerging technology also became considerably important.. Centralmost of
which was Acme Mapper. This image of the Twin Boobs Canal played
a central role in greatly expanded current discoveries and exploration.
The Garmin eTrex handheld GPS receiver also continues to serve as
a key resource.

Much more here.

This site  is an interesting resource for units and conversion factors.

At first glance, our prehistoric mountain canals seem to be rather small,
but if you run them long enough, they move a surprising amount
of water.

Ferinstance one gallon per second equals an interior attack firehose
of 60 GPM or just under 100 acre feet per year or over 31 million
gallons per year. Enough for 40 acres of modern drip irrigated cotton.
And great heaping bunches more for teacup-at-a-time prehistoric watering
of desert acclimated crops .

And one cubic foot per second equals a fire cannon of  nearly 500 GPM,
over 700 Acre Feet per year, and nearly a quarter billion gallons
per year.

Otherwise proven by a leaky faucet racking up an incredible water bill.

Aparently certain military suppliers still stock the older HP
oval line cords. But you have to have a security clearance to
buy one!

As near as I can tell, these old line cords have become unobtainium.

If you integrate a pulse width modulation PWM waveform,
you can extract its analog equivalent. This has an enormous number
of uses, not the least of which is LED dimming.

I just ran into a pair of PWM variants that you may find of interest.

We might call the first "phase alternating PWM". In which you
output a PWM waveform and then its mirror. This cuts the number
of transitions in half and raises efficiency by reducing your switching
losses. But probably is not all that much different from reducing your
PWM frequency by two.

The second is called "binary code modulation". Picture a PWM waveform
with a 256 microsecond period. And a microcomputer or whatever that
has an 8-bit parallel input. Bit 0 makes 1 microsecond of the period
high. Bit 2 makes 2 microseconds high. Bit 3 makes 4 microseconds high.
Up to Bit 7, which makes 128 microseconds high.

This is an all digital approach that avoids any analog involvement. It is
less efficient that conventional PWM, though, in that up to eight transitions
are needed per cycle, compared to two in conventional PWM and one
in phase alternating PWM.

A tutorial on binary code modulation appears here.

Variations on PWM for digital sinewave generation that allow you to force
any number of low harmonics to zero can be found here. And a newly
revised calculatior here.

Further expanded our Gila Day Hikes library page.

Yup, White Streaks Canyon is still at the bottom of the
list. Definitely an acquired taste. Recommended for hikers
that bring their own Catclaw just in case there is not enough
along the trail.

A rock scramble best viewed from a distance. Right here is
more than close enough.

365 main entries might make a reasonable goal or
waypoint. But it is getting harder and harder to find
( and verify ) quality new stuff. Yet I feel some obvious
stuff is glaringly absent.

Please email me with your suggestions.

A reminder that consignment sales simply do not work
on eBay and that they are a certain formula for disaster.

Optimum results on eBay demand something like a 30:1
SBR or sell buy ratio. This would pretty much be the
equivalent of receiving a 97 percent comission.

A 50% commission would equal a pitiful 2:1 SBR, while
a 20% commission would be a laughingly absurd 1.25:1
SBR.

In addition, people's expectations of the value of something
sold on eBay are often wildly excessive. And they fail
to realize the an eBay sale is an enforcable contract under
the uniform commercial code.

Offering a commercial consignment service is even worse,
because you add employees, rentals, and other fixed brick
and mortar costs to the mix. Besides raising questions of
product legality and theft.

Much more on viable eBay routes here.

The next step in the Shingle Mill Canal research   
would be proving or disproving its prehistoric origins.
The canal is certainly in an expected place, an expected
size, seems to fill an expected need, and has the expected
engineering.

And "fairly large" mesquite trees mid channel certainly
suggest little recent use. But the adjacent "service road"
raises serious questions. As does its overwhelming
obviousness on Acme Mapper compared to the others.

The canal and road combination already occurs nearby
in Marijilda Canyon. Where we clearly have an anglo
rebuild and reuse of a known prehistoric ag feature.

But following a prehistoric canal would be a rather
obvious "steal the plans" route for a much more
modern road. Perhaps one that was Weech Toll Road,
McEniry Scam, Cluff Ponds, or CCC related. Or
even had something to do with the shingle mill itself.

Cluff Pond access from the west was cut off in the 1980's
by locked gates. The amount of "overgrownedness" on
this "lost" road is consistent with this time frame. But
the two could certainly be unrelated.

Logical next steps are to visit the portions of the canal
that do not have an adjacent road to see what differences
if any exist. And to check any road crossings ( at least one
is suggested by Acme Mapper ) to see if they run roughshod
over the canal without accomodation. As is the case in
many nearby examples. 

A closer look at the actual road construction might prove
useful. At first glance, it does not seem graded, nor does
it have any evidence of heavy equipment cuts and fills.
But there are nicely arranged edge rocking in places.

Obvious direct CCC involvement seems absent, although
there clearly are other minor CCC works in the area.

Totally separate would be a historical approch with
appropriate interviews and family journals common
to the area.

Your participation is welcome,

Managed to actually visit and verify the Shingle Mill
Canal. It appears to be an anglo rework of a prehistoric
water project similar to the dozen others that seemed to
use world class engineering to exploit nearly every drop
of available Mount Graham water with as much as FIFTY
MILES of water channels!

Using a mix of stone age technology and mind-boggling
engineering. .

As usual, the find raises many more questions than it resolves.
The site appears to clearly be pre Cluff Ranch and not at
all in the CCC style. Five and six inch diameter mesquite
trees are common in the middle of the water channel.

While clearly sited exactly where a needed prehistoric
canal would be expected, there seems to be an adjacent
roadway that raises issues. Just to complicate matters
further, this "lost" or "forgotten" roadway could in some
manner be associated with the Weech Toll Road or the
McEniry scam.

Aerial photos also show some rather ambiguous areas
further south that could be grids or  sites. These remain unvisited.

Only about a mile has been verified. The source is believed
to be Shingle Mill Creek, and delivery is believed to be
in the area of one of the smaller and now drained Cluff Ponds.

Your assistance in both field work and funding would be
most certainly welcome.

A credible source for global warming appears to have
been located  Per this analysis.

Updated and expanded our Gila Valley Day Hikes
library pages.

A new variation on GPS appears to be tested and ready
for commercial rollout. It seems to offer dramatically improved
sub-inch resolution, especially vertically.

What you do is build a few fake GPS satellites at WiFi
frequencies and spread them at the edges of your farm
field or survey area or wherever. Much higher resolutions
can cheaply result. The fake satellites can track each other's
stability for much cheaper time accuracy.

Most newer scanners can easily be used as a superb
magnifying glass or microscope.

This can be very useful for identifying hard to read
integrated circuits and SMT devices. The trick is
to first raise the resolution for an initial 8X magnification
perhaps by using something near 575 DPI. You can pick up
a further 8X in paint with its pixel magnification for a
non-half-bad 64X total,

Further dinking with the gamma, brightness, and contrast
can yield additional enhancements. And the accuracy
can be much better than writing down what you saw.

It can be real easy to confuse the watermarked copy
protecton on the Google Maps used by Acme Mapper with true
ground features. In general, these appear as subtle lettering
that runs diagonally up the image. Perhaps repeating eight times
per screen.

The resolution of local features ( typically 100 feet per inch ) sure is
frustrating when tryng to interpret archaeological detail. Apparently
doing anything significantly better has to be flown rather than
satellite. And thus outrageously expensive for remote areas.

... And a reminder that our mid-level navigation has
been improved as an ongoing project. With full
details here.

Several additional links remain to be added.

Those old Hewlett Packard 3 pin "oval" line cords seem to
have just gotten a lot more difficult to find. Seems that the
Belden/Volex 17280 has been discontinued. And that recent
surplus offers have been gobbled up.

The few that are available show up on eBay for $20 or more
each . Rumor has it that some older sewing machine
repair stores have some pricey ones that also may be called
"recorder cords".

The Molex .093 crimp terminals are somewhat small, but possibly could
be spread. A replacement could be made from a standard line cord and
three of these. Followed up by sealing them in polyethelene from a hot
glue gun.  But the terminals are rather fragile.

Someone else suggested crimping yellow butt splices to a cord.
These apparently also can be made to fit with a slight crimping and
also would be hot glueable. But neither of these methods seem very
reliable or promising.

There are two flavors of the cords with opposite sequences of the
outside pins. Using the wrong one could raise some safety
issues during servicing ( the switch would be on the low side ),
but otherwise would not cause unreasonable problems to
most people most of the time.

Probably the best solution is to replace the connector with
a standard IEC jack. Such as the sixty two cent Mouser
562-701W-X2/02. But these may have nibbling and rear
clearance issues. Not to mention panel space availability
for their much larger size .

Yet another possible route would be to use a fixed line
cord and a Heyco strain releif.

Please email me with any suggested solutions.

What was an excellent website for the TEP Springerville
solar pv array seems to have gone dark with no obvious
upgrade or update.

This despite a recent modest expansion. A curious fact: The
generated power was never used at all for consumer
distribution. Instead, its entire output went to partially
powering the pumps on their adjacent coal fired plant.

Two additional sources of local VHF scanner frequencies
can be found here and here.

Per this price tracking directory, the prices of pv solar
panels have been dropping significantly. Partially due
to overcapacity caused by the word getting out that
these are still gasoline destroying net energy sinks,
and partially due to some obscene European subsidies
mercifully being discontinued.

Present pricing averages $3 per peak watt with some
prices around half that. The value needed for genuine
no-bullshit net energy production is a quarter per peak watt,
so present prices still miss by a factor or six to twelve.

Even after the quarter per peak watt is reached, it will take
at least another decade to pay for all of the previous
subsidies and industry wide energy consumption. Plus the
amortization of obvious heavy new investment Or possibly
a lot longer per this analysis.

An interestng free solar industry trade journal can be
found here. With more on energy fundamentals
here. And more on pv in particular here.

Found yet another bug in the Black and Decker power
monitor. Their outside temperature sensor is encased
in black plastic. During times of direct sunlight, the
reported temperature is ridiculously high.

Expanded and updated our Arizona Auction Resources
link farm.

Your own custom regional auction finder can be created
for you per these details.

Additional auction support can be found here.

A map of when Netflix comm is up or down can be found
here. And a list of their "watch now" movies here.
Or here. And TV shows here.

At present, a reaonable guess of their watch now DVD's is
about 4000, compared to many more than 40,000
mailable DVD's.

You can create your own watch now of stuff they
don't have by using these links. Especially Fox, CBS,
and ABC.

Your network ready "smart tv" will have to allow
unrestricted internet interface to access these. Otherwise,
input via VGA ( from your own computer ) is sometimes
possible. Note that HDTV resolution is normally much
less than the best computer monitor resolution. But the
results will be the same for any HDTV sourced material.

I am eagerly awaiting CSI Gila Bend. Plus the cross-genre
classic Godzilla versus the Night Nurses. But the restraining
order from the Tapioca Pudding Institute caused the latter to
only be released directly to 8-track.

A reminder that our Magic Sinewave Calculator has
been updated and improved.

Quantization effects are now much more accurate and
are more conveniently exported.

Much more on Magic Sinewaves here.

There seems to be a fatal flaw in the Black and Decker
Power Monitor. If the conventional meter wheel does
not make it around once during the calculation interval,
the reading rounds down to zero.

This gives a tremendous credibility gap when a zero
is displayed during obvious power consumption time.
It also lowers the monthly totals dramatically. It also
gives you the least information at the time it would
be the most crucially useful.

Accurate measurement info should be provided at
least down to fifty watts, and prefereably less.

The solution has been known for years: To measure
a low frequency event, calculate its time period
instead. Run the calculation separately and
continuously so that the rest of the algorithm always
has something useful to work with.

Such a solution would also allow "Your sensor is not
properly set" error messages as well.

Here's some of the local Gila Valley VHF comm frequencies.
Many of these are now plain talk and newly narrow band...

122.8000      SAD Safford Airport
146.8600      Heliograph Ham Repeater
146.8800      Ham Radio Skywarn Net
146.9000      Heliograph Ham Repeater
151.3175      RAC East ( Transmit )
151.4975      RAC West ( Transmit )
151.4000      AZ State Forestry Helio ( Receive )
153.0200      Pima Schools
153.6800      Graham County Electric
153.7400      Safford Utilities
153.7850      Graham County H2 ( Transmit )
153.9500      Fort Thomas Fire Department
153.9950      Graham County SO1 ( Transmit )
154.0400      Safford Municipal
154.0850      Thatcher Fire Department
154.1150      GC Search & Rescue ( Transmit )
154.2125      RAC East  ( Receive )
154.2800      Fire Mutual Aide
154.3250      Safford Fire Department ( Receive )
154.4250      U/A Observatories
154.7250      EAC Secuirty
154.8750      Safford Police Department
154.9100      Thatcher Public Works
155.0400      Bylas Police Department
155.0550      Graham County H2 ( Receive )
155.1450      Pima Municipal
155.1750      Southwest Ambulance
155.1900      AIRS5 ( Transmit )
154.4525      VTAC12
155.4750      AIRS5 ( Receive )
155.7150      Graham County SO1 ( Receive )
155.7600      Pima Fire Department ( Receive )
155.8350      Gram County SO3
156.8175      RAC West ( Receive )
158.8050      GC Search & Rescue ( Receive )
159.0825      Pima Fire Department ( Transmit )
159.1050      Graham County H4 ( Receive )
159.4050      AZ State Forestry Hekio ( Transmit )
162.1625      BLM Law Enforcement
168.1500      USFS Fire Helio  ( Receive )
168.5375      BLM ( Receive )
169.6000      Coronodo NF Fire Net
172.2750      USFS Fire Helio ( Transmit )
172.4250      San Carlos & Phx Fire Dispatch
173.8250      BLM ( Transmit )

Traffic on most of these channels is quite infrequent. Reception
is usually limited to line-of-sight from the Gila Valley. Some
additional frequencies may appear here.

Arizona law allows you to listen but in no manner interfere.

One of the unintended consequences of continual video
on demand becoming and being the entertainment norm:

What possible point could there still be in having "seasons"
to television series offerings?

There would be no competing for time slots because
there would not be any time slots for most viewers.

Managed to pick up a Black and Decker Power Monitor
and am rather disappointed with it. I'm not the least
surprised their EM100B is selling at deep discounts.

The alignment of the meter sensor ( It looks for the
black bar coming around on a stock older power
meter ) seems to be extremely critical. And the
display seems to consistently read low, especially
during low power usage times.

I suspect there is a fundamental algorithm problem
involving very low rotation rates.

And I feel the programming and interface is not all
that well thought out.

I've yet to take it apart to see if Flash Memory
storage can be added. Or if a full web interface is
even remotely possible.

Nonetheless, at $25 it seems eminently hackable.

Just added four precision slides to our eBay pages,
three of them possibly super premium SKF/ Gillman
brand.

One has a lockable slider, one has a dial style manual
advance. One is rotary with a half degree calibration,
and the final one has an unusual pitch and yaw
capability.

Added a new fast navigation feature to our Santa
Claus Machine llibrary page.

I keep lowball bidding on these, and hope to
eventually end up with one. As with real print,
it looks like inkjets will literally run away with
all the marbles.

Looks like its ecopornography time again.  A  
major electric car company has put a gasoline
destroying net energy sink pv solar panel on their
factory roof so they can claim their cars are
solar powered.

There are several fundamental reasons why solar
powered cars are both ludicrously absurd and
clearly ain't gonna happen.

The first, of course, is that not one net watthour of
pv solar electricity has ever been produced. And will
not be produced till a decade after the fully burdened
panel costs drop under a quarter per peak watt.

Actually, it may be a lot more than a decade. At least
per this analysis

We, of course, are still missing by a factor of twelve
or more, not counting the hideous subsidies that skyrocket
the additional true cost. Per details here and here.

The second is that on-board solar energy recovery
makes no sense whatsoever. At best you could come
up with two meters of effective area. Which even in
the Arizona sun would only be good for a kilowatt hour
of electricity per day. For a 30 kW ev, this translates
to a stunning TWO minutes of run time.

That is assuming your air conditioning energy costs do
not skyrocket, that the cv aerodynamics do not change,
and that the added panel cost is zero.

Thirdly, even if stationary and off vehicle solar energy is
a very poor time match. Instead of one set of batteries in
the car, you need one in the car, one being charged, and a
third set floating around charging stations. Tripling the
battery cost of an electric car is unlikely to aide in its
adoption.

But what about those student solar racers? Yes they
can be outstanding engineering project challenges. But
they have no relationship whatsoever to real world
transportation solutions.

Meanwhile, find that turkey that put the panel on the
roof and stake them to an anthill.

Not surprisingly, Borders has gone belly up. I still
maintain that most people will be astounded with the
upcoming sudddenness, speed, and totality of the
utter demise of books and traditional publications.

eBooks are already far outselling traditional print
titles. Despite eBook readers still having grevious
and obvious flaws.

A decent eBook reader would present a display
comparable to a full page printed color ad in Newsweek,
would cost less than $49 unless it was a free promotional
giveaway, would have only the mildest forms of DRM
as the rare exception rather than the rule over totally open
content, would allow totally portable and exportable
content over all platforms and all countries, and,
most definitely, would allow many different and totally
open formats that must include .PDF or its improved
offspring .

I can see several unavoidable consequences of where
the eBook revolution is heading. First, Book-on-Demand
publishing is pretty much stillborn, owing to eBook formats
having overwhelming cost and convenience advantages.

The earlier fatal flaw in BOD appears to have been the
total lack of cost effective trimming and binding solutions.

Secondly, libraries are clearly doomed and headed to
extinction. Because the very concept that a control
freak, rare material, and an end user need be in the
same place at the same limited time is now ludicrous.
At present, the best library in town is usually
Burger King's free WiFi link. Not to mention their
having far fewer food and drink restrictions.

Thirdly, scholarly publication is about to implode.
It is absolutely inexcusable that older scientific and
technical publications are not instantly and freely
available. Meanwhile, open sources such as
Wesrch should shortly utterly dominate.

Fourth, unexpected collateral consequences are likely
to hasten the eBook conversions. Such as a student
revolt against backpacks or wholesale bootleg publications
of outrageously overpriced info.

My own free eBooks and videos can be found here.

While I try to have a new and eclectic blog entry
every day, it is usually easier and more creative
to do several entries at once and "backfill" the list.

Please always look back a week or so to pick up
anything you may have missed.

In a monumentally stupid move, eBay is attempting to
ban most email links ( live or otherwise ) from their
seller listings.

The theory is that there will be fewer off-eBay sales.
The reality is that there will be MORE off-eBay sales
and that the amount of seller pissing off will vastly
exceed any possible imaginary benefit or savings.

eBay is entitled to any actual sales made through their
site. They are not in any manner entitled to payments for
any additional later sales conducted offsite between
buyer and seller newly communicating through email.

It does seem that the Business and Industry category
is exempt from the email link ban. One of the unintended
consequences, of course, will be that the listings in this
category will now go through the roof.

Posts to many Usenet newsgroups seem to be sharply
down, as does their quality. I suspect some of this is due to
the eyeball siphoning of newer social media such as Facebook
Twitter, and the rest of the gang.

The degradion of the sci.energy.hydrogen newsgroup was
pretty much expected, since word has finally gotten out that
the "hydrogen economy" is a ludicrous scam. Per these
details.

Less obvious is the drop in popularity and quality of the
alt.marketing.online.ebay newsgroup. Caused in part by
eBay itself no longer being a leading edge phenonomen.

Bruno does continue to be the AMOE attitude relateralization
facillitator. Since he is so big on multitasking, he frequently
combines his duties with those of being a product durability
tester for a major New Jersey Baseball Bat manufacturer.

And AMOE pioneered resolving the hassles at New Mexico
customs and language barriers through the use of truck
tires that can now be insided out to meet the different
spacing and size requirements without needing reloading
at the border crossings.

AMOE, of course, was instrumental in making sure that
the illegal aliens used in the Alabama grits harvest
were being used for flavor only. And that, contrary to
popular belief, Oliver Hazard Perry was not responsible
for sinking Lake Erie. It instead was Ednumd Fitzgerald.

I still find the sci.electronics.design newsgroup highly
useful, although even it sometimes degrades into
unhousebroken playground squabbles.

A reminder that we have a program going to make many
of my classic electronics books newly available as free
ebooks, along with constuction projects and videos.

A current list and links to what is already available can
be found here.

Present in-process projects include Micro Cookbook volume I,
Applewriter Cookbook, Cheap Video Cookbook, and
Enhancing your Apple II volume I. Along with bunches
of "golden era" Popular Electronics tutorials and
construction projects.

I'd dearly love to do full "Directors Cut" restorations
on all of these. Such as this example. Or this one
in process. But the required time and effort would
require your further support as a Synergetics
sponsor or Banner advertiser.

We've usually have several different flavors of solid
state relays on our eBay website. It is important to
understand what each of these does, because they
most definately are NOT interchangeable.

The most common version is called a logic level
driven, triac output ac only zero crossing switched
solid state relay. The input signal will typically
be 3 to 24 volts of isolated dc of a low enough
level to be microcomputer or logic level compatible.

The high current bipolar output will turn on and
STAY on, starting with the next ac zero crossing and
staying on for the remainder of the ac half cycle.

Other ac solid state relays will respond only to
ac inputs in the 90 to 240 volt range. And while
suitable for traditional industrial relay circuits,
are definitely not microcontroller or logic friendly.

A variation on the logic controlled input versions
can be turned on at any time in the ac cycle and
staying on for the remainder of the ac half cycle.
Zero crossing switching gives minimum RFI and
optimum line cleanliness. Random turnon instead
lets you do such things as lamp dimming or solder
iron heat controlling.

These do not normally work well on brush type
universal motors and not at all on induction
motors when speed control is intended. Instead,
special feedback circuits may be needed for
universal motors, and induction motor speed control
demands much fancier controls in which, besides
feedback, control of both amplitude and frequency is
needed.

Further, ac solid state relays do not normally work
well with highly inductive or highly capacitive loads,
although some circuit tricks can help.

The triacs or SCR's in ac solid state relays also have a
MINIMUM current requirement. Thus they are not
suitable for very low current or highly variable loads.
Proper testing also demands that reasonably heavy
loads be used. Fifteen watts is often a good choice.

An ohmmeter is totally useless for testing ac
solid state triac or SCR style relays.

DC solid state relays instead use power MOSfets
for their output stages. The stages turn on when the
controlling signal is applied and back off when it is
removed. These are typically a lot more expensive
and are usually intended for a single output polarity.

One newly popular variation is called reverse phase
control. In which a lamp load is turned on at the
zero crossing and off in proportion to the brightness.
By slew rate limiting the turnoff, the usual dimmer
RFI can be dramatically reduced. Special devices
and techniques are required for reverse phase control.

Some of the local emergency services are switching
to a narrow band VHF system that only has one half
the FM deviation of the previous systems.

On one hand, this doubles the number of available
channels. On the other, it adds millions of dollars of
unneeded costs to new comm systems.

If you have a newer pager, usually narrowband can
be reprogrammed. The frequency and squeltch codes
may also need changed.

On older pagers and some older scanners, these
may still work and may still have the same sensitivity.

But will have only one-half the output audio volume.

The future ain't what it used to be.

I've pretty much convinced myself that legitmate
energy efficiency breakthroughs continue to be
of neglible interest and have zero market.

Yes, there are all sorts of "steal the money" scam
subsidies for business-as-usual that still do the wrong
things in the wrong way. And all sorts of grant
writing scams that bypass the real breakthroughs.
And money being thrown at "ain't gonna happen "
pseudoscience.

Proof of negligible interest are smart power meters
that are closed for utility use only rather than open
for owner energy conservation , power sensors that
are being abandoned as unsellable in warehouses,
deep distress discounts on Black and Decker energy
monitors, Google cancelling their entire energy powermeter
awareness program and, of course, my own energy
efficiency breakthroughs largely being ignored for decades .

Just discovered an amazing piece of technology called
a Rogowski Coil. These have been around since 1887
but only recently have become a major new way to measure
ac current ( and thus energy and power ) for energy
conservation and awareness.

The concept is simple enough and is based on good
old e = L di/dt. The problem in the past had been
building a decent integrator. Doing so digitally
appears to solve the problem. For another definition
of an integrator is a digital network that has a
slope of 6 decibels per ocatave and a phase shift of
90 degrees.

Compared to a classic current transformer, you have
an air core loop that is linear and and can not saturade.
The usual form is a is a coil that that is a toroid that
completes nearly one turn and then closes directly back
through itself.

One popular version is the Fluke i2500-18
Useful general tutorial can be found here
and here.

One of the key secrets to dealing with pseudoscience is
being able to separate the useful adjuncts to porcine whole
body cleanliness from the total hogwash.

More details here.

I've pretty much convinced myself that the recent
stimulus boondoggle did not have so much as one net
cent of economic benefit.

In this analysis, the cost of providing rural broadband
ended up around seven million dollars per home. But
I suspect few of those homes actually needed or ended
up using the web all that much in the first place.

Our local fiasco Du Jour involved trucking plain old
dirt nineteen miles to slightly alter the turnoff to an
obsolete road that nobody uses anyhow.

We picked up half a zillion Allegro Microsystems UDN2597A
octal interface chips at an aerospace auction and will
continue to offer them on eBay. at less than fifty-six
cents each.

Although no longer current production, these remain
absolutely ideal for mid level microprocessor output
interface, and are almost indestructable, even in
student labs.

The octal outputs are all saturating NPN for cool
operation and high efficiency. And are rated a full
50 volts and one ampere.

The eight outputs can handle a pair of medium power
steppers or up to eight power relays at once.

Analog Devices now has a wide variety of power line
monioring chips. Some of which sell for btween $4 and
$8 dolars each.

Some examples appear here. With tutorials here and
here. With bunches more onsite.

Most of the television sets being sold today are totally
worthless scams.

There is not the slightest doubt that the overwhelming
majority of media delivery will shortly be web based,
interruption free, and able to be be started, stopped, fast
forwarded, and rewound instantly at your command.

CRT televisions are, of course, obsolete. As are
plasma displays and all 720P resolution displays.

A "smart" TV that is going to have a useful lifetime
must be wireless internet capable, and must have instant
and no hassle access to the entire web. And not just to
a limited number of paid subscription services.

The TV also must have some sane method of conveniently
and remotely entering alphanumeric data.

Our own links to web based tv appears here.

We just picked up some rather impressive aerospace
grade electronic components at a recent auction. I'll
start listing them shortly on our eBay auction site.

Typical pricing will be one-sixth of the usual
electronic distributor net. Some of the items
have recently become unobtanium and might
be extremely difficult to find elsewhere.

Most items appear to be new, unused, clean,
and still in original packaging.

I'm wondering if there might not be yet another hanging
canal linking the Shingle Mill drainage with the area around
Cluff Resevoir #3.

The topography is favorable and canal systems are known
to the east and west. Others have mentioned this possibility.
Acme Mapper does give some tantalizing hints.

Some auctioneers are now offering only a fraction of their
live auction lots online. Items that do well online usually
have a well defined object, include definitive photos,
have value easily searched, and are of broad interest.

Lots that tend to do poorly online will be somewhat vague,
such as "contents of room" or "contents of cabinet" Besides
getting lower online bids, these are more likely to be
noshows at pickup time.

All of which shades points in your favor to attend live auctions
rather than bidding online.

More on auctions here and on eBay opportunities here.

I'll be giving a talk on the Mount Graham Tramway tomorrow
Thursday June 30th as a BLM Brown Bag presentation at
12:00.

BLM is located at 14th Avenue and 8th Street in Safford AZ.
You are welcome to bring your own brown bag lunch.

The CIA World Factbook site has bunches of useful and
intriguing info on it.

These comparison pages are particularly scary.

Very few stainless steel alloys are magnetic and the
even fewer that are will be only weakly so.

A magnet is thus a very useful tool at an auction to tell
whether you are looking at plated steel or stainless.

Note that aluminum will usuall scratch very easily with
an ordinary knife, and stainless will not.

A very useful websites for tracking down electronic
and semiconductor logos appears here.

Outside of that Missus Lincoln, how was the play?

Texas Instruments has come up with a new MEMS
thermopile system that lets you do non-contact
termperature measurements cheaply and efficiently.

The TMP006 specs can be found here and a USB
evaluation unit here.

The definitive location of the Mount Graham Tramway is a
knob once called Alabam Point. This name apparently has
fallen into disuse, but can be found on current topo maps as
an elevation point 7565.

Looking North from Alabam Point, the entire tramway
route down shingle mill canyon can be viewed at once.
Looking South across a deep canyon, the Graham
Escarpment looms above you. Crossing the canyon
was done in a single spectacular tram jump that was
2600 feet horizontally and 800 feet vertically.

There are hints of a tension station on this point
as well.

The access terrain to Alabam Point would make a
marine drill seargent blanch. You can't get there from
here. Which is one reason that numerous artificats
still remain in the area. Most buried in high brush.
Amazingly, recent fires have pretty much spared
the Shingle Mill canyon area.

Further modified and improved our ultra speed
Magic Sinewave Calculator.

You can now directly export quantized delay values.

Please email me with any suggestions, improvements,
or problems.

Going back to our "Throw another few kilobytes of
memory at it", with larger 16 or 32 bit embedded micros...

You can easily do a precision delay by using some
seemingly ridiculous brute force table lookup programming.

Just set aside, say, 4K worth of NOP's followed by a RTS
subroutine return. The rest of your precision delay code
then becomes amazingly compact, clean, and largely
free of "pinch points".

More on the Magic Sinewave use of precision delays here.