Temperature measurement

His plastic brain no longer has a modulus of elasticity. He is hard wired dumb, and hard frozen as such.

I doubt it was the drugs. Far more likely the lack thereof.

He is too dumb to do "prude" though.

Hear, hear.. there, there... Good job, dude.

Likely wrong to guess too, so I am likely wrong since the odds are low that I guessed correctly.

Far out. Probe on a schedule, and plot it out perfect for data logging. Broken parts would show immediately too. So it HAS to work.

Yes, and you can put some 'gear' in that circuit for ambient compensation(s) as well.

Make that sucker track straight in any environment.

Absolutely.

Same thing works if you want to increase hi value resistance string value accuracy resolve.

Ten one percent, 1G Ohm HV flat sticks in series will be closer to 10G Ohm than a single one percent, 10G Ohm HV resistor typically will. They also get stressed less under use, which makes them shift value and throw 'observed' figures.

Does anybody have any data on that?

I'd expect the machine making resistors to have repeatable errors so that if you pick them off the same reel, they will probably have similar values.

I'd expect something like that might work if you purchased reels from different vendors, or vastly different times if you use the same vendor.

--
These are my opinions, not necessarily my employer's.  I hate spam.

[...]

Nice. Maybe like this?:

.--------. . | . PIO1 |---[Rt]--------. . | | . PIO2 |---[Rref]------+ . | | . | --- . | --- . | | --------' --------

I suppose you can use one PIO to sense the voltage while the other is switched high for charging. Making sure to disable any internal pullups! I wonder are the input thresholds of a typical micro matched well enough for this? Or use a single third pin as the sense.

.--------. . | . | Sense . PIO0 |---------------. . | | . PIO1 |---[Rt]--------+ . | | . PIO2 |---[Rref]------+ . | | . | --- . | --- . | | --------' --------

PIO0 could be the input to a on-chip comparator.

How accurately can you measure the resistor using this technique? It looks like it would be limited by digital supply noise? But that could mostly be averaged out too.

--

John Devereux

So how accurate *are* they? (At around body temperatures, the subject of this thread.) Do you have any numbers? Like +/- 0.1K?

[...]

--

John Devereux

You don't get it. A one percent move from one G is lower, but the device is still likely to be close to right on the money, whereas the 10G is going to be up to 1G off and is less likely to be right on the money.

ergo, my declarations about it, and that matches years of lab observations too.

My load resistor test banks were a few feet long, and had big round solder nodes on in as the take-off points when one chooses the value for the load. I'll post a photo link when I find it later today.

So, nearly three feet by about 5 inches wide, two runs of perf board. The resistor banks are mounted *under* the exposed 'selection nodes', and the second plate is separated from the first by epoxied spacer posts about 4 to 5 inches apart. It can be left open air (up to a certain excitation level), or one can spray it with transformer varnish.

It is a precision load bank.

Anyway, 1G resistors are typically closer to 1G than a 10G is to 10G.

And the less stress (per element) thing makes it automatically capable of handling higher voltages and holding its value while doing it for any given voltage.

It is like one sting of .3 x 1.25" flat silica bars with thermally conductive nodes between each (not that they heat *much*), compared to a single .4 x 2" flat silica stick. (with resistance media, of course).

I'll post the pic link shortly.

HV resistors are larger form factor devices (and different as well) "flatso" is what they are called.

We made an old analog non-contact that had a one inch focus, and a ball roller to keep the space accurate that chiropractors used to look at spine temps back before imagery moved in.

They were bath calibrated, In a bath and pointed at one of our known bb sources for the upper range at about 43° C, and accurate to 0.1° through the entire range. That was back in '85.

I am sure that a properly configured modern imaging system would print out a proper cal range on the right or left edge of a 'photo' that would be just as accurate. We had a $90k (1985 $) unit back then that had to have a LN2 mini dewar inside of the camera filled that chilled the back side of the IR CCD. That was maybe 640x400 at 4 fps. It was accurate though. There was an option for a microscopy stand and headend for it. That would take it far over $90k though.

Nowadays, it ain't shit to get room temp imaging arrays (talking about the CCDs/CMOS) inside premium optics with great software and recording.reporting/plotting capability for far less.

We had to have a big dewar of LN2 on hand (have to admit that was fun).

I am surprised I didn't see an imager at Harbor Freight!

Well, it was two feet long.

s from

I've

which

expect

can

t by

you

't a good

=A0No,

ss

I don't particularly care what you call it, but it is worth reminding the lurkers that you are clueless idiot, and your opinions a waste of bandwidth. It would help if you did desist from posting you bizarre opinions, but you are a little too far gone for there to be much hope of that.

-- Bill Sloman, Nijmegen

An electron beam tester is an electron microscope modified to measure the energy distribution of the low energy secondary electroncs scattered from the surface of a working integrated circuit (with the lid taken off its package, obviously). The energy distribution of the low energy secondaries is a direct reflection of the electrical potential at the surface at the time the secondaries were ejected, so the system functions as a sampling oscilliscope for integrated circuits, with the focussed electron beam taking the place of the scope probe.

It really isn't suited to testing tubes. Those particular machines incorporated the patented electron beam deflection electrodes that I'd invented for the job.

More pretentious bullshit ...

None that you would know anything about. I compared my experimental results with predictions of H.S. Johnston's "ultra-simple" rate theory, as set out in his book "Gas Phase Reaction Rate Theory" published by the Ronald Press in 1966.

but he was more of an experimentalist than a theoretician - he's famous for his work on the destruction of the ozone layer.

Scarcely. People with bachelor degrees do earn more while the more academic members of their cohort are studying for higher degrees, but once you've got your post-graduate degree you collect more money from the start, and - on average - your income goes up quite a lot faster than the incomes of your less studious undergraduate colleages.

Small quantities of ethanol from time to time.

Or so you think. It has the same status as your - implausible - opinion that you know something about phsyics.

But most dingbats call it something different. And I think you are confusing chemistry with mechanical engineering. A couple of our lectures touched on viscosity, mainly because it is one of the - many

- properites of gases that can be predicted by the kinetic theory of gases, but the deformation of solids was left to mechanical engineers and soild state physicists.

-- Bill Sloman, Nijmegen

ld

It doesn't. It just increases the voltage you get to measure.

-- Bill Sloman, Nijmegen

uld

He's talking about temperature measurment, idiot - that is the ambient variation in almost all real situations. Changing atmospheric pressure can change the thermal resistance from junction to ambient, and thus the self-heating in the diode (or diodes) but that's a little too complicated for either of your tiny little minds.

-- Bill Sloman, Nijmegen

from

which

expect

good

You, of anyone here, calling anyone a clueless idiot is simply rich. That's like DimBulb calling someone stupid. ...too funny!

G y

Only if the ten 1G resistors are drawn from a population whose resistances are normally distributed around 1G ohm. In practice, resistors from the same batch tend to be distributed about an average value that differs from the nominal. One well known technique for making close tolerance resistors is to make lots of resistors whose resistances are loosely distributed about some value close to - say -

1G ohm, and measure the lot, selecting only those resistors that come within +/1% of the desired tolerance. If there is also a market for 0.5% tolerance resistors, the batch you will buy will likely all lie between 0.5% too high and 1% too high - while someone else will get the resistors from the bin that measured between 0.5% low and 1% low.

And the distribution will not be normal, but a narrow slice out of the normal distribution, with a much higher than "normal" incidence of parts at either end of the range.

-- Bill Sloman,Nijmegen

Only if the distribution of the resistance values 1G ohm resistors is normal (Gaussian), which is an assumption that you should have tested.

-- Bill Sloman, Nijmegen