Temperature measurement

Are you krw posting under a pseudonym? I never claimed a specific offset.

Five degrees would be a large offset. Since I was talking about the way the defects in the description of the way that the calibration points were set up reflected the defects in the way Ken S.Tucker thought about the problem, rather than the defects of the "calibration" he produced, this is totally irrelevant.

You are remarkably confident about what his - poorly described - method might have achieved.

Perfectly true. But it's easy to set up a temperature calibration point so that does set up a reliable and predictable temperature, and pointless to set up one that doesn't.

It's ironic that you - of all people - should think that.

-- Bill Sloman, Nijmegen

But how would Gibberman know if anything was accurate enough?

-- Bill Sloman, Nijmegen

And your estimate is?

-- Bill Sloman, Nijmegen

=A0To

Not 1% of the absolute temperature?

They'd be black body radiators. The temperature depends on the fourth root of the power you dissipate in them.

And you wouldn't use one to calibrate a diode-based thermometer unless you were a remarkably extravagant idiot.

-- Bill Sloman, Nijmegen

Yes, you did. You stated in your first response to me that it was "a few degrees".

Now, you want to claim great error, but also want to claim no offset?

Make up your mind, asshole.

Are you Sloman posting as Sloman? Oh yeah... that's right... You are.

What is pointless is to define those points at a greater accuracy than is needed for the purpose.

You lose, again, you pointless fucktard.

It would have been stated differently then.

There are different "types" of tolerance on stated ranges of devices.

There are declarations of a given percent of the reading, and then there are accuracies declared as a percentage of the span.

It is probably 1% of reading, which would have to have been stated differently.

Yes, and we manufactured them. We are, in fact, one of the few places you can get such a source.

Of course, the bulbs can be found, but the calibration instruments are what one needs.

The device I mentioned is for IR thermometry, you retarded twit.

For you to think that it was referenced as a tool for a diode transducer makes you the idiot... that wishes he was extravagant.

The only thing extravagant about you is the level of your stupidity.

"a

Really? I can't find any such statement, and I've been through the date-ordered thread from you first post. You need to identify the relevant post by date and time if you want anybody to believe you.

I've never claimed "great error". I claimed that he wasn't setting up his calibration points correctly. I didn't make any claims about the size of the error involved - it's not a mistake I happen to have made, so I simply don't know how bad the error gets, except that it is big enough that the NIST researchers worried about it.

-- Bill Sloman, Nijmegen

But - unless you do it right in the first place - you can't know how accurate your improvised approximation is. Once you have done it right, you are in a position to measure how wrong a sloppier set-up is, and compare this error with your error budget.

You have guessed a figure for the error, and on the basis of that guess you thing you can define an "adequate" temperature calibration set-up for a project that you know very little about.

That doesn't make even a glimmer of sense.

-- Bill Sloman, Nijmegen

The universe is full of black-body radiators. Your device may have advantages in particular applications, but every filament lamp is a black-body source, and filament lamps are not - at the moment - all that hard to get hold of.

To do what?

No need to tell me again. It is an expensive method to measure temperature, and not all that precise. If you don't need non-contact temperature sensing you can get better accuracy with much less expensive hardware.

-- Bill Sloman, Nijmegen

I thought of that, too. But then you need a micro, as opposed to just counting. ...Jim Thompson

un

Poorly written specifications should be returned for clarification.

You made an inference of "great error", asshole, as a "small error" would not have been worth mentioning.

Who needs an A to D?

Use a micro. Charge a capacitor through a reference resistor. Discharge capacitor. Charge through thermistor. The ratio of the charge times is the ratio of the resistances. Use a look up table to find temperature from resistance.

As the thermistor is only powered while charging the cap dissipation is not a problem. And the capacitor makes it fairly noise immune. It just requires 2 micro pins.

I've got about 100,000 units out there using this technique and haven't found a problem with it yet.

good

No, clueless gibberish is clueless gibberish. If you don't want clueless gibberish called such, STFU.

I have seen more analog scales, d'arsonval movements, taut-band movements, and the circuitry that drives them than you have.

What do you want? Linear? mV/degree? 4 to 20mA?

And we are not even mentioning the digital models. $2500.00 each back in '85 dollars. And everything was traceable. We worked with NIST directly, and were among few IR companies that did. To this day.

They are actually doing record breaking black body 'surface panels' that are used to calibrate FLIR's imaging arrays. Uniform temperature over a *very* large area is *very* difficult to do due merely to random, Brownian air currents alone. Of course, in use, it gets no such exposure, because the arrays are very close to it.

I had small scale designs for a thin faced, oil bath backed, Al ingot based source, at 4" with a working port opening of about 3.75". More accurate than our thick faced, soak and settle ingot based job with a heat rod shoved in the center of the rear of the ingot. Very stable, but the settling time from a set point change is pretty slow. It was (they are) about 13" W x 9" H x 11" D. Pantented paint formula (and surface texture) on the face, and it was at or better than .98 emissivity across the entire face and variance from center to edge was nil.

We made (they still make) one for 4500°F that is a wad of silica (plaster of Paris) laid onto a basketball (literally). Build two hemispheres of that, take the ball out, and bond them together, then wind a shitload of heavy (18Ga) coiled Nichrome wire (spring :-)) around it and lay out the termination leads. Be sure to make the port for the silica tube. THEN, plaster up the exterior with an additional gob of wadding till you get an inch of it over the wire coils. Cure that, and then pump some current into the wire to cure from that source as well. It looks like a lolly pop with a short stick on it. Then, take that assembly and mount it into a cabinet that is wider than any standard rack cabinet (and squat too).

Suspended inside the cabinet, pack around that with about 1 foot of standard fiberglass kiln insulation wadding.

Wire up the power, the Solid State relays, and the pid switcher to pump them. Option of a high temp calibrated thermocouple probe port.

The front of the ball has a 6" x 2.25" ID quarter inch thick silica tube that makes the black body source port. The body is biased toward the front of the unit a bit.

Calibrated it, and you have an extreme temperature black body calibration source so folks can calibrate their sensors.

How's that for a glimmer?

You lose.

The lamp has a CALIBRATED filament. You need to know what that is. Obvious that you do not with that remark.

Then, the electronics behind the driving of it need to be there as well as the cabinet it would be mounted into.