At 12:36 PM 8/14/02 -0700, you wrote:

>I hope you don't mind too terribly much if I just disagree with you a
>bit there.  The specifications for the Fluke 80TK thermocouple module
>seem to belie your opinion about where thermocouples can be used.

Excuse me? I did NOT say it could not be used or would not "work" (in
the sense of giving some indication of temperature), I did
say that it was *inappropriate* for this application.  Straw man?

>Measurement Range:
>
>- -50=BAC to 1000=BAC (-58=BAF to 1832=BAF), depending on thermocouple=
 probe
>used.
>
>- -20=BAC to 350=BAC: 0.5% =B1 2=BAC


Yes, not very good, for this kind of biological application, is it?
And what warm-up time do they specify to achieve this level of
accuracy? And what thermal gradients can be expected outside this
tub of relatively constant temperature liquid?  What kind of
errors in any panel jacks etc. that might be involved? Errors
due to the compensation leadwire not being quite the same alloy?
There is an inaccuracy due to the probe as well, typically
special limits of error alloys are used, but still the errors
are significant. A decent RTD system would be around five times
better accuracy at a similar price, and have completely
interchangeable probes. Which is why they are used
almost exclusively in industry for this sort of thing.

Do you know anything about thermocouples and their conditioning
circuitry?  There is a temperature sensor inside the Fluke that
must *track* the junction between the copper wire inside the
circuit and the thermocouple materials more accurately than
the desired system accuracy. This can't be done perfectly,
particularly not on a very inexpensive instrument like this one
which doesn't have a lot of effort put into maintaining isothermal
conditions around the junctions. A 2' error in tracking means
a 2' error in the reading (the thermocouple voltage is (to a
first approximation, it's not exact, nor is it linear)
proportional to the difference in temperature from one end to
the other. This introduces lots of sources of error.

>You will notice that the 10 to 30 degrees falls well within the
>optimum range of -20C to +350C (-40F to +662F)

Optimum? Hardly.

>So, I will have to
>say that I am of completely the opposite opinion on this one.  10 to
>30 degrees, be it C or F, falls well within the prime operating range
>of thermocouples.  Well, Fluke ones, at least.

No, it is an area of temperature that is *usually* best served by
other types of sensors. Most temperature sensors will "work" in this
temperature range- semiconductor, precision thermistor, base-metal
RTD, precious metal RTD, most thermocouple types (not type B or some
exotics), IR, etc. etc. T/C's are best reserved for applications
where their ruggedness, speed of response, high or low temperature
capability (or potential, anyway) etc. justify their disadvantages.
This, quite simply, is not such an application.

>With regards to the immersion probe which I mentioned,
>80PK-2A K-type immersion probe
>Measurement range: -40 to 982=BAC (-40 to 1800=BAF)
>
>It looks like it fits too.

Yes..  believe it or not, you can get probes and/or protection
tubes made to whatever length you want, within reason. Cheaply.

>Now, if you were to tell me that the problem was something along the
>lines of there being an issue with heat flow, I'd accept that, but,
>having read fluke specs before, "Thermocouples don't work in that
>range" just doesn't do it for me.

"Heat flow" would be B.S. A big tub of liquid like this is about the
easiest type of system to measure accurately.

>Electrosonic shows the 80TK thermocouple module at $124 each, and the
>80PK-2A immersion probe at $83 each.
>
>All in all, I think it's quite a viable solution.

If you don't look too close.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the=
 reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
9/11 United we Stand

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu