Eventually, I figured out that the Oxford VT controller was
using a couple of voltage-to-frequency converters to measure
the output of the probe thermocouple and of a junction
temperature compensating chip mounted on the back of the unit.
The register used to store the output of the temperature
compensation voltage-to-frequency converter was apparently
overflowing under some room temperature conditions in a way
that made the probe temperature look negative. I forget the
details, but changing the value of a resistor inside the VT
unit cured it, with the only side effect being that, when the
thermocouple becomes disconnected, the run up towards maximum
indicated temperature is slower than before.
I remember that for several months, before making the
modification, we did not dare leave the instument unsupervised
with VT running.
John M. Wright, Dept. of Chemistry & Biochemistry, M/S 0314, UCSD,
La Jolla, CA 92093-0314; email: jwright@ucsd.edu; phone: (619) 534-3049
> Date: Tue, 6 Jan 1998 12:07:52 -0800 (PST)
> From: lew cary <lcary@chem.unr.edu>
> To: ammrl@wwitch.unl.edu
> Subject: More Vt problems!
>
> When I walked in Monday, the unity500 had , the day before, toasted a
> sample! Overnight, the temp had gone up enough to pop the cap on a cdcl3
> sample, and evoporate the solvent, and cover the top of the spinner with
> solid sample. With slight cleanup every thing seems normal.
> We always run the vt on and at 25C. One probe has an intermittant
> with the thermocouple. The cable is wonky, but with care works with all
> other probes. The symptom is always the same: the appearent temp. goes to
> max., but the sample is not heated.
> My question is: how could a thermal runaway have happened? Over
> temp shuts off the heated in all cases that I have seen. No one was there
> Sunday, of course.
> What a great way to start '98, Lew
>
>