REFLECTOR: Thermocouple Calibration

[Brian Michalk]

This weekend I had the second engine start, and the goal was to dial in 
the electronic fuel injection system.  I was going to do that using 
EGTs.  I had picked four sites on the fuel table and was going to run up 
to peak EGT, then go to 25 rich of peak.

Well, Murphy said it wasn't going to work.  My panel gauges didn't work, 
and my engine monitoring EGT thermocouples didn't work either.  A long 
time ago I decided to terminate the thermocouples in the cabin, thus 
locating the cold junctions in the cabin which is a more stable 
temperature environment.  The gauges are calibrated by Westach for 75F. 
    When I did this, I had to add thermocouple wire to my probes, and in 
doing so made probes that had more resistance than what the gauge was 
calibrated for.  I trimmed them for the same resistance, and just forgot 
that I needed to send the gauge back for a recalibration.
The reason the electronic EGT probes didn't work is because my CPU runs 
on 3.3 volts, and the A/D converter on the CPU can only sample up to 
it's VCC (3.3V).  The AD595 I was using outputs a signal of 10mV/C, so 
above 330C, the A/D sampler is saturated.  So for the sake of dialing in 
my engine, both systems were inop.

I have now added a gain resistor to the AD595 to move it from 10mV/C to 
a more manageable 1mV/C.  However, resistors being what they are, I want 
to calibrate my thermocouples.  I've thought about using a solder pot 
full of solder.  I've thought about using a hot air gun, and another 
thermocouple as a reference.  Boiling water is okay, but I'd like to 
calibrate to something over 500 degrees.

Do any of you guys have any thoughts as to a cheap temperature reference 
I could use for calibration?