A while back, I showed a pic of the back of my pyrometer gauge, whereon I had penciled the colors of the thermocuouple leads.
Well.... It turns out that Tom B. knows something about thermocouples and knew damn well that I had the leads hooked up wrong. It seems that in American "K type" thermocouples,
the red wire is always negative. Since it had been hooked up this way for nearly 15 years, and all of my heat treating experience in that time had been based on numbers from this pyrometer, I was less than delighted to contemplate the repercussions of this revelation.
Although there was a strong temptation to leave things as they were, I decided that willful ignorance is never the correct path.
So... I opened things back up to try to get to the bottom of it all. I turns out that "K" thermocouples
https://en.wikipedia.org/wiki/Thermocouple#Type_K are made of the alloys Alumel and Chromel. Of these two alloys (both of which sound like they are probably Superman's uncles, or perhaps lesser known Archangles) only the Alumel is magnetic. This gave me a way to determine which of the two legs of the probe was which.
I used a rare earth magnet to figure it out. Here we see that the probe leg on the left is Alumel. The same test showed that the red wire was also Alumel. Now.. thermocouples being what they are, the Alumel probe leg should be attached to the Alumel lead wire, and vice versa for the Chromel side. I had wired the leads up backward at this junction block, as well as at the meter!
My first thought was to get the leads hooked up correctly and start in on a series of experimental hardenings to reestablish my protocol. This was madness, of course. In a bit, I realized that I needed to leave them like they were and run a test firing with another pyrometer so that I could compare the temperatures and get a sense of how far off my numbers had been. I knew "how I was wrong" and now I needed to find out "how wrong I was".
Dave Rylak kindly put his spare pyrometer in the mail and I had that in hand the next day. Thank you, Dave!
The first four columns in this chart show the temperature as displayed on Dave's digital pyrometer, the display on my (incorrectly wired) analog pyrometer, the time of the measurement, and finally the amount high or low that my pyrometer read.
The results were somewhat surprising. At room temperature, my pyro was more or less correct, but it soon came to read low. By the middle of the tempering range it was about 50°f low. This error gradually diminished as the temperature increased, until at about 900°f I was getting good readings. Above than that, it began to read high. The error increased gradually until it was 50°f or more too high in the austinitizing range. I had assumed 15 years ago, that if I was getting "plausible" temperature numbers, than I must have the thing hooked up correctly. That turns out not to be true.
Having found out the extent of the terrible truth; I rewired my leads at the junction block and the gauge, and ran another test firing. I made a couple of recallibrations to get the analog gauge to agree with the digital one, and made sure they were pretty similar at around 1500°f. For a while it looked like "the man with two pyrometers never knows what temperature it is", but I seem to have that straightened out. The two gauges now more or less agree from 1750°f down to about 100°f. That's were the kiln is now. If they continue to agree at room temperature, so much the better... but that does not really matter. The important thing is that I can now trust my pyrometer to be within a few degrees of the truth across the range useful range.
Now.... Now, I can get on with the hardening experiments.
Mac
