Dear Moderator,

What about establishing a subject category for "Control History"?

I am intrigued by the ingenuity of those who have come before me. As neat as a modern algorithm is, the solutions to problems that today are not even seen as "problems" are fascinating.

I was inquiring recently about how thermocouple cold junction compensation was accomplished back before my era of cheap non-linear thermistors, microprocessors and memory.

I got some references to the L&N Micromax recorder as using a standard cell against which a dry cell was periodically referenced and nulled, but I didn't get the whole story as to how that related to cold junction comp.

As a side note, I was informed that it was not uncommon to physically bury a thermocouple junction X feet, Y inches below ground grade, knowing that the ground temperature was stable at, for example 57°F. That stable 57°F provided a constant temperature reference point, similar to that of an ice bath, i.e. a known temperature. The difference between the ground temperature and freezing was adjusted as a mV calibration.

So, are there any old timers out there who know the whole story of the L&N Micromax as it relates to cold junction compensation for thermocouples?

Dan

 

Excellent idea, a lot of smart people have forgotten more than you or might ever know.

Bill Sturm

 

I can't help you with the L&N Micromax, but if you are interested in old technology, then have a look in used book stores for old electrical engineering handbooks. I have copies of hand books and text books dating back to 1908.

You'll typically find that the older the book, the more deeply it will go into the details of how something actually works. By comparison, many if not most of the newer books published today are complete rubbish.

 

I'll bet that many of the folks who would know are also buried x feet y inches under the ground. But I agree, many of the early pragmatic solutions are intriguing and thermocouples have had a whole world of paraphernalia, old and new, to make them work. It's hard to believe the lengths that the early users went to to use them. Tiny output, non linear, difficult to connect to and requiring great understanding to use with accuracy and reliability, they were (and are) nonetheless one of the best options, especially when many points need to be monitored. Judging from the traffic and questions here and the reverence you receive if you know how to apply them, they are still not well understood. I never saw the "ground point" reference, but I remember first using a ancient leather bound wheatstone bridge type instrument to verify oven temps and the like in a lab in the 70's. That later culminated in a diode and op amp ice point reference just to see if I could do it. At one time you could have bought dozens of the special switches and connecting blocks and metering and miles of wire surplus for pennies after the war years, but it's getting spendy again and there is still no better option for many situations. I haven't put my twin carbon couple melter away yet.

Regards
cww

 

My 1935 "Industrial Electronics" handbook limits its discussion of thermocouples to the problems of amplifier circuits:

" Electronic tubes may be used to amplify the output of temperature-indicating circuits operating on the variable resistance or the thermoelectric principle. When the indicating circuit operates on the resistance principle, the exploring resistor should be arranged as one branch of an a-c. operated Wheatstone bridge and control circuits similar to Fig. 215 should be arranged. When thermoelectric indication is used, the problem of amplification becomes more complicated because the output of the thermocouple is a d-c. voltage of the order of millivolts. A d-c. type amplifier for the high amplification required is not very stable, as outlined in section 33. Better results are obtained by using an a-c. capacitor-coupled amplifier. The thermocouple voltage can be modulated by means of a continually vibrated microphone which varies the resistance in series with the thermocouple."

Another sort of temperature regulator not involving thermocouples is suggested:

"... the breakdown characteristic of a grid-glow tube depends upon the temperature of the mercury vapour. This characteristic has been used to control the temperature of an oil bath in which the grid-glow tube is submerged. A regulator of this type is inexpensive and is sensitive within a few degrees. The range of a regulator of this type when a standard grid-glow tube is used is limited to a temperature range of approximately 25 to 90 deg. C."

 

My first exposure to thermocouples was with a (then new) Honeywell potentiometric recorder. These used a vibrator - like the ones used in car radios to turn battery DC into AC for transformation to valve anode levels. I think they were driven by the mains but may be wrong as this could give some interesting effects with induced mains voltage.

The t/cs were in series with the amplifier input, across the centre arm of a bridge. To detect up-scale burnout a small potential was applied to the circuit between the t/c and amplifier input - unfortunately, this put sufficient current through the t/c to give an appreciable voltage drop. With a 36-point recorder on a reasonably large boiler, there were some interesting variations between recorded temperatures depending on how the wires to the individual couples were run. We had to spend a couple of weeks measuring the lead resistance and making up individual ballast resistors to balance out all the circuits. Those were the days...