Thermocouples

[Dick White]

This is a technical question about thermocouples for the collective expertise of the excellent hive mind we have here. Is it possible/likely that a thermocouple will fail by outputting a steady unchanging voltage regardless of a change in actual temperature? As an example, a kiln is firing in an ramp up segment to, let's say just for the purpose of the discussion, 2200℉. The current firing status is at 2000℉. The thermocouple outputs xyz micro-milli-volts (whatever, not going to look it up, exact number is not the question) which the controller correctly interprets as 2000℉, and because it is in the midst of an up ramp, closes the relays to turn the elements on for more heat. Is it possible/likely that said thermocouple might suddenly freeze its output voltage and remain steady at xyz micro-milli-volts despite the presumed actual increase in the kiln temperature resulting from the controller having turned the elements on? If such a scenario were to occur, I would think the built-in safety of the controller's E1 error will turn everything off after the appropriate period of no material increase in apparent temperature as measured by the thermocouple's electrical signal, so no harm no foul. Rather, I am having trouble wrapping my head around having heard that inside company technical reps from two of the major kiln manufacturers have told customers that an appropriate resolution to an E1 late in the firing is to replace the thermocouple. How do we get there from here?

[Bill Kielb]

I am not aware of that failure mode of thermocouples so I believe constant output is highly unlikely. Thermocouple circuits these days contain open and shorted circuit detection and many detect  the resistance of the circuit in operation. Often comes right on the op amp. Thermocouple drift is a thing after many cycles and is generally downward. Once the junction fails it becomes an open circuit. Most kiln thermocouples I see fail open  as they corrode completely through.

The kiln rate of rise, or rate of decline is detectable and the controller ought to detect that error. Error 1 is failure to rise 12 degrees per hour.  Highly unlikely in my view to be constant output under varying junction temp. It would be interesting if the failure could be reproduced as well. Simple visual examination should reveal a well worn thermocouple as well. Generally the failures are obvious, not always, but generally. For exposed tcouple we usually replace when the ceramic at the tip cracks and begins to expand  due to the corrosion underneath. In advance of the inevitable failure.

[neilestrick]

I don't think a thermocouple can fail by creating a static output. When they fail it's because the connection between the two metals is no longer viable. That can be because it's been fired to death, or because the weld has failed prematurely. I've had brand new thermocouples with bad welds. I can be a total fail where it stops sending any signal at all, or it can cause a fluttering situation where it's sending an erratic signal. The erratic signal can cause an error code, but I've also seen the flutter actually cause the kiln controller to stall out, where it's just holding temp for an extended period, like hours, without putting up an error code. Basically the thermocouple is jumping in and out of the accurate temp, but never long enough that the controller recognizes it as an error situation, and never long enough that it gets any temperature rise.

If the controller is giving an Error 1 at the high end, there are a few possible reasons. My first inclination would be to check the relays, as when they are on their way out they tend to stick on or off when they get hot (both from the heat of the kiln and the heat from power going through them) near the high end of the firing. It could also be the thermocouple, by way of the metal expanding as the kiln gets hotter. If the weld has failed, it may only open up enough to cause the fluttering or inaccurate reading once the kiln is super hot. I would also check all the thermocouple connections. I have seen several kilns in the last few months that were having problems because the screws in the thermocouple terminal block had loosened up from the expansion and contraction of the metal. Also double check that the thermocouple wires haven't broken off under the screws in the block. 

[Dick White]

Thanks Bill and Neil, your interpretations are consistent with the logical construction of my non-engineer's mind. Failure through corrosion or breakage into an open circuit (or a wiring fault) will produce the FAIL error, but that's an all or nothing proposition. Drift as a failure condition is an overall shift of the voltage pattern, but the output will still vary proportional to temperature and would simply cause an inaccurate firing that should be detectable by using witness cones. And weird flutter could send the controller on a joy ride, with no particular error code. But these were claims that Skutt and L&L factory personnel told the customers to resolve an E1 (which is quite specific - no material temperature increase, i.e., no voltage variance from the thermocouple, over a modest (18 minutes?) continuous period of heating demand) by installing a new thermocouple, and if that didn't work, then replace elements or relays. Huh?

[liambesaw]

Why not, that's an easy sale!  

I have gotten err1 my past 2 cone 6 firings, even though witness cones show it reached cone 6.  But they were very long firings and I already knew the err1 was coming soon.  

Also, the controller doesn't take a snapshot of the current temperature and then make a decision, it is continually monitored and averaged, and you can check the current average rate of climb by pressing 5.  So even if it's varying up and down a bit it should still fire somewhat normally.  

I could see a situation, however, where the thermocouple begins to act more like old elements and the overall resistance drifts at high temperature, making it read lower and lower, causing an err1 at peak.  So might be one very specific situation where changing the thermocouple could get you back into business.

I don't know though, I'm not an expert at this stuff

[Bill Kielb]

Hah, I like the elements and relays ending. Sounds like a scripted response. I think error 1  is at least a 12 degree per hour increase BTW.  Elements worn to that level are done!

Drift as I understand it is a changing material composition thing, so it does not mean drift at a temperature, just drift over time from calibration.

[Dick White]

Looked it up, E1 is <12℉/hr over 22.5 minutes, or less than 4.5 degrees total increase in that period. In my experience, yes, either the elements are done (or one of them is failed and the others can't take up the slack) or a relay is stuck open causing the same effect as if that element is broken and the rest can't take up the slack. The TC didn't do that, it just reports reality.

Thanks, you guys keep me sane in an insane world.