Controller Output

Why do you need two thermocouples for point sensing? I wonder.

well, I dont know either why they used the two-sensor-for-one-controller design.

I forgot to tell, its service job. I calibrated instruments for other companies, specifically, this company is a cookie factory. The instrument are used on their HUGE oven, where the oven is located in line with the conveyor.

Thanks

Thermocouples are just non-linear voltage sources remember. Two different voltages in parallel should average, although I have never tried to put two into a single input. It should be ok.

Como to think of it they put to thermocouples on a single input for redundancy. If one tc fails there´ll be no loss of loop integrity.

I agree with the idea of putting them together in parallel for redundancy. I`ve seen that in other applications. But these paralleled sensors had always the same location - you should never put them together to average temperature depending voltages because the nonlinearities can force unacceptable errors if the temperature difference is too big.

Hmm, that's never occurred to me, connecting two thermocouples in parallel. I've never seen it done or recommended in any document. I've always assumed that it should never be done since it's never mentioned.

We use dual-sensor thermocouples in our factory but each goes to a separate controller. One is used for control and the other is used for the safety system. There are instances where the second sensor is not used, in which case it becomes a spare sensor in case the other one becomes defective. In my experience, however, when one goes bad, the other goes bad soon after.

So what's the answer?

I lack the expertise to determine what the result would be (averaged/lowest-of-the-two/highest-of-the-two). They won't add, I'm sure. In any case, since there's no mention of such a practice in any document, it's probably not recommended to parallel them. I'll have to verify that statement, however.

Okay, I Googled thermocouple +"parallel connection" and got this. It says that parallel connected thermocouples will average the temperature readings of the two sensors.

It's not something I've done and maybe something I might not recommend. For one thing, if one of the thermocouples go bad, the controller will not know it. Some controllers will tell you if the thermocouple has broken but if they're in parallel and only one breaks, you won't get that alarm. If this is just for monitoring, it might be okay. For control, I wouldn't even consider it.

I did not try before telling you this, I think avaraged out put from TCs will be reaching the controller. One more point is, if you disconnect one TC still the temparature will be same, provided they identical and located at same place. Suppose one TC is at higher temparature and the other at lower say 63 mV and 43 mV the indicator will read temparature little more than that of 43 mV. The voltage drop across 43mV TC when (63-43) mV applied + 43 mV this much more temparature.

Hello,

How big is your oven?

for big ovens, and for each control loop one may use one temperature input for the quantization of the control action (heat input) and the other for the reference tracking on the other side of the oven, where the temperature difference may be significant.

In this way you can assure that the temperature is not too high (at the input) nor too low (at the other side of the oven) and your cookies are tasty.

In other processes, you could need even more thermocouples for detection of spontaneous exothermic phenomena.

Or if your cookies are really spensive you could need redundancy in the above mentioned sensors to assure you don't stop the process and spoil the cookies in the middle of it.

Hope this helped

ARE THEY CONNECTED TO THE SAME INPUT. IF THEY ARE THE CONTROLLER COULD HAVE THE ABILITY TO AVERAGE, YOU DON'T SAY WHAT THE CONTROLLER IS OR HOW THE TC ARE CONNECTED. IF THEY ARE ON THE SAME INPUT, THEY WON'T AVERAGE, YOU WILL READ THE HIGHER INPUT. TO AVERAGE THE INPUT YOU NEED AT LEAST 4 WIRED SERIES PARALLEL.

Dear OP,

At a company I used to work at, we actually used 3 TCs in parallel. Some testing revealed that the result was an average of the temps at the three measurement points. The accuracy was well within the accuracy of the TCs themselves.

Mike