Thermocoupler

[McLane]

Yet another question.

I have communicated with Jim off-forum about this, but thought I'd post this question here as well to see if other people have experienced a simmilar issue.

One of the people on the design team has brought a concern to the table that involves thermocouplers. The thought is that these things need replaced on gas oven and similar appliances from time to time. There is a concern that if a thermocoupler goes bad, that we will not be able to replace it.

Here's the question:

Have any of you had a thermocoupler fail?

I have 2 ideas that could help solve a potential problem. One would be to install more than one thermocoupler in a "zone", so if one fails, the other could be simply plugged in. This is what Nasa does with critical systems on the shuttle. Build multiple systems so that if one fails, there are 2 more that can perform. The other idea is to build into the oven, a series of 1/2" stainless tubes that a thermocoupler attached to a long metal rod could be inserted into the area where they would reside. This way, if the coupler failed, a simple removal and replacement could be performed.

What say you folks about these two options?

[n2iko]

The thermocouple itself is just a small piece of glass with a bimetal strip and
a set of wires comming from the different metals. This then in pushed into a stainless steel tube and the tube is inserted into a hole drilled into a brick.

To replace the TC you just withdraw the packing and replace the bimetal unit.
then repack making sure you do not have a short. Leave the tube undisturbed.

The wire set that is attached to the unit must be used as is. It is calibrated that way. Specify the wire length when you order the TCs.


--mr.jim

[Mypsi]

I've been wondering about this, too. Our oven is close to 20 years old, and was around long before I started using it. There are 5 thermocouples, but only 4 give readings. They go into a switch box, then to a thermometer, and all the wires are a little shaky, so it's a balancing act of fiddling with wires to get good readings. I suspect that they are reading low, because I picked up a cheap dial oven thermometer to put inside, and it actually reads around 50 F higher than the thermocouples (that's not right, is it?).

So I have a few questions:

What happens when a thermocouple goes bad? Does it just stop reading, or does it get less accurate?

mr.jim, can you explain a little more how they would be replaced? I've got these wires coming out the front of the brick, that I imagine were placed during construction, and are snaking through inches or feet of brick. Some notes from a previous baker seem to imply that at least one might be in the back of the oven. Are you saying that I should be able to pull this out, replace it, and feed it back in?

If I replace the outside wiring, what kind of wire should I use? Right now it is a thick single strand that seems kind of brittle, and I think I could get a better connection with a thinner, braided wire.

[wlively]

First, therocouples are nice thing to have, but unless you are doing some serious baking or commercial baking they are not really neccessary (good thing my wife does not read this ). I put 3 in because I "just had to know exactly how my oven works", but have not used them of late as I have learned to "read the oven" and have a good idea of what temp it is. So unless you are in the mode of above, you won't be using them much after you have cooked a dozen times or so and if one goes bad, no big.

Ok you want to use them. I put my 3 on the same side or slice of the oven (top, mid, floor) and installed a removable door in the roof, so all I have to do is remove some shingles, then the door, and pull out a TC and install a new one. I would recommend TC's in stainless sheath, they are sealed, so will last the longest and easily slide in/out of drilled hole in brick.

TC's do fail over time. They tend to "drift" or read lower values as they progress to failure, but I doubt you will notice since they tend to drift in single degrees at a time. Even on a TC's worst day it will most likely be much more accurate than a "cheap oven thermometer" so that would not be a good test. Also must remember that the bead, or disimilar metal junction at the tip is what is reacting to temp, so the TC only reads what the temp that bead is exposed to. You need to know where exactly that tip is. There will be different temps at different parts of the oven and that is excluding the convection currents of the open interior.

Here are a couple of failure modes for TCs and two are specific to K-type (because that is the best choice for our ovens, IMHO)
1. Ground loop failure, TC circuit becomes shorted to ground somewhere or shorted to the other wire. TC will read low or close to zero and will change temp.
2. Corrosion due to oxygen of Alumel/Chromel bead, the higher the temp the faster the rate and can be seen by bead becoming a greenish color. Will read gradually lower temp and then fail "open circuit". The main reason I prefer stainless sealed sheaths.
3. Grain growth in Alumel at Chromel junction, at our temps this would probably take tens of years, more applicable to temps above 1000 deg C. Fail "open circuit".

Originally Posted by Mypsi

I've been wondering about this, too. Our oven is close to 20 years old, and was around long before I started using it. There are 5 thermocouples, but only 4 give readings. They go into a switch box, then to a thermometer, and all the wires are a little shaky, so it's a balancing act of fiddling with wires to get good readings. I suspect that they are reading low, because I picked up a cheap dial oven thermometer to put inside, and it actually reads around 50 F higher than the thermocouples (that's not right, is it?).

So I have a few questions:

What happens when a thermocouple goes bad? Does it just stop reading, or does it get less accurate?

mr.jim, can you explain a little more how they would be replaced? I've got these wires coming out the front of the brick, that I imagine were placed during construction, and are snaking through inches or feet of brick. Some notes from a previous baker seem to imply that at least one might be in the back of the oven. Are you saying that I should be able to pull this out, replace it, and feed it back in?

If I replace the outside wiring, what kind of wire should I use? Right now it is a thick single strand that seems kind of brittle, and I think I could get a better connection with a thinner, braided wire.

I would suspect loose wire connections or even some oxide build-up on connections. You can try and open connectors, clean and retighten. The same for the rotary switch if you can access it. As for replacement, you need to know what type of thermocouples they are. You can replace the wire running it as close to the original as possible or replace the TC's and the wire at the same time. As stated above you must get the wire made for the type of thermocouple you have, generically matched. The wire itself is not match to that exact thermocouple and does not have to be the same length as long you are not adding tens of feet then any temp difference readings will be less than a degree or so.

[Mypsi]

Originally Posted by wlively

First, therocouples are nice thing to have, but unless you are doing some serious baking or commercial baking they are not really neccessary

Yeah, it's mostly from obsessive curiosity, but we are doing some serious baking, and trying out a variety of different firing schedules.

Originally Posted by wlively

As for replacement, you need to know what type of thermocouples they are. You can replace the wire running it as close to the original as possible or replace the TC's and the wire at the same time. As stated above you must get the wire made for the type of thermocouple you have, generically matched. The wire itself is not match to that exact thermocouple and does not have to be the same length as long you are not adding tens of feet then any temp difference readings will be less than a degree or so.

So let me talk this through and make sure I understand. Thermocouples are calibrated for a specific thickness and length of wire; a thicker wire would lower resistance, giving a higher reading, and a thinner wire would raise resistance, giving lower reading; a longer wire would raise resistance, giving a lower reading, and a shorter wire would lower resistance, giving a higher reading. Is that correct?

So we have five k-type connectors (the little forks coming out of yellow plastic plugs) sticking out of the front of the oven. They plug into the rotary switch, and another length of wires (a couple inches long) with connectors on either end connects that to the digital readout. Most of what I have to fuss with is that small section of wire. The wire currently used is a single strand, kind of thick (1mm?), somewhat inflexible, brittle wire. I opened up the connectors to make the connections a little tighter, but it's not a very easy wire to work with. They broke a few times when I made the loops, and it's a little too thick to work well with such small screws. Since it's such a small length of wire in respect to the grand scheme, would it cause significant problems to replace that with a multi-strand braided wire, that would be more flexible, and wrap around the screws more easily?

[wlively]

Originally Posted by Mypsi

So let me talk this through and make sure I understand. Thermocouples are calibrated for a specific thickness and length of wire; a thicker wire would lower resistance, giving a higher reading, and a thinner wire would raise resistance, giving lower reading; a longer wire would raise resistance, giving a lower reading, and a shorter wire would lower resistance, giving a higher reading. Is that correct?

So we have five k-type connectors (the little forks coming out of yellow plastic plugs) sticking out of the front of the oven. They plug into the rotary switch, and another length of wires (a couple inches long) with connectors on either end connects that to the digital readout. Most of what I have to fuss with is that small section of wire. The wire currently used is a single strand, kind of thick (1mm?), somewhat inflexible, brittle wire. I opened up the connectors to make the connections a little tighter, but it's not a very easy wire to work with. They broke a few times when I made the loops, and it's a little too thick to work well with such small screws. Since it's such a small length of wire in respect to the grand scheme, would it cause significant problems to replace that with a multi-strand braided wire, that would be more flexible, and wrap around the screws more easily?

Not quite. The disimilar metal juction of the thermocouple outputs a very small (millivolt) voltage in proportion to temperature. The meter is simply reading that voltage and converting it to temp based on a table for that type of TC. The material make-up of the wire is specific to the type (K or R, ect) to not interfere with that millivolt signal. There is no real specific length to the wire. In a calibrated system a calibrated input is used (in place of the TC) and the meter is adjusted to offset any errors created by wire length. But we are talking about less than a degree here, probably alot less in the relatively short runs around an oven. And oven use it is not that "exact" temp critical an application, you will not notice a degree or two. You are correct about longer wire more resistance.

I understand what you are talking about now. TC wire is usually a solid conductor (single strand) and it sounds like it is getting very old and has some breaks in it. I would replace your single pair of junction wires, is a simple and easy thing to do. Use any K-type wire you want ( I would use the simple plastic covered 2 wire) and polarity is important, red is negative (-), have no idea why. You could also just skip the rotary knob and plug directly to the meter.

[n2iko]

My hands are ruined today so here is a link to a nice summary:

Thermocouple - Wikipedia, the free encyclopedia

I've ordered leads to ten feet with only a few dollars difference in price. But that was some time ago.

--mr.jim