PID thermocouple reading is inconsistent

[CGT80]

I ordered the parts from Auber to build a PID. I used the 2352 PID, a 25 amp SSR, and the WRNK-191 thermocouple for heat treating, in 6" length, and a small box for the PID.

I managed to fit the SSR in the small box, where others mounted it to the back. I had to leave off the clear plastic cover, and I used electrical tape to cover the contacts just in case something moved or came loose. It is screwed to the back of the box and has some thermal grease on the back. There is just enough room for the electrical connections. I ran my cords through rubber grommets on the side of the box.

The PID appears to work correctly and I have the polarity correct on the thermocouple. Yep.......I had it backwards when I first hooked it up.

I used ice water to test the accuracy of the thermocouple and I also tried boiling water in my 10 pound dripomatic, after cleaning it out real well. It actually held water, after I thoroughly cleaned the spout and rod.

The TC read correct air temp at first. I put the probe in the pot and got a reading, then put it back into ice water. I have played around with it quite a bit. When I first got temps too high or low, I adjusted the input offset by up to + or - 20 degrees. I was able to set it to 32 degrees when touching the ice in the ice water. After reading room temp or the boiling water in the pot, it would be off again. At one point it showed 220 degrees and the water wasn't even boiling yet. Tonight it showed 59 degrees when I help the probe in my hand. Other times it showed in the mid 90's.

I am wondering if the TC is damaged or if a connection is not good. I used the spade connectors that came on the TC to connect directly to the set screws on the PID. I notched the metal box where the two halves join, to allow the braided jacket of the TC to pass through. I didn't bother with a grommet for the TC cable.

The aluminum box was damaged when I got it via USPS flat rate box from Auber. One of the tabs on the rear was bent up toward the back of the case. The cardboard box was wrinkled up a bit too. I straightened that metal tab and used the box anyway. Maybe the TC was damaged.

I was just about to order the big green egg thermometer that has been mentioned here, but I hate to spend 30 bucks on that when I have a PID.

Chris

69684

69685

[CGT80]

I just ordered a big green egg thermometer so I will have a reference for my PID. I might call Auber to see what they can say about my temp issues.

[Lee W]

Try this.
Make sure you controller input is set to type K, then set the offset to 0. Place a copper wire across the input terminals.
This should read room temperature, or as close as the controller itself is. If this it OK, the probe has a problem.

[CGT80]

Thanks, Lee.

I tried the copper wire and set the input to 0. I had checked before that it was set to K type. This test showed my actual room temp. I moved the unit outside and it lined up exactly with the readings from two weather stations that I have. One station said 60 degrees F and the other 61. The PID bounced back and forth between 60 and 61.

I tried the thermocouple in the outside air, inside air, under a halogen lamp, and held in my hands. Sometimes it showed higher than I expected (also higher when TC was outside) and other times it showed lower than expected. I wonder if it could have a bad connection or just a manufacturing defect. I wiggled around the wires all along the TC and it didn't make a difference. It didn't seem to matter if the braided jacket of the TC was touching the box for the PID or any other surface. I did these tests with the box open and the tests the other night were with the box closed and mounted near my lead pot.

After messing with the TC, I tried the copper wire again. It showed 65 degrees from being inside my loading room. I moved the whole unit outside and the temp slowly dropped to 60 F which is the current temp.

It appears that the PID is working, but the TC is either bad or just fluctuates and is inaccurate at low temps.

I ordered a casting thermometer as mentioned before, so I have to wait for that. I will try to call Aubur now that I have more info on my side.

I was excited to cast some nice booits with the stash of lead I just picked up. Needless to say, I'm not too thrilled with Aubur as of right now. A damaged project box and now a bad TC. I wonder if the two are related. How fragile are TC's? How accurate should my TC be at low temps?

PS. The red wire with red terminal ended up being negative on my TC. White wire with blue terminal is positive. I used terminals 4 & 5 on the PID. Polarity must be correct since a heat source causes the temperature to go up.

I did not try grounding the jacket of the TC to pin 5 as stated in the instructions for the PID. Has anyone else had to ground the braided sheath of the TC?

Thanks,
Chris

[bgoff_ak]

It may be that the TC is working fine with in its error tolerance. I can’t find any technical data on that TC , neither mouser or the likes has datasheet for this item. 30° over the scope of +1000° seems well with in the spec. I will say i had to get a diffrent TC for our egg incubator so it would work. same setup diffrent appliation.

[Frozone]

A K type thermocouple is inherently inaccurate.
It has a rather wide average margin for error +- 2º IIRC.
That and 'cheap' PIDs are only good to around .3% of the thermocouple range of ~ 2200º or ~ 7º of error.

A K type that holds with in 3º of 0 - 100 C is doing pretty good.
The other issue with The K type is, it is Most inaccurate at the temps we use to cast.

Why do we use them if they are so bad, you might wonder.
Well, 1. they are cheap and available, 2. they don't rust, and 3. they work in the range we need.
J type is a more accurate choice for our temps but it has to be shielded - it's iron and rusts.
B,R, and S types would work except they are made from rare metals (platinum) and are rather expensive.


The best you can do is what Lee recommends: replace the faulty TC and set the PID for Ambient temp.

The other problem you're seeing is 'drift'. As that enclosure doesn't vent, the heat builds up inside.
In other words what the controller 'sees' as ambient is higher than actual.
While it may not cause failure it is effecting the function.
Look at the terminal end of that Auber unit, see the small glass package 'diode' in one of the open holes?
That is the cold junction compensator. That is what reads ambient temp.

<edit> Is your PID sitting on it's side?? Do you mount he box on the wall?
There is a reason there are vent slots in the top and bottom of the PID unit you know </edit>

[CGT80]

I made some progress, but it put me back about 5 steps.

I talked to a guy at Aubur. He said to connect the red connector (and red wire) on the TC to the #4 pin on the PID which is positive. He also said to make sure the probe of the TC is completely covered by the ice water. This gave me good readings. I checked hot water and got the expected temps. When switching between different containers of water (with different temps) the temp reading can sometimes go the wrong way for a couple seconds before heading the correct direction and stabilizing at the new temp. It seems I had the polarity backwards.

The bad part: I can only get accurate readings with the 6" probe if the upper portion of the probe is in contact with the material to be tested. Putting just the tip in melted lead will cause an error on the PID which states the TC is bad. It also will cause the temps displayed to be opposite of what they should be. In this case the melted lead read out at -126 degrees F. My lee 10 lbs pot is not deep enough to immerse the probe up to the top, to test this theory. This also means that any level of lead below the top of the TC probe would not give accurate results. This seems to not have been a problem for other casters here who used the same TC and PID. What gives?

I'm going to find the thread of someone else who used the same parts and re-read their posts for the 10th time. I may PM them as well.

At this point the PID appears to be good and the TC works if completely submerged, but this will not work for my application, unless I mount the probe through the bottom of the pot and let the probe stick out of the melt.

Auber made a good attempt at being helpful, but the lack of any documentation with the TC really sucks, and communicating over the phone can be rough due to the guy's foreign accent. The instructions for the PID sounded a bit foreign as well, but so far I didn't find any major problems. It just makes it a bit hard to understand sometimes.

So far this whole PID setup is a major pain. It figures that others would have smooth sailing and I would end up with the problem child. It isn't even like I can say that the TC is faulty so far. To say that I may have misunderstood the limitations of the TC might be more accurate.

Who uses a TC that will read properly just from the tip? I may have to silver solder a short TC to the bottom of my pot, although I would like to move up to a lee 20 pound pot when they become available again.

[CGT80]

<edit> Is your PID sitting on it's side?? Do you mount he box on the wall?
There is a reason there are vent slots in the top and bottom of the PID unit you know </edit>

I mounted the PID in the "cbox16" aluminum extruded box from Aubur, just as many other people here did. The box is screwed to the wall and the PID is right side up. Actually the bracket on the back of the PID would not have let me turn the PID 90 degrees in the box, without modifying the bracket anyway. This box doesn't have vent holes but it does have ribs or slots to act as a heat sink. The SSR is in the box and lee pot is 500 watts IIRC. The unit hardly even gets warm to the touch. I could see this effecting accuracy, but I can't even get a temp reading within a few hundred degrees of my melt, so it looks like it is the wrong thermocouple for this application.

[Frozone]

Don't underestimate how much heat those PID units generate.

I have a aluminum cased unit out in the garage. In winter it gets well below freezing.
Since the Auber units won't power up below 32º F, I added an internal heater with a cutoff switch.
The internal heater would warm things to ~45º than cut off. (bimetal thermostat 5ºC)

After I switched to Mypin units (which don't care what the ambient temp is) I now longer needed the heater but I tried to turn it on after the unit had been on a while.

With the ambient temp at ~ 20º, the heater wouldn't come on - the internal temp of the case was at least 20º warmer than when I turned it on!
BTW, My case has ventilation slots in it too.

I have thermocouples that work perfectly for this app @ $10 if you need a new one (and you likely do).

[CGT80]

Thanks for the info Frozone. I will contact Aubern again tomorrow to see about replacing my TC and another member was kind enough to offer to help with a sensor also. Between the two, I hope to get up and running. If not, then I will contact you for a new TC.

I will also keep an eye on the temp. I may have to move the SSR to the outside of the case and vent the case as well. My bbq thermometer is in route, so I will have another high temp thermometer to compare my PID to.

Thanks to everyone who has helped out so far. There are some outstanding people on this forum.

Chris

[Cane_man]

another way to check your temp probe is to see what it reads as the lead is melting... should read 621... mine reads 619, and in boiling water it reads 211, so my TC reads a little low but no biggie

[CGT80]

Auber is sending me another probe and a shipping label to return the bad one. I will try the new one when it shows up. I will use ice water and boiling water for baselines. This probe would return to withing a few degrees of 32f when put back in the ice water, but only if the entire probe was covered with water. A few degrees off will still be far better than the Lee temperature control, but the fact that it won't read from just a portion of the probe is a deal breaker.

Water actually boils just over 210 degrees where I am at (1237' above sea level).

Thanks for the suggestions.