Another PID question

[bosterr]

I built my PID a couple years ago, before Frozone did the group buy, or I would have bought one from him and maybe wouldn't have this situation. Anyway, it sits in my cold basement, around 56 degrees, and the PID says it's 32 degrees. When immersed in ice water (mostly ice) it reads 12 degrees instead of near 32 degrees. Can I assume that it's off by 20 to 24 degrees? When I run my Lyman 20 pound pot, I adjust the set temp by 20 to 25 degrees. Can these things be adjusted to show correct ambient temperature? The controller is a SYL 2352 I bought from Auberins. I'm VERY electronically challenged and re-read the instructions twice and didn't see anything about this. I know it dosen't sound like a big deal, it still works fine while casting, but it kind of bugs me. Any suggestions?

[Spawn-Inc]

straight from the manual,

"4.10 Input offset “Pb”Pb is used to set an input offset to compensate the error produced by the sensor or input signal itself. For example, if the controller displays 5ºC when probe is in ice/water mixture, setting Pb=-5, will make the controller display 0ºC."

assuming you know how to get into the settings menu, simple set the number (default should be 0) to what ever you think the error is. so i would set it to 20. then verify with a bowl of ice.

hope that helps.

[felix]

It's is always best to calibrate at a known temperature that is repeatable in the area of interest, like when the lead pot is at the most average temperature when casting. Maybe some certified linotype, for example, when the known lead temperature is fully liquid, and one degree lower the lead is fully solid. Not much off-the-shelf consumer grade stuff is absolutely linear and as such makes 32 degrees too far off from a valid calibration point. ... felix

[bosterr]

Okay, Spawn, I adjusted the Pb and moved it 20.5 and got it to read 32-33 degrees in a bowl of ice water (mostly ice). Then I boiled a cup of water. I know water boils at 212 degree at sea level, where I'm not (the highest it would go is 197). Felix, I agree with what your saying that ice water is too far from actual temp I need to monitor. When we calibrated spring gauges at the gas company, for example, they were calibrated to read about 3/4 of the capacity of the gauge. I don't have any certified linotype to check. I wonder how Frozone calibrates the ones he sells.

Thanks so much guys, I think I'm closer that I was before.

[kitsap]

bosterr,

A couple of other things can introduce an error in temperature readings. One is the setting in the PID controller for the type of thermocouple you are using. Most likely you are using a type K thermocouple. Verify the setting in the PID controller for the Sn parameter is 0.

The other is the method of electrically connecting the thermocouple to the PID controller. The controller is designed to connect directly to the thermocouple leads. If you cut the leads and used some type of switch or connector between the controller and the thermocouple you could have introduced an error. In particular if the switch or connector was not specifically designed for use with thermocouple leads.

DougF

[theperfessor]

To follow up on kitsap's comments: When I made mine Frozone was very helpful and suggested that I cut off some of the thermocouple wire and use it to wire from the plug-in jack to the PID terminals. This insures that you don't introduce an error by having a different composition wire in the circuit.

I checked mine using some relatively pure lead; it reads out within just a couple degrees of being the right temperature (651F) as the lead melts.

[Frozone]

Start by removing the input as a source of the error.
At the PID unit, remove the TC leads and use a short piece of copper wire to jumper the input terminals together.
The Unit should now read nothing but ambient temperature.

Do you still have an error and if so how much?

The PID unit itself has a input error of 0.3% of range.
For a K type that is ~2500º about +-7.5º of acceptable error in the unit itself.
However, in practice the error is rarely more than half that.

The K type TC itself has an error also. This is due to small impurities in the metal use to make it
Depending on temperature, that error is a few degrees (max of about 4º)

Irregularities in forming the 'normalized junction' that is the sensor itself can also cause problems.
(I'm suffering from that with my TCs from 'TOKO', I'm getting a 8% failure rate and it can't be found without bring the TC up to operating temp.)


As theperfessor mentions, how you wire the TC up matters.
One thing that can cause rather large errors is to reverse the wire types from the jack/plug connection to the unit.
Not using TC wire for that connection is a big error source as well.

FYI
You Can get away with a screw type connector that is not for TCs.
This is because the screw, at the 2 points it contacts the wires and there for the metal to metal junctions, is at the same temperature.

[bosterr]

Hi Guys, I'm back.

When I built my PID I opted for simple. By that I mean I didn't use any type of plug, on/off switches, alarms. All connections were made with 1/4 inch 2 prong spade connectors. The TC I used is a 1100 or 1200 degree that had these same spade connectors installed, which is what I used. In other words, everything is wired directly to the controller, the power cords, both in and out. Like my first post said, I tried adjusting to a bowl of ice water, maybe not the correct thing to do. Then I adjusted to ambient temp in my basement, and finished casting some 400 gr. 50's a little while ago. Last night, though, I re-read post #238 that Frozone wrote in "Casting Equipment", under "Project PID on Lee Pro 4-20 furnace. In it he gave all the correct settings, and I adjusted everything to that. Like I said, I have no kind of alarms, lights, or buzzers, but I set my controller that way anyway.

Now, I just checked two thermometers, 1 is a glass tube/ mercury and the other is a pretty good round dial/ spring thermometer. They're within 2 degrees of each other at 57 and 59 degrees. I then used a 14 ga. 2 inch long piece of copper wire and jumped across the TC terminals on the controller and got 73 degrees. Are you saying that the correct setting is 73 degrees? Also, help my pee brain understand, is 73 degrees a pre-programmed setting that is used to calibrate the TC? Since the TC isn't connected to read ambient temp, how does it know it's 73 degrees?

BTW, I just finished casting 400 gr. 50 cal. Truncated Cone boolits, started the controller at 800 degrees, then slowly worked down to 760 degrees. This is a double cavity magnum sized aluminum mold from Mountain Molds and it worked out really good, virtually all were perfect. Never one to be completely satisfied, "good enough" is not my philosify. I would really like to get it set up the way it's supposed to be. Thank you all for your time and effort to help me, it's greatly appreciated!!!

[Frozone]

No, when you jumper the terminals #5 and #6 together, the unit reads just like a thermometer.
The upper display is what the unit thinks the ambient temp is; in this case 73º.

If you are sure the room temp is 58º and that the unit is at that temp, then the "cold junction compensation" circuit is bad.
Holding the unit in your hand can warm it up, be sure it has had a chance to cool back to room temp before testing.

The unit 'knows' what the ambient temp is by the reverse leakage current across a diode.
If you look at the end of your Auber unit, you will see 2 open slots (no terminal screws).
One of those 'holes' should have a small diode in it. That is the CJC diode and that is what reads ambient temp.

[bosterr]

Now both of my thermometers agree at 57 degrees, so now we know what actual ambient temp is. The control unit is still wired up inside the enclosure, so the best I can do is tip it up part way to put my hand around it. It will go up to 75 from the previous 73 by holding it for about 2 minutes. Like I said before, I had it running my Lyman pot earlier today for about 2 1/2 hours, but it all feels cold to the touch now. Maybe I should let it sit overnight and try that again tomorrow. I think I see the diode you describe as a tiny orange thing with what looks like black stripes. If the cold junction circuit is bad, does that mean the controller is no good or can the Pb setting be adjusted to compensate?

[kitsap]

bosterr,

Do you still have the Pb offset programmed into your controller? If you are using the short wire and comparing to ambient you should have Pb set to zero.

DougF

[bosterr]

Yep, the Pb still is offset, I'll put it back to zero and see what I get.

[bosterr]

The thermometers still say actual ambient in my basement is 57 degrees. Backed up the Pb setting from 40 to zero. Jumped the TC terminals and get 33. Are you saying that if it was working correctly, it should now read 57?

[Frozone]

YES

Adjust Pb to compensate for the offset (+24), then hook the TC back up and check the reading.

I don't like to adjust the offset by that much, it's out of spec by a bunch, but at least it should get things better.

[bosterr]

Hi guys,

It may be time to admit defeat. Dialed in 24, reconnected the TC and ended up with 75. Then while holding the TC in my hand, the display temp was actually decreasing. So I reset "PB" with 30 now instead of 40 like before to get my 57 degrees ambient. There's no doubt this controller has issues. I guess if I can dial the correct ambient temp into it, I'll continue to use it that way. Still bugs me that it's not right. Thanks so much for all your help! Bob

[Spawn-Inc]

if the temperature is decreasing, you may have it wired backwords/wrong polarity.

[bosterr]

I didn't think there was polarity on the thermocouple. Anyway, at least it's working the way it was.

[Frozone]

If you hold a thermometer that is reading 50 in your hand and it goes down you think it's right???

Thermocouple have polarity in the wires!!!!

Lets try again.

Disconnect the TC, jumper the TC terminal.
Adjust Pb for the correct ambient reading.
Pull the jumper off and reconnect TC.
Check that displayed temp RISES when you hold the TC tip.
If not reverse the TC wires.

It may still not be perfect, don't worry if it's < 4º off!!
The TC can have that much error over it's range.

[bosterr]

I'm to jump across both TC terminals, is that correct? With the jumper attached in that way, when I adjusted Pb to 57, when it returned to the regular readout, it said 116 degrees. Then I tried again, only this time, I had to adjust Pb to -4 and when the display returns, its' right on the target temp of 57. But when I re-attached the TC, it now reads 22. Yes, now when I hold the TC in my hand temp rises.

[Frozone]

Oh for god's sake.

Just set Pb to "24". we already established what the offset was.