Does your PID controller lie to you?

[el34]

Not long ago I added a Mypin PID, same kind Frozone and others use. In 650deg lead it reads about 40deg higher than 2 reference thermometers. Fortunately you can program in a 'correction factor' and I've done that, but it still bugs me.

I know a bit about thermocouples and that they need a reference point of 32deg. Modern electronics fakes out this reference by taking the ambient temp and adjusting the TC output.

My TC plugs into a mating socket on my controller box and I used TC wire, observing polarity, to hook up the socket to the PID.

Anybody run into anything like this?

[blikseme300]

I treat the readout on my PID as just numbers. I record what number works best with which alloy and mold and set to that prior to casting. The actual temperature is of no concern to me, just good boolits are my concern.

[btroj]

That is what I would do. If there is an offset does it matter?

If an alloy casts well at setting 104cxs then as long as you are consistently at that setting does it matter what temp it really is?

I don't have a PID but have seriously considered one. I want it to keep pot temp steady, not to keep it at a predetermined temp.

[el34]

I completely agree- consistency and good boolits is 99% of it. But my operating system wants to know the other 1% stuff, kind of like an engineering viewpoint.

The controllers aren't designed to be loosy-goosy and I can't help but want to know why mine is 40deg off. I have another one but haven't yet hooked it up for comparison.

I'm working on a simple method to measure mold temp for every cast, based on the gut feeling that lead temp and mold temp are equal partners in consistently creating close to perfect boolits. Seems like hot lead can compensate for a cold mold etc and I'd like to monitor and regulate mold temp too.

[Springfield]

Are you using an in the lead TC that goes where the regular thermometers go, or a TC that attaches the the outside of the pot via a welded on nut?

[el34]

My thermocouple is a stainless probe, 4" in length. It is mounted to the pot and the probe is in the lead.

I see the error just in air, where the TC is laying outside the pot right next to a cooking thermometer. The PID reads higher than the thermometer. I have another TC but for some stupid reason haven't tried it yet.

My controller has a socket connector for the TC to plug into. There is a potential error caused by connectors or final wiring not being made of the same (weird) metals the TC is made of and I did the socket to PID connections using excess TC wire, a tip from Frozone I saw somewhere.

[KohlerK91]

My PID was suppose to use a K type TC and had a J type in the pot. The temp was off like you say. I got a the correct K type and it more closely matches my thermometer.

[snuffy]

Well, I just looked at the inside air temp here in the house. A very old mercury thermometer said it's 74 DEG F. The AC thermostat says it's 73. I just turned on my PID, it's TC is in solid room-temp lead, it read 70 degrees. An error? Or is the loading room cooler? Yes it is, the back of the house is always cooler.

So what? I'll trust the PID readout before the spring type bi-metal type of casting thermometer. IIRC, a couple of times I put my Lyman thermometer in the same pot as my PID, they were 20 degrees apart, the Lyman was 20D higher.

As said, it matters not if you have good luck with a certain alloy @ 710 degrees with a certain mold, write it down, you can repeat it at a latter date.

The best part of the PID is; it maintains temp +- 2 degrees! Compared to the lee pro 20's huge temp swings.

[el34]

My PID was suppose to use a K type TC and had a J type in the pot. The temp was off like you say. I got a the correct K type and it more closely matches my thermometer.

Yeah, I dbl checked that. Thanks. My PID and the 2-3 others I'm familiar with will accomodate several TC types as well as other temp sensors. It's possible yours would work with a J type if you set it up that way, dunno.

How closely does it match your thermometer?

[el34]

As said, it matters not if you have good luck with a certain alloy @ 710 degrees with a certain mold, write it down, you can repeat it at a latter date.

I don't care if it's up to 10deg off, but 40 deg error makes me think something is wrong. I understand the thing about the number is just a number but I'd still like to know the real temp for comparing to stuff I read. I have that because I forced my PID to match a couple of thermometers but it still bugs me. I'll give it a rest and move on probably.

[Mal Paso]

My Auber PID is spot on from room temp to 750F but I had a problem at first. I had silver soldered the TC in the bottom of the pot but the "Junction" was too close to the outside of the pot and it was reading about 40 degrees low. Auber called it "thermal drain". I moved the probe into the lead 1/2" and that solved the problem. Could the hardware holding your probe be stealing heat from it?

[el34]

Could the hardware holding your probe be stealing heat from it?

Thanks Mal. It is supported at the top, the bead at the bottom is about 1/2 in above the bottom and 1/2 in away from the side.

I had about a 20deg error compared to a thermometer just in free air, 75deg or so. It's like a puzzle I want to solve. Well, someday.

[kitsap]

el34,

No, never have had that much of a differential. I have built three using the Auber controller and Auber thermocouples. As part of the building process I would check the temperature with crushed ice and alcohol (32 degrees) and then boiling water (distilled water and I live at sea level for 212 degrees). I do not have access to a reference thermometer in the range of lead melt. At the most the offset was in the range of 3 to 4 degrees. When compared in an alloy melt against two different RCBS lead thermometers it was within reading range.

In two of the controllers I used the plug/receptacle kit from Aubers and used a short piece of the thermocouple wire cut off and stripped.

While I concur the readings are a reference and not absolute measurement, it would be worthwhile to know why the differential. I recall the controllers are thermally compensated for ambient temperature. Maybe something wrong there.

DougF

[el34]

I recall the controllers are thermally compensated for ambient temperature. Maybe something wrong there.

Yep to internally compensated. With the TC laying on the counter at room temp, next to a thermometer, they both read the same when power is initially turned on. The PID starts to drift upward over the next several minutes making me think like you that as the controller generates its own internal heat the compensation isn't doing its job. I have another controller and another TC, all I gotta do is swap them around. Heck maybe I better check that the controller is still told it has a K type TC.

I also used TC wire to connect the socket to the controller. Thanks for all your comments.

[Dale in Louisiana]

You might want to check around if any of your circle of acquaintances works in industrial instrumentation. If so, they typically have access to test equipment that will simulate a thermocouple input to your controller, as well as read your thermocouple. Both these operations are pretty standard things in industry.

Not much can go wrong with a thermocouple itself. Problems usually arise at connections and within the device that reads the thermocouple. Since the advent of equipment that can be programmed to accept various different thermocouple types, a fairly common problem is to find that the unit is programmed for the wrong TC.

dale in Louisiana

[500MAG]

I've been lucky. I have 2 of the same ones that I built and once I auto tune them they are accurate to within a couple degrees.

[NineInchNails]

el34 I quickly scanned through this topic, but I didn't see anyone mentioned this ... Does your thermocouple plug into a panel mount connector OR is it wired directly to your PID?

Panel Mount Connector:


I noticed that you are aware about proper polarity, but IF you are using a panel mount connector you have to use thermocouple wire to connect the panel mount to the PID. Polarity is also important with this wire as well.

[el34]

Does your thermocouple plug into a panel mount connector OR is it wired directly to your PID?

Thanks for asking 9inch. My TC came has the standard small connector on the end, and it also came with an banana plug adapter for connecting it to a multimeter. The adapter was yellow, the official color for type K, but who knows. I glued that adapter through a hole in the rear panel of my enclosure. Initially I wired the inside (banana plugs) to the PID with twisted wire but replaced that with some TC wire leftover, about 12", from shortening the original cable and matched the wire colors.

Next weekend I plan to open up everything to make sure I did what I said I did, and to maybe hook up another PID, same type.

[kitsap]

el34,

As you work through this issue I recommend you focus on that banana plug adapter as part of your problem and get it completely out of the circuit. The two terminals on the back of the PID controller should be used to connect only thermocouple wire. Your banana plug adapter was designed to plug into a multimeter and not a PID controller.

Many thermocouples come with terminal ends crimped onto the end of the wire and that is OK to connect to the PID terminals.

DougF

[el34]

el34,

As you work through this issue I recommend you focus on that banana plug adapter as part of your problem and get it completely out of the circuit. DougF

I'm with you Doug, that banana plug adapter is the only non-TC part in the circuit. I still want to have a disconnect on my rear panel (http://castboolits.gunloads.com/show...D-4-20-project) so I'll get a panel mount connector designed for this.

Thanks for pointing that out. I got used to using the programmable fudge factor in the controller but if I can make it right, well, you know.