PID controller accuracy?

D

Dana Hackney

Joined
Nov 1, 2005
Messages
174
Hi folks!

It's been a while since my last question!

I'd like to ask your opinon on something. I've got one of
those PID controllers setup on my small paragon kiln ($200 :)). The kiln was basically on/off at 1600F, so figured I could use it hooked up to the controller for termpering too.

I can set the controller at 177C and it heats up and stays right at the set temp...BUT...
if I put a couple of oven thermometers in the kiln they register close to 420F.
177c is supposed to be pretty close to 350F (for the 5160 blades I want to temper).
Which would you trust: the thermometers or the controller?

Many thanks!
Dana
 
Can you get access to a handheld pyrometer? I'd trust that over everything else. Were the thermometers close to the elements and the thermocouple farther away? That might offer some explanation... The thermocouple might be better at accurately reading high temps, but I'd have trouble swallowing that. Only way to know for sure it put a pyrometer in there, I think.
 
Might have something to do with the placement of the probes - similar how on a sunny day one feels so much cooler in the shade, while 2 feet away, in the sun it is hot, yet the air temp is the same .
 
Might have something to do with the placement of the probes - similar how on a sunny day one feels so much cooler in the shade, while 2 feet away, in the sun it is hot, yet the air temp is the same .
Click to expand...
420 F = 215.5 C; 177 C = 350.6 F

I agree, the placement of the senors is important. Oven thermometers are probably good to only +/- 10 degrees, so accuracy is an issue too. When using a PID, there may be differences in observed temperature around the oven. This depends on where the PID's temperature is taken -- that temperature will be really close to the setpoint because it is being controlled to be so. Also, the air in the oven must be circulated well to even the temperature.
 
I would also think that the thermocouple itself may be sending a slightly off reading to the pid. Thermocouples work by generating a bit of electrical current due to degradation from the heat. I find it hard to believe they are all perfect.
 
I would also think that the thermocouple itself may be sending a slightly off reading to the pid. Thermocouples work by generating a bit of electrical current due to degradation from the heat. I find it hard to believe they are all perfect.
Click to expand...

Very true. There are many types of thermocouples, and some are better suited to high temperatures. Also, the PID probaby may be setup to use only couple of types (sometimes only one). You would need to check that the correct thermocouple is being used.
 
I used to install temperature controllers as part of my job. Most specs for the ones that I have used show them to be accurate within a few degrees. To speak to the accuracy of the thermocouples, +/- 2c is common. Some are more, some less accurate.

As far as what you are experiencing, it is possible that your controller was initially set up with an offset in it, or that it was set for use with a different type of TC than what is currently in use. If you set up a controller to use a K type, but use a J type, you will see similar offsets to what you have. Other potential issues are electrical "noise", and damage to the thermocouple.

When trouble shooting controllers the first step we used was a digital calibrator. It sends a low voltage signal that is equivalent to the voltage that would come through the thermocouple for a given temperature; displayed temperature should match what the calibrator is set for. You do this for multiple points throughout the intended range of use for the application. Hopefully the temps match, or it they are off, they are consistent in their offset. This can compensated for in most controllers.

The others are correct in saying that it is very important that you are taking the temperature in the same place as where the ovens thermocouple is located, as temperatures can vary widely throughout the chamber (but shouldn't be as far off as what you are seeing).

Don't know if any of this helped. Im not awake yet.
 
It has been my experience that most small heat treat ovens with a single type K thermocouple give inaccurate readings at lower temps (below 1000 F).
I have an old Hupert retrofitted with a new PID controller and a new TC that is 30 degrees off at 500 and dead on at 1800-2000.

Jim
 
this is all helpful stuff.

I went down to the local H.F. store and bought one of their el -cheapo hand-held
digital temp reader that comes with a k-type thermocouple. I set the kiln's PID controller
to 177 and then let it cyle for a while and then stuffed the theromcouple of the hand-held
unit in the kiln. It was fairly close to what the PID was reading but the two oven thermometers
got fried and both were reading past their max of 500F when they died!

I asked the guy that sold me the PID unit about this and he said, "It's just matter of tuning
or the location where you set the thermocouple to do sampling.Try to play around with the
PID controller. It should set to K type. If it doesn't work right, try to set to J type just in case
the thermocouple is J instead of K."

So, it looks like what you guys are saying meshes with the seller. NOw I gotta find out
how to test the J vs K setting :grumpy:

I dropped back for now and am using the house hold oven. I used the hand-held
unit to see what temp setting I needed for the oven that gets me the closest
to the 350f mark. Turns out abut 360F for the oven is close, sort of. At lest that is
what the digital thermo dealy-bob reads.

Many thanks!
Dana
 
If you are unsure as to the probe's type, it will certainly be a problem.
Only K-type is fit for the temp one expects in heat treating ovens.

For some steels you will be pushing real close to 2000F.

You can figure out what type you have by watching the voltage coming
out from your TC, by borrowing a good digital meter from the same store.

Google for tables that have K-type temp-to-voltage values, use your
other probe (from HF) to tell you what actual temp is.
 
Look at the wire colors for the thermocouple in your oven. If they follow the US and Canadian coding methodology they will be yellow and red for K type: white and red for J type. If they are Japanese, both types are red and white. Other areas use different codes ~ these are the only ones that I am personally familiar with. I am sure that you can look at Omega or one of the other sites for more details. You can measure the voltage of the TC as said, however, if it has damage that happens to increase resistance in one of the wires (or the junction) you will get haywire voltages.

Again, hope this helps.

Steve
 
I bought the controller and it was supposed to come with a K-type thermocouple.
I got it from the same place that Ed Caffrey purchased his - Cold Fusion off e-bay.

I worked pretty well when it was hooked up to a simple coffee pot heating water.
I sould bring it just to boiling point and was able to hold just under that mark without
a hitch. The trouble started when I hooked it up to the small Paragon kiln. The kiln
has no control for temp. It's 1600f or off.

Thanks guys,
Dana
 
You must log in or register to reply here.
Back
Top