Thermocouple Placement with PID

robs · August 12, 2006, 08:58:35 PM

1. Where do you place it?
2. What brand?
3. How much swing?
4. Are you using an offset on input?
5. Type of PID (just curious)
Thanks

I'm having a hard time getting mine setup properly. It seems like the only placement where I don't get a swing is 2" above the heating element. This won't work though for obvious reasons. 1. The TC will likely fry out over time. 2. Too much of an offset is needed. On my PID, the maximum negative offset is -199.99 - not enough.

So I tried auto-tuning with the TC in several different places. It seems that no matter where I place it, I get under and overshoots of 10 deg. f or so.

I contacted the manufacture and they suggest that the TC is as close to the heat source as possible. Like I said, the swing is gone when very close to the heater, but it's not feasible.

I've read on the forums that Auber TC my be subject. This is what I have. Is this the problem? Even if this is the problem, I'd still like to know where everyone is mounting the TC.

My setup:
1. Almost everywhere
2. Auber TC
3. 7-12 in both directions
4. Lord I've tried
5. Auber PID

Thanks everyone!
rob

Arcs_n_Sparks · August 13, 2006, 07:54:18 AM

robs,

Cannot help you with the Auber setup. However, at these temperatures, you will not fry the TC. Not sure what type of TC you are using (J, K, etc...), but they are hardy devices.

Arcs_n_Sparks

nodak · August 13, 2006, 08:22:42 AM

Robs, it's your Thermocoupler, Buy an Omega (about $30 and It's worth it) the one you bought with ypur PID is probably about $7.00 and you got what you paid for. I went through the same thing until Bubbagump advised me.

TomG · August 13, 2006, 10:06:02 AM

Or you can buy 2 for $25 inc. S&H.

http://cgi.ebay.com/Thermocouple-2-K-3-16-x6-w-8ft-lead_W0QQitemZ290018058698QQihZ019QQcategoryZ50926QQrdZ1QQcmdZViewItem

http://cgi.ebay.com/Thermocouple-2-J-3-16-x6-w-12ft-lead-cf_W0QQitemZ290018060472QQihZ019QQcategoryZ50926QQrdZ1QQcmdZViewItem

Rob are you giving AT enough time to find the set. My units, neither an Auber, will swing 5-10* for 10-15 mins. before hitting and staying on set. :)

TomG · August 13, 2006, 10:20:49 AM

BTW Rob, I just drop my TC down through the vent so that its tip is close to cabinet center, obviously avoiding any thing that's on the racks. AnS as always is correct, the temp ranges on a J type TC is 32-1382* F and -328-2282*F for K type and its unlikely that the BS elements will reach those temps. I'm almost certain the Aubin cheapy is either a J or K.

nodak · August 13, 2006, 11:25:29 AM

http://www.omega.com/config/probeconfig.html

Thermocouple - Omega #TJ72-CASS-316U-3-SB-SMPW-M(can modify to cord length of your choice by going to link http://www.omega.com/config/probeconfig.html,
these are custom built)

Mini Panel Jack - Omega #RMJ-K-R

The wire on TC is heavy duty compared to most. Order an extra 6 inches on your TC if you want to do the mini panel jack so you have the right wire to add it in.

I ordered the extension cord also for the TC so I can stick the cord under the door and put my PID inside in winter @ 0. Believe minimum ambient operating temp for PID is 32F.

robs · August 13, 2006, 03:52:55 PM

Quote from: TomG on August 13, 2006, 10:06:02 AM
Or you can buy 2 for $25 inc. S&H.

http://cgi.ebay.com/Thermocouple-2-K-3-16-x6-w-8ft-lead_W0QQitemZ290018058698QQihZ019QQcategoryZ50926QQrdZ1QQcmdZViewItem

http://cgi.ebay.com/Thermocouple-2-J-3-16-x6-w-12ft-lead-cf_W0QQitemZ290018060472QQihZ019QQcategoryZ50926QQrdZ1QQcmdZViewItem

Rob are you giving AT enough time to find the set. My units, neither an Auber, will swing 5-10* for 10-15 mins. before hitting and staying on set. :)

This PID starts and finishes the AT on its own. It over and undershoots 3 times and then it's done.

robs · August 13, 2006, 03:54:46 PM

Quote from: nodak on August 13, 2006, 11:25:29 AM
http://www.omega.com/config/probeconfig.html

I ordered the extension cord also for the TC so I can stick the cord under the door and put my PID inside in winter @ 0. Believe minimum ambient operating temp for PID is 32F.

Yup, not sure what I'm going to do about the cold weather here in ND yet...

robs · August 13, 2006, 04:13:26 PM

It took a bit of messing around, but I'm finaly getting aceptable results.

TC pretty much anywhere in box:
I'm not even sure if I'd call it a swing. It maintains a constant temp. Every few minutes, it drops or raises a deg for just a few seconds. This could be from electrical noise?

I reset the values for P, I, and D. And started running tests from there without running AT.

Then I started modifying the I value. The larger the number (to a certain extent), the smaller the swing got. I didn't mess with P and D values. I just kept tweaking I until it stayed constant with damper open full and 1/3.

Now, I hate to go against math and just tweak things until they work. But maybe I'm not? I mean if P, I, and D are all greater than zero, then it's still acting like a PID, right? Maybe just not optimaly. I think that because of other variables (possibley the TC), the AT just wasn't cutting it.

I do need to mention that Auber's has very, very good support. Suyi helped me via email on Saturday and Sunday and offered to troubleshoot with me over the phone.

I'm not going to discount this TC just yet. The whole sollution seems to be working after tweaking the I value manualy. I haven't asked Auber's about this yet, but will update this post later. What would be intersting, is comparing the P, I, and D values from other users using the same PID from Auber. Actualy, from any PID users may be useful to everyone?

rob

TomG · August 13, 2006, 05:46:10 PM

Quote from: robs on August 13, 2006, 03:54:46 PM
Yup, not sure what I'm going to do about the cold weather here in ND yet...

Hell man that's easy, move to California :D :D :D

Arcs_n_Sparks · August 13, 2006, 06:01:56 PM

Let me make a few control system observations.

PID loops (and their parameters) depend upon the time constants of the overall system and their stability. If you took an empty BS, vent half open, and tuned it, you would have one set of parameters. If you took a BS with 15 pounds of butt in it from a cold start, you are going to get different parameters. Tweak the vent, and you affect the heat balance and system response as well. In the overall scheme of things, this probably doesn't matter a whole lot (unlike fly-by-wire aircraft).

You can always tune to have an overdamped system that will not overshoot (lower value of P). In this case, it will take more time to reach setpoint (the I term will integrate out the error eventually).

Of course, this is why we buy gadgets and play with them, and compare notes.

Arcs_n_Sparks

robs · August 13, 2006, 07:06:36 PM

Quote from: Arcs_n_Sparks on August 13, 2006, 06:01:56 PM
Let me make a few control system observations.

PID loops (and their parameters) depend upon the time constants of the overall system and their stability. If you took an empty BS, vent half open, and tuned it, you would have one set of parameters. If you took a BS with 15 pounds of butt in it from a cold start, you are going to get different parameters. Tweak the vent, and you affect the heat balance and system response as well. In the overall scheme of things, this probably doesn't matter a whole lot (unlike fly-by-wire aircraft).

You can always tune to have an overdamped system that will not overshoot (lower value of P). In this case, it will take more time to reach setpoint (the I term will integrate out the error eventually).

Of course, this is why we buy gadgets and play with them, and compare notes.

Arcs_n_Sparks

Good post. I'd like to hear others chime in.

My 2 cents here is that I was using a full water bowl that I "think" would give me enough mass in the box.

Very very good point though.

nodak · August 13, 2006, 08:02:09 PM

tomg says

QuoteHell man that's easy, move to California

The reason we have cold temps here in ND to keep the undesirables out (not to mention any names :D). ) :) :)

TomG · August 14, 2006, 07:51:47 AM

Quote from: nodak on August 13, 2006, 08:02:09 PM
The reason we have cold temps here in ND to keep the undesirables out (not to mention any names :D). ) :) :)

Yup, I imagine the denizens of Hell use the same rationalization. In either case, "you get what you pay for". ;)

bubbagump · August 14, 2006, 04:01:07 PM

Hi Robs,

What do you have your cycle time parameter set to and are you using an SSR output? If you are using an SSR output you should set your cycle time to either 1 or 0. The lower the number the faster the cycle rate and the less over or undershoot you will have. That's the beauty of using an SSR.

My PID is not the Auberin one but I did install one for another person. I had no luck at all running auto tune with Auber TC hooked up to it. I had temp swings and inaccurate readings. But then again you can't expect much of a TC for $5.00. As far as where to mount it goes I eventually mounted it to the inside of my garbage can.

After purchasing the Omega (I'm partial to Omega because I've had such good luck with them) I only had to run the auto tune once and it dialed right in. I didn't have to mess with any indivdual PID settings or program any offset in. The only parameter I changed was the cycle time. I think the default is 2 and I lowered it to 0.

There should be no reason that I can think of that auto tune would not work. When I run it I will place two bricks in the preheated smoker. Then I do not start autotune until the smoker is within 10 degrees of setpoint. If all parameters are set correctly and you have a good TC it should work fine.

I would compare the Auber TC against another digital thermometer to be sure it's accurate. You could also do the boiling water test instead.

Bubbagump