PID Controllers - Everything You Wanted to Know, But Were Afraid to Ask

[Cushing H.]

Because the nature of forging violates almost every condition above, how would you suggest we tweak both ones expectations and operation of these units

William - good question .... here is a shot at thoughts in reply. Since the fundamental trade-off is fine control versus fast recovery from big disturbances - the first thing would be to simply choose one or the other. The derivative term alone (when it is big) can produce fast recovery ... but at the risk of oscillations and taking a while to settle down. the derivative term alone (when it is small) can give really good stability ... but really slow recovery from big disturbances (or a start up).

I think you comment about turning off the PID controller and resetting the integral when you reach your control point is a really, really good idea, and might start to buy you the "best of two worlds(though no on in their right mind would turn off a controller on a distillation column ... as things tend to go "boom"). you could use a fairly large proportional band, and an integral term that will give some "oomph" during startup. on that simulation software, that would look like this if you did not turn off the controller when you reach setpoint:


if you could re-set the integral when you reach setpoint (your turning it off and on), you would instead have something like this:


even better, you could turn off the controller when you are a few degrees below setpoint (I can not emulate that on this software).

Also .... because oscillations or other bad behavior can occur when you take a "stable" system and then disturb it (like putting a hunk of metal into a forge) ... there is really no harm at all, if those oscillations occur, in just flipping the controller off then on ... if the integral is producing those oscillations, doing that might well kill off the oscillations

 

[Cushing H.]

So the way I look at it, the most efficient use of the fuel gives the least stable environment. How should we tweak our settings and/or expectations to best deal with this reality of forging?

Hmmmm .... my first reaction is that your expectations should be based on how much you know about the tolerance of what you are doing to the metal to large temperature swings. If you know that you can get basically the same results from forging if you accept the forge maybe overshooting by, say, 30-40 degrees, then you could make your proportional band really small (and use a fairly large integral term), and get a rapid return to temperature with the acceptance of that overshoot. If you know you just can not accept an overshoot, then you would need to accept a larger proportional band, a much smaller integral term, and a slower return to temperature. "pick your poison" I guess..... ???

 

[Dfarr67]

Would like to give this thread a nudge- lots of good info, PID threads like this are hard to come by. I have a powder coating over that will be used at 350F normally with 450F used to gas off parts. Controller was inherited (I would go another direction now) Inkbird ITC 106RH which has solenoid outputs to a breaker style switch (220V with two 1500w elements). On the whole it operates just OK +- 8F either way, it's the operation that I can't seem to smooth out- although I am open to use in manual mode. In auto tune- things go quite smoothly until it times out, switching on and off normally through the temperature swings- when it exits and goes to PID mode it will heat to 345F and shut off, tops out at 358F then starts to cool down, when it hits 350F it will short cycle several times until about 343F when it finally turns back on 100%. The the cycle continues- since this is not a SSR I have concerns about wear and just not working correctly. I've set Ctl to max at 120.
Another issue is when the PID is shut off for a period of time (overnight?) it seems to lose it's thermocouple (k in this case) setting resulting in ORal fault- it is hard to reset as it is difficult to get back into the sub IP menu to change back--wiring, does the PID have an internal memory battery fault? This is the second unit with this same issue, I do have third to try if it is the PID faulting (and I have read of multiple units binned). The first unit has a warranty 106VH on the way (may try out the SSR)- I'm starting to be wary of Inkbird and I don't know if I'm being fair or not. If this continues- I would probably look for a used higher end unit on fleabay. I have to digest- just wondering on the experience out there.

 

[Cushing H.]

Not the best idea to use PID mode without an SSR. it can work, wear and tear on the mechanicals could be a factor (or might not ...depends on luck), but you definitely won’t get the fine control that a non mechanical system can deliver. depends on what level of control accuracy you want.

I note you say the unit you have loses its settings. That should not happen ... I would question the health of the unit. As you said, there are several reports of inkbird controllers apparently being defective....

 

[Dfarr67]

First time working with PIDs, had an electrician friend wire up the control box but I'm on my own from here----> painful. And with more reading realizing the cheaper PIDs aren't easy to deal with.

Loses the thermocouple setting, yesterday I had a devil of a time getting back into the sub menu to change it then it scrolls through the types not letting you pick one. I get it....having a defective unit- but two in a row makes me think I'm doing something wrong. But once that is reset- it works. I do have a third on hand.

Yes, I know the system design as-is is antiquated- but it is what it is, and I need it to last for a year or so. SquareD contactor with Inkbird 106RH which is the internal relay version of the VH. Functions very smooth in the autotune mode takes forever tho to cycle through AT, did two rounds last night from cold then at temp. Next step is to play with the P/I/D settings manually and then if there is still an issue with short cycling will change to manual mode. What is the difference between 'manual' and 'manual inhibit' ?
I have another PID 'SET' brand that came with the oven- but it is for an SSR, the contactor was already installed at the time, I have a warranty replacement Inkbird 106VH coming also. So I am pivoting towards the SSR.
What is a good brand 40amp SSR that I could pull off of Ebay- is a used pull out advisable, I know they do have a finite life span? What is a functional equivalent PID model number for the 106VH in a Watlow?

Also interested (and searching) in wiring up the Inkbird timer to start countdown when temp is reached so I can prep oven and start up cold without opening door when hot. Now it is full manual with pause and reset and shuts oven off when run down.

Thanks for the replay and glad you kept going after a few negative comments- I find these threads few and far between, especially with specific PID brand issues.

 

[Cushing H.]

Apparently manual inhibit causes the controller to not accept being switched into manual mode.

the rest of your post I need to spend a little time making sure I understand what is being said ... but ... if it works “well enough”, even with the short cycling, I would just leave it be. The noise might be annoying, and you might get some reduction in lifetime of the contactor (and also get more stress on the heating element) ... but if you just need it to last a year or so, then maybe just leave it as it is?

if you really want to go to an SSR at this time ... they are much more straightforward/simpler than the controller, so likely any brand will be ok ... just make sure you get one rated appropriately for the power running through it.

 

[Dfarr67]

Apparently manual inhibit causes the controller to not accept being switched into manual mode.

the rest of your post I need to spend a little time making sure I understand what is being said ... but ... if it works “well enough”, even with the short cycling, I would just leave it be. The noise might be annoying, and you might get some reduction in lifetime of the contactor (and also get more stress on the heating element) ... but if you just need it to last a year or so, then maybe just leave it as it is?

if you really want to go to an SSR at this time ... they are much more straightforward/simpler than the controller, so likely any brand will be ok ... just make sure you get one rated appropriately for the power running through it.

Well, still working through options/settings before I settle for 'good enough' I would prefer to maximize component life if I can, it is far better than it was a month ago for the cycling maybe 3-4x on the down slope before heating and I've increased the Ctrl time so it's not cycling as fast as it was previously. For component shopping- I strongly prefer non-china (or mexican for that matter), I go out of my way for NOS or used.

 

[C L Wilkins]

This has been an interesting read. I come from the same discipline, having worked in a large industrial complex. I was fortunate enough to have known Bobby Ewing. He was on the original PID development committee way back when. He worked at the Humble Oil Refinery (now ExxonMobil) at the Baytown, Texas refinery and was a wealth of knowledge.