PID controller minimum On time ?

[brewman !]

I'm using a generic PID controller (JLD612 PID ). It has the usual array of settable parameters. The user manual for it is here: https://mythopoeic.org/misc-files/JLD612_Manual.pdf

I'm using it to control a DIY 5KW induction element. Meet Thing 1. A 5Kw Induction All In One...

Everything works great except that the induction element takes about 3 seconds to turn on from the time it is requested. And when my mash gets close to the set point, the PID controller will request heat from the induction heater in very small time increments. Maybe 2 seconds on, 3 seconds off. (I have ot, control period, set to 5 seconds, meaning it checks temp every 5 seconds). When the PID controller does this, the element doesn't even get turned on before the PID controller is turning it off.

I had ot set to 30 seconds, but that was too slow and the temp would overshoot. What I need is ot set to 5 seconds and some say to set the minimum time that the output can be on.

Is there a way to do this ?

 

[Bobby_M]

So you have it set to 5 seconds and that's too short and you tried 30 seconds and it was too long. I suspect there are a few settings in between 5 and 30 that would get you closer to your target but that's so obvious that I can't assume you haven't tried it. I would try 10.

 

[Vale71]

With electric as there is no inertia one should have the control period set to the shortest possible value for better overall response.
If you are still experiencing over- or undershoot you need to better tune your PID parameters and not play around with the control period.

Everything works great except that the induction element takes about 3 seconds to turn on from the time it is requested.

Don't know what you mean by that. As electricity travels at the speed of light unless your controller is literally in outer space the element will turn on instantly as soon as the PID commands the SSR to close the circuit.

 

[brewman !]

So you have it set to 5 seconds and that's too short and you tried 30 seconds and it was too long. I suspect there are a few settings in between 5 and 30 that would get you closer to your target but that's so obvious that I can't assume you haven't tried it. I would try 10.

Ot is the control cycle time. It is how often the PID can turn the output on and off. It has nothing to do with how LONG the PID turns the output on. But probably by default longer ot times = longer minimum on times.

 

[brewman !]

With electric as there is no inertia one should have the control period set to the shortest possible value for better overall response.
If you are still experiencing over- or undershoot you need to better tune your PID parameters and not play around with the control period.

Don't know what you mean by that. As electricity travels at the speed of light unless your controller is literally in outer space the element will turn on instantly as soon as the PID commands the SSR to close the circuit.

There is a time delay between signalling the induction driver board to turn on and when it actually sends power to the induction coil. There are some capacitors that get charged and then it slowly builds the induction coil current. I think it does this to sense the inductance of the coil and the pot combined before it goes to full power. It doesn't do a step from zero current to full current.

The induction coil driver circuit is resonant and every pot will result in a different resonant frequency. Even moving the pot changes the frequency. I think part of the start up process is rediscovering the new frequency.

Residential induction cook tops do this too.

We had a high end KitchenAid cooktop that had a timer in it to allow you to remove the pot for 3 seconds before it shut down and restarted. I think it was to allow the chef to toss whatever was in the pot/pan and put it back on the coil without the coil having to do a cold restart.

 

[Vale71]

My bad, I was thinking of resistive heating elements which clearly don't have this issue.

If that's how the cooktop works then I'm afraid a PID controller with PWM output is really unsuitable. You should see if there is a way to change the power setting directly, without ever turning the unit off, using some tipe of proportional output.

 

[brewman !]

The driver operates in a narrow range of voltages. You can't control it by turning down the input voltage.

 

[brewman !]

Just to clarify... the PID controller works as it is now. I'm concerned about blowing up the induction driver board when the PID repeatedly signals it to start the coil and shuts it off before it does. Maybe it will last forever doing this, I don't know.

This driver board has another input to turn the output on and off, but it doesn't work properly and the supplier is of no help whatsoever. The turn on time with the other control input would definitely be shorter, but probably still not short enough. It works intermittently enough that I know.

 

[brewman !]

Brewed last night.

0

 

[BrunDog]

You could switch to a standard thermostat. I know we like to use PIDs, but the reality of heating water is they are not necessary - water temp stops climbing the instant the heat is turned off, so there is no loading effect like a coiled spring. Depending on the hysteresis window, it would reduce cycling and practically assure a certain amount of ON time.

 

[Vale71]

The driver operates in a narrow range of voltages. You can't control it by turning down the input voltage.

I meant you should find a way to regulate the board's internal power setting, not regulate the voltage itself which obviously would not work. Unfortunately this requires hacking the board and without the manufacturer's cohoperation, which you already know you're not going to get, this will most likely prove impossible.

 

[Vale71]

You could switch to a standard thermostat. I know we like to use PIDs, but the reality of heating water is they are not necessary - water temp stops climbing the instant the heat is turned off, so there is no loading effect like a coiled spring. Depending on the hysteresis window, it would reduce cycling and practically assure a certain amount of ON time.

Seriously?

 

[brewman !]

You could switch to a standard thermostat. I know we like to use PIDs, but the reality of heating water is they are not necessary - water temp stops climbing the instant the heat is turned off, so there is no loading effect like a coiled spring. Depending on the hysteresis window, it would reduce cycling and practically assure a certain amount of ON time.

You are right. I wonder if I could get the PID controller to act like an on/off thermostat by changing the gains ?

 

[brewman !]

I meant you should find a way to regulate the board's internal power setting, not regulate the voltage itself which obviously would not work. Unfortunately this requires hacking the board and without the manufacturer's cohoperation, which you already know you're not going to get, this will most likely prove impossible.

I think that would be pretty hard to do. Let me think about it.