Adapting a PID to Probat L12

[CCafe]

Has anyone tried to do a aftermarket PID to a commercial roaster? Our Probat L12 has a Watlow SD Limit and I know the SD series is in the same family with their PID. I talked with a Watlow distributor and he said it was very doable.

I know I would need a PID and thermocouple. Should I consider a variable electronic gas valve or a simple on off valve?

I would still want to leave the SD Limit in place just in case of an over limit.

 

[La Crema Coffee]

And the reason would be?

 

[CCafe]

The biggest reason would be to loose the manual gas valve control and have it be controlled electronically.

The reason we want to try a PID is that it can more accurately judge and hold the temperature to a finite degree. On top of that it is variable and the temp could be adjusted on the fly. Also the electronic log file could be easily downloaded and examined if needed.

I understand roasting is an art. We are just looking at the possibility of adapting electronic control to help in the process.

 

[ElPugDiablo]

If you have just a simple on off valve, how fast can you restart the flame once it is off? On a Probat it may not be a problem, on an IR Diedrich it would be a bad thing.

On a maybe somewhat related topic, anyone try to measure the drum's actual temperature throughout the roast? I don't mean the atmospheric or beans surface temperature, but the metal drum itself.

 

[CCafe]

Well when I talked the Watlow guy about the PID he said a variable gas valve really wouldn't shut down the flame. He described it more like it would drop the gas down to a low flow.

As for the actual drum temp. I don't think it would be to hard to do with a laser temperature finder.

 

[ElPugDiablo]

I have seemed Renegadee roasters with their Agtron system III. I think they use on off, but I am not sure. Very impressive roasters.

 

[CCafe]

I see they keep a lid on Agtron Scientific Roasting System II. I guess they really want you to call.

 

[johng99]

Although we all talk PID, I am sure we are probably using no "D", very little "P" and mostly (I) due to the long time constants involved in heating a roaster. I have never tried to measure them (no time), as this would be very difficult for a number of reasons (many variables, changing process with time, environment, etc.) and usually when this is the case a fairly "detuned" PI control algorithm is used if my old process control thinking is right.

Many of these controllers have "self-tuning algorithms" but you don't really learn much from this and you can end up with problems.

This may all sound like nonsense, but if anyone has taken a process controls approach to controlling the roaster I would be interested in learning from you.

 

[CCafe]

I'm assuming when you were referring to the self-tuning algorithms you were talking about the initial setup of a PID?

 

[td]

Interesting discussion. Since I have some experience with PID functions and coffee roasting I think I might be able to help a little.

John is correct about D (derivative) not be used in coffee roasting because roasting is not that dynamic.

P and I are the two functions that you use to \"trim out\" your controls. But one cannot successfully be used without the other.

One set of PIDs will not be valid through an entire roast as the heat sources and the heat absorption rates of the coffee vary throughout the roast.

Additionally, the PID settings need to be different for different ranges of batch sizes.

In our case we use 3 sets of PID settings per roast, and have 4 sets for each roaster ( full, half, micro and pico) as each roaster has slightly different dynamics.

The above information would be true of systems using bean temperature as the control temperature- things are slightly different for those systems that use environment/drum temperature for control.

Finally, just buying a Watlow SD or anyother PID controller will not automatically give you the results you want. Furthermore, while buying a proportional modulating valve will certainly give you a wider range of control- without the burner being off- you must have some way for the controller to talk to the valve, usually via a signal conditioner. You cannot just hook one up to the other. With an On/Off valve you can control it with just a switched relay (a function that is in all PID controllers). Also, the techs at Watlow are awesome, but they are not coffee roasters and they do not understand the coffee roasting process. Meaning they may mislead you without actually trying to do so.