PIDing my TO/UFO

[Tex]

Putting a PID & thermocouple into the mix is simple. Now what can I do with the info it provides?

Do I put the PID setpoint at 465?F and let it blast away at full power until it's roasted? Maybe set it @380?F and let it run at that point for x minutes AND then bump the setpoint to the final temp?

I'm so used to running my TO/UFO with so little feedback that I'm not sure where to go with all of the control I now have. When does this begin to get simpler?

Suggestions/ideas?
 :icon_scratch:

[peter]

Full blast 'til the temp reads around 280.  Cut the heat then, since it will coast up to 300.  Let it sit there for 3-4 min., and go full blast 'till around 345.  Cut the heat then, since it will coast up to 350.  Let it sit there for 3 min., and go full blast 'til the outliers of 1st crack.  Then have it pulse the heat to slowly rise through 1st, and afterward hold or slight rise for a couple 2-3 min.  That's where a lot of my beans wind up, but you could go full blast again toward 2nd, and kill heat at the outliers of 2nd, since it will coast right into a light Vienna.

Not sure if that's the input you were asking for.

[Tex]

Just the kind of info I was looking for Peter! But it's been ages since my statistical analysis studies classes, so help me understand outliers within the context of first & second cracks - initial stages (when pops are slowly ramping up) or last stages (when pops are slowly ramping down); both have outlier possibilities, right?

I'm going to wire the fan directly to the line-in power so it'll always be running at max velocity. I'll try the PID with Auto Tune (Watlow 935A) with the TO power @500?F. Any suggestions for setting the PID parms manually?

[peter]

Outliers could indeed refer to either end of the crack, but I generally use the first few snaps to define outliers.  With some coffees, the outliers start really early, and the actual crack hasn't even started. 

Sounds like you're on the right track.  Will the PID be able to factor in the slow response time of the TO heating element?  Maybe it'll keep it pulsing on and off in an effort to keep the element at the current temp so that there isn't much of a lag.

[milowebailey]

Putting a PID & thermocouple into the mix is simple. Now what can I do with the info it provides?

Do I put the PID setpoint at 465?F and let it blast away at full power until it's roasted? Maybe set it @380?F and let it run at that point for x minutes AND then bump the setpoint to the final temp?

I'm so used to running my TO/UFO with so little feedback that I'm not sure where to go with all of the control I now have. When does this begin to get simpler?

Suggestions/ideas?
 :icon_scratch:

Tex

I know peter is the expert at both ends of the crack.    but can you roast a few batches with the thermocouple installed and record the results of a manual roast.... listen for the first crack... second crack and see where your PID is measuring the temp... ramp... etc.  Start there and then tinker with what you think could improve on a manual roast... maybe ramp and hold... etc.  Automation only helps if you know what you want to do differently that you can't do easily manually or have to do repetitively... again and again... over and over... you get my point..  Just my 2 cents.

[J.Jirehs Roaster]

I have thought about PIDing the TO but I could never get the ramp hold on my PID to work so it makes a great temp probe... I do use it on the hot air popper and just manually raise the target... would the P, I, and D settings need to change that much for the responce time of the TO verses air popper??

[Tex]

Those are all good questions, and the short answer is "...beats the sh#! outa me." If Jim Gallt is a member here I'd recommend asking him. If he's not then someone should invite him. I know that he mentioned using J&B Weld to affix a t/c to the bottom of his roaster.