Wrong dterm formula ?

Discuss tuning PID coefficients to maximize system efficiency.
Post Reply
JImram
Posts: 12
Joined: Fri Jan 02, 2015 8:03 am
Bot?: No

Wrong dterm formula ?

Post by JImram »

I optimalized my PID koeficients and everything work well, only in temperature ramp mode the regulation is very slow. I think that is because the formula "dTerm = dGain * (curr_temp - previous_temp)" is used. The better formula is "dTerm = dGain * (curr_err - previous_err)". I thing that with this formula the PID regulator would work better.
Why the formula "dTerm = dGain * (curr_temp - previous_temp)" is used instead of standard "dTerm = dGain * (curr_err - previous_err)"?
brahn
Posts: 543
Joined: Thu Dec 13, 2012 11:01 am
Bot?: No

Re: Wrong dterm formula ?

Post by brahn »

You're correct that it is using dGain * (curr_temp - prev_temp). I don't know why that formula was chosen. I'm a bit hesitant to change this before releasing the 4.0 firmware, but I'd be open to making a build of it for you to test out with the other formula to see if we should switch post-4.0. I'd have to do a little more work to see exactly what would be involved.
JImram
Posts: 12
Joined: Fri Jan 02, 2015 8:03 am
Bot?: No

Re: Wrong dterm formula ?

Post by JImram »

Thanks, I´d like to test it out.
Post Reply