blondica73 wrote:I have a question about sliders in Artisan: I use the Python routine to send button commands to the Arduino board where I specify the value; see my screenshot above. How could I use the sliders portion to do a similar thing? Meaning, as I move the slider, a new value to be passed to the Python routine?

See below. Use your artisanprog.py file here.

blondica73 wrote:My first attempt to roast using this setup failed miserably. Delta BT kept switching between 0 and a high value, see below.
Please keep in mind that this is a drum roaster.
Any suggestions?

As I mentioned in my PM's to you, I only use PID for warm-up and BBP and it works great for this purpose. I played with using it to automate roasting itself but it was a fool's errand for me. And judging from other posts in the Roasting forum others have had similar experiences. I much prefer incrementally tweaking a background profile from roast-to-roast for a particular green to hit my desired targets.

That said, suggest you set D=0 because rapid changes don't occur in roasting and it will likely just amplify noise.

blondica73 wrote:One anomaly I noticed is that my stepper motor seems to have lower torque than expected. It seems to work ok when I power up the stepper motor, Nema 17 Stepper Motor 42BYGH, and during the first few moves. However, as the servo moves the shaft back and forth, I hear a grinding noise, and the shaft does not spin. The gas valve seems to be hard to turn, and I think that the motor does not have adequate torque; it is 42Ncm(60oz.in). I'm considering going to a Nema 23 stepper motor, 269oz.in/1.9Nm. Did anyone experience this?
I've seen setups using the 20kg servo motors, which have a high torque, 22.8kg*cm (316.63oz*in).

My NEMA 17 servo is connected to a gas needle valve and provides more than enough power. Some thoughts before you get a more powerful servo which might damage your gas valve:

1) When the EN pin is low on the TB6600 controller it is designed to lock the position of the servo and it won't turn. So ensure the binding shaft isn't due to that.

2) If you manually turn your gas valve on/off does it require more than light finger pressure through the full range? Does the flow stop/max out only when the valve is fully mechanically closed/open? On my valve the flow stops before the needle valve is fully screwed in tight, and there is still free play too when fully open to my max gas setting. So little torque is needed at all by the servo.

3) I have only had grinding when my valve was out of synch with the position of the servo. Before I begin a roast session, I manually turn the gas valve to a known flow-off position if needed. Then I calibrate the servo code to this flow-off position using an Artisan event button that sends a "CAL" command to it. (Think of this like taring a scale before you use it). If you aren't doing this it can cause grinding because the code otherwise doesn't "know" where the gas valve position is.

4) Have you ensured there is no binding in the servo connector between the servo and your gas valve? Have you also ensured there is some play along the longitudinal axis of the servo so it can move forward and backward? (I need this on my needle valve because as you turn it the shaft moves in/out.)

blondica73 wrote:The second attempt was much better; still off, but better. The drop in ROR is due to turning off curve correction. My sampling rate is 3s as my system, BC-5, is slower to respond, and if I don't do smoothing, delta BT and ET see wide swings.

Hmmm.. Getting PID to work is challenging enough, but if your TC/metering setup is slow I think it would compound the challenge. Perhaps consider upgrading this at some point. I'm a fan of thinner/faster TC's and RTD's in particular.