PID tuning / bed heater control

Hello! I'm after a little help with PID tuning on the duet.

I've installed a 200x300 heated bed on my Ormerod 1 and the default PID settings seem fine for PLA printing at 55 degrees. I'm looking to print in ABS, so as a test I've commanded it up to 110 degrees. It's taken 15 mins to go from 55 to 75, and I don't have time to leave it running to see if it ever reaches 110. I'm hoping some PID tuning will remedy this.

Did a bit of reading and it seems the bed is using bang-bang control instead of PID as the P term is set to -1. This is where I'm stumped, and I can't find much online about how to tune such a controller.

Default settings (using dc42-1.09m):
Heater 0 P:-1.00 I:2.500 D:1000.00 T:2.70 S:1.00 W:255.0 B:5.0

I've also been experimenting with a 24V power supply and E3D hot end as well with little success, though have now moved back to 12V and stock hot end. I don't believe the problems I had were directly PID related but it was dropping a couple of degrees once printing started, so any advice on tuning the hot end PID would be welcomed too. I gather the S value can be used to account for changes in supply voltage, but since the hot end design is very different I think a full retune would be in order.

Many thanks in advance

How did you mount this bed?

---

Slicer: Simplify3D 4.0; sometimes CraftWare 1.14 or Cura 2.7
Delta with Duet-WiFi, FW: 1.20.1RC2; mini-sensor board by dc42 for auto-leveling
Ormerod common modifications: Mini-sensor board by dc42, aluminum X-arm, 0.4 mm nozzle E3D like, 2nd fan, Z stepper nut M5 x 15, Herringbone gears, Z-axis bearing at top, spring loaded extruder with pneumatic fitting, Y belt axis tensioner
Ormerod 2: FW: 1.19-dc42 on Duet-WiFi. own build, modifications: GT2-belts, silicone heat-bed, different motors and so on. Printed parts: bed support, (PSU holder) and Y-feet.
Ormerod 1: FW: 1.15c-dc42 on 1k Duet-Board. Modifications: Aluminium bed-support, (nearly) all parts reprinted in PLA/ ABS, and so on.

If the bed heater LED is permanently on (as I suspect), then it has nothing to do with PID control and means that your bed heater is not powerful enough to reach 110 deg. You probably need to supply it with a higher voltage.

Dave

Thanks for the responses. I've mounted it to the aluminium bed plate using the same bolts - I simply got an aluminium heat spreader cut by an ebay seller and drilled holes through both it and the heater plate (the edge of the plate has no copper). Seems to work beautifully, though I had to reduce speed a little as the extra weight was giving the bed a lot of momentum and introducing wobbles. I also have an extended Y axis (400mm) to accommodate this.

There's no LED on the bed. I could wire one in to check but I suspect you are correct in saying it's on all the time. I was hoping 24V wouldn't be the answer as I've had trouble getting the 24V E3D end to print as well as the stock end. I guess I'll have to give that another try. No matter what I did I couldn't stop it leaving horrendous levels of stringing on the print. The bed is both 12 and 24V compatible, just requires changing the + and - connections presumably to use a track with the appropriate resistance.

I've got some Ormerod E3D profiles somebody offered up so I'll see where I get with exploring that. I think for once I don't have any projects on the go that need printed parts, so can afford a little downtime to get 24V working! Any tips for PID tuning the E3D hot end? Is it just a case of pushing P up until it heats up in a reasonable amount of time, then using I to reduce the overshoot? Presumably then backing P off if the overshoot is difficult to correct.

I didn't have to change any of the PID parameters when I changed the heater cartridge on my E3Dv6 form 12V to 24V. I did find that the E3D needed more retraction when I was using previously. I think this may be because the push-fit collet for the Bowden tube allows a little movement.

What is the resistance of the bed heater, either between the 24V terminals, or in your present 12V configuration?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

You may get away with just increasing the voltage to 14V or so if your PSU has an adjustment, and you can keep everything else the same. Power increases as the square of the voltage, so a couple of volts extra has a surprisingly large effect on the amount of heating.

Dave

Quote
dc42
I did find that the E3D needed more retraction when I was using previously. I think this may be because the push-fit collet for the Bowden tube allows a little movement.

yes thats what i noticed aswell ,i have put some hard coper wire just under the plastic push fit collet and it does not move anymore and dont see any blobs anymore( altho i have not teset with 2mm retraction yet)

My large delta printer has a 660mm long Bowden tube and needs 8mm retraction using the E3DV6. The J head I was using before needed only 6mm.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I'm going from a 40W 12V cartridge in the stock end to a 25W 24V cartridge in the E3D end, would the lower wattage account for the lacklustre performance at default PIDs? I'll take a look at securing the bowden clip, but even with 8mm of retraction I was getting strings. With the same PLA in the stock end I needed only 4mm. I tried experimenting with retraction speeds and a higher speed helped but still wouldn't eliminate it.

The bed measures 1.6 ohms in 12V configuration and 5 ohms in 24V.

I'm using the standard ATX power supply that came with my Ormerod1 for the 12V setup, which isn't adjustable (as far as I know). I have a 480W LED supply for the 24V setup.

I have tried with 2mm retraction(with ormerods 300mm bowden i think) and no blobing ,but i did also change 0.4nozzle for the 0.5 nozzle.

Edited 1 time(s). Last edit at 01/28/2016 03:58PM by Darathy.

Quote
static65
I'm going from a 40W 12V cartridge in the stock end to a 25W 24V cartridge in the E3D end, would the lower wattage account for the lacklustre performance at default PIDs?

If by that you mean you are getting undershoot when heating up, then yes. Increasing the S parameter to about 1.5 may be sufficient to fix it. If not, try this. Set S back to 1.0. Heat the nozzle to 200C and allow the temperature to stabilise. Then use M573 to find the heater PWM. Divide that by 180 (i.e. 200C minus room temperature, then use the resulting values as the T parameter in the M301 command. That should get rid of the worst of the undershoot.

One day I'll implement PID auto tuning, but for now I have more important features to implement.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks dc42. I've had a fun morning of testing. It ended when my 24V power supply exploded! It appears the cooling fan on the PSU died, and it presumably went into a thermal protection mode. I heard the hot end fan stop so went to investigate. As I got to the table the hot end fan started again, which is when I noticed the PSU fan wasn't spinning. A couple of seconds later before I could even unplug it there was a BANG and some sparks and that was the end of my 24V supply Plugged the 12V supply back in and the Duet seems to be fine - phew.

Anyway, I tested with S=1.5 and that seemed to give the best performance with the least undershoot. I reset S to 1 and then tried adjusting the T value but I'm not sure if I did it correctly. M573 P1 gave me 0.715, divided by 180 gave me 0.00397. I rounded that up to 0.004. This seemed to take even longer to reach 200. See graphs:



When I've got my hands on another 24V supply I shall get back to this. It was an eBay job so hopefully I can get a replacement with relatively little hassle.

Quote
static65
I reset S to 1 and then tried adjusting the T value but I'm not sure if I did it correctly. M573 P1 gave me 0.715, divided by 180 gave me 0.00397. I rounded that up to 0.004. This seemed to take even longer to reach 200.

OK, I got that wrong. I was assuming M573 P1 returned the PWM on a scale of 0-255, but evidently it uses a scale of 0-1. So you need to multiply by 255.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].