Auto-tune heater



  • I just launched a M303 auto-tune command. Here is the M307 resulting params:

    M307 H3 A748.3 C962.6 D4.5 B0

    It improved a lot the over-shoot, but now, the temperature is not very stable: it oscillates +-1.7°C; was only +-0.2°C without the tuning…

    Which param can I tune to reduce this oscillation?


  • administrators

    To reduce the oscillation, increase D.



  • Thanks.

    I had to increase D up to 10 to keep oscillations below +-0.2°C. But now, it takes muuuuch longer to reach the setpoint; Is there another param I can change to fix that?



  • Assuming H3 is a hot end (as H0 is normally the bed, then your M307 shows a gain of 748.3 which means that if left unattended it could potentially reach a temperature of around 750 deg C! That would mean a massively powerful heater of around 80Watts plus. Is that what you have? If not, then another explanation could be that you have a 12V heater running on 24V.

    You ought to check that out but one way to "tame it" would be to tune it with a PWM of 0.5 or so.



  • Yes, I saw the warning. As I was running at 12V, and now switched to 24V, I checked if I didn't forget to switch the cartridge, but no: 14.5 Ohm -> 41W…

    It's a E3D Lite6, with the silicone insulator. How could it reach 750°C? Is it really possible with 40W?



  • I just make a new auto-tune cycle, with PWM 0.5. Now params are:

    M307 H3 A352.5 C275.7 D3.2 B0 S0.5

    It takes longer to reach the setpoint, and I still have a huge oscillation (+-1.4°C)!



  • I checked, and default params are:

    M307 H3
    Heater 3 model: gain 340.0, time constant 140.0, dead time 5.5, max PWM 1.00, mode: PID

    It overshoots 2°C, then takes 15s to get back to the setpoint, and stay there within 0/-0.2°C…

    Should I let the default values?



  • It will take longer to reach the set point using 50% PWM because it's running at half power (more or less). I know that the silicone sock makes a difference but I'm surprised it makes that much difference. I guess you could try tuning without the sock, just to see. TBH 2 degrees overshoot but recovering within 15sec isn't too bad and is unlikely to result in huge problems I would have thought.

    I'm still very surprised that the sock makes that much difference so maybe try tuning without the sock but with PWM of 1.0 just to check. If you still get a warning and up with a high gain, then it has to be the heater cartridge that is too powerful.


  • administrators

    If it overshoots the set point, increase the A parameter in the M307 command.



  • I will try to tune without the silicone sock, to see its effect…

    And I will test with different values for A.

    Thanks.



  • Here are the results without sock:

    Heater 3 model: gain 878.2, time constant 287.3, dead time 4.9, max PWM 1.00, mode: PID

    BTW, as I mostly run around 190-200°C, should I use 200 for S param in the auto-tune sequence, instead of 240°C?



  • @fma. For info, I know I have a different hot end to you (Diamond 3 and 5 colour) but I get the following with the 3 colour version

    40 Watt cartridge without Silicone sock, A349.7 C233.5 D7.0
    40 Watt cartridge with home made silicone sock A 360.6 C213.5 D6.0

    So it makes a bit of difference but not much.

    Now, the Diamond 5 colour comes with an 80Watt cartridge (I don't understand why) but tuning that with PWM of 1.0 I get:

    A647.7 C190.4 D6.4 so the gain doubled and was similar to yours which is why I thought your heater must be very powerful.

    with PWM of 0.6 I get A350.4 C194.8 D5.2.

    HTH

    Edit. Just seen you last post - that gain is higher than my 80Watt cartridge!



  • Strange! As said, my cartridge has a resistance of 15 Ohm, which gives 40W under 24V…

    I will monitor current and voltage during heat up to see if there is not a problem (short circuit inside the cartridge, lowering the resistance when heated up... Not convinced, as I would have seen a change in heating curve).



  • I had a cartridge fail once. Measured the resistance and it was fine but it wouldn't heat up at all. Double checked the resistance, treble checked it, wiggled the wires around, checked the connector. No problem but it still wouldn't heat up. Fitted new cartridge and all was fine. Weird…......



  • Well, my cartridge is OK: I measured a current of 1.6A during all the heating… So, 40W. And 0.38A to maintain 190°C (~9W).

    It is normal that it heats up much faster than the Diamond, as the E3D is smaller. But why such a difference with a standard E3D? I don't know.

    I'm using a Lite6, with the last heater block...



  • FYI, it takes 77s to reach 190°C on my E3D Lite6. Is it something similar to you?



  • @fma:

    FYI, it takes 77s to reach 190°C on my E3D Lite6. Is it something similar to you?

    No nothing like, but then my 5 colour Diamond hot end is great big lump of brass with 5 heat sinks screwed into it and a high flow (20cfm) fan blowing over them. Currently with the 80Watt heater running at 60% PWM it takes around 120 seconds to reach 190deg C but then it overshoots by around 3 degrees and takes about 20 seconds to stabilise. This is about the same time it used to take the 3 colour Diamond with a 40 Watt heater but I only used to get about 1 degree overshoot. When I get time, I'm going to ignore what RepRap.me say and fit a 40 Watt heater cartridge instead of the 80Watt that they recommend.


  • administrators

    Ian, I presume you tuned the new heater? You can increase the A parameter to reduce overshoot.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.