Duet3D Logo Duet3D
    • Tags
    • Documentation
    • Order
    • Register
    • Login

    Bed PID Tuning: How long should it take?

    Scheduled Pinned Locked Moved
    Tuning and tweaking
    5
    15
    1.3k
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • DonStaufferundefined
      DonStauffer
      last edited by DonStauffer

      I just insulated the bottom of the bed and tried to run a PID tuning. Almost 2 hours and still going. Bed temperature slowly seesawing above and below the target, on about a 4 min 20 sec cycle:

      9/8/2024, 3:59:07 PM M303 H0 S70
      Auto tuning heater 0 using target temperature 70.0°C and PWM 1.00 - do not leave printer unattended
      9/8/2024, 3:59:12 PM Auto tune starting phase 1, heating up
      9/8/2024, 4:01:39 PM Auto tune starting phase 2, settling
      9/8/2024, 4:16:43 PM Auto tune starting phase 3, measuring
      9/8/2024, 5:49:26 PM M303
      Heater 0 is being tuned, phase 4 of 4, measuring

      Also, why does it say "phase 4 of 4" when it's only on phase 3 according to the original command?

      DonStaufferundefined 1 Reply Last reply Reply Quote 0
      • DonStaufferundefined
        DonStauffer @DonStauffer
        last edited by DonStauffer

        @DonStauffer Finally:

        9/8/2024, 6:03:08 PM Warning: heater behaviour was not consistent during tuning
        Auto tuning heater 0 completed after 3 idle and 25 tuning cycles in 7440 seconds. This heater needs the following M307 command:
        M307 H0 R0.307 K0.119:0.000 D21.47 E1.35 S1.00 B0
        Edit the M307 H0 command in config.g to match this. Omit the V parameter if the heater is not powered from VIN.

        What does it mean "was not consistent"? What does one do about it? I've always found the bed heater to be consistent as far as I can tell.

        I've been doing lots of perusing previous posts about PID tuning and didn't find anything helpful in my situation. I did notice discussions about where the bed thermistor is. In my case (Railcore) the thermistor in use is in a slot on the bed top surface and the heater is on the bottom. So I'd estimate there's 4mm of aluminum bed between heater and thermistor. Keenovo 700W 120VAC heater controlled by SSR. Heater does have its own thermistor but it's not being used.

        moth4017undefined 1 Reply Last reply Reply Quote 0
        • moth4017undefined
          moth4017 @DonStauffer
          last edited by moth4017

          @DonStauffer the bed PID tune does take a long time i have a 340mm x 340mm x 8mm thick Keenovo 700W /220Vac , if you have a heated or insulated chamber even longer .
          here is my PID tune if it helps

          My sensor is in a hole drilled in the edge of the aluminum bed . not using the one on the silicone heater

          M307 H0 R0.838 K0.331:0.000 D3.08 E1.35 S0.90 B0

          if you send M307 H0 in the console you will get a report
          Heater 0: heating rate 0.838, cooling rate 0.331, dead time 3.08, max PWM 0.90, mode PID
          Predicted max temperature rise 198°C
          PID parameters: heating P69.2 I2.156 D149.1, steady P69.2 I6.103 D149.1

          <

          1 Reply Last reply Reply Quote 0
          • moth4017undefined
            moth4017
            last edited by

            from the settings above i ran an auto tune to see how long it would take , ambient air 18C
            M303 H0 P1 S100
            the PID tune took 37 minutes and i got back this response

            Warning: heater behaviour was not consistent during tuning
            Auto tuning heater 0 completed after 4 idle and 25 tuning cycles in 1620 seconds. This heater needs the following M307 command:
            M307 H0 R0.812 K0.268:0.000 D3.35 E1.35 S1.00 B0
            Edit the M307 H0 command in config.g to match this. Omit the V parameter if the heater is not powered from VIN.

            My first auto tune took much longer

            so my bed has been working fine with those settings above and the results were not much different .

            previous settings working ok "M307 H0 R0.838 K0.331:0.000 D3.08 E1.35 S0.90 B0"
            settings after the PID tune "M307 H0 R0.812 K0.268:0.000 D3.35 E1.35 S1.00 B0"

            <

            DonStaufferundefined 1 Reply Last reply Reply Quote 0
            • DonStaufferundefined
              DonStauffer @moth4017
              last edited by DonStauffer

              @moth4017 Mine took 2 hours. I'm a bit disappointed that the bed doesn't heat any faster with the insulation. I've done this before almost the same way and did get a massive decrease in heat time (it was a Marlin machine). Maybe this tuning algo doesn't handle insulated beds very well. If so, maybe tuning it manually might help.

              The bed is around 300x300mm. No big deal, everything works. Once the tuning finished I put the line in config.g.

              engikeneerundefined 1 Reply Last reply Reply Quote 0
              • Phaedruxundefined
                Phaedrux Moderator
                last edited by

                What M307 did it give you and how did it compare to what you had before?

                Z-Bot CoreXY Build | Thingiverse Profile

                1 Reply Last reply Reply Quote 0
                • engikeneerundefined
                  engikeneer @DonStauffer
                  last edited by

                  @DonStauffer the PID tuning cycle measures both heating and cooling rates. Its always going to take longer to cool down, especially with the additional insulation, hence the tuning cycle took longer over all despite the heating being slightly faster.

                  As the bed goes through the tuning cycle it kicks off heat and warms up the air around it, more so if the printer is partially enclosed. This means the cooling rate changes slightly over time. This may happen more with the insulation applied as less heat comes out the first few measurement cycles.

                  Your tuning took the full default 25 cycles to complete. I'm guessing this is because the cooling rate was changing over time (as explained above), meaning it wanted to use more cycles to collect more data, and hence the warning message you got.
                  I'm guessing previously it did not need to use as many cycles, so also took less time.

                  The important thing is how it performs in a print. Does it heat up any faster/slower? Is it stable over a print? Those are all that really matter

                  E3D TC with D3Mini and Toolboards.
                  Home-built CoreXY, Duet Wifi, Chimera direct drive, 2x BMG, 300x300x300 build volume
                  i3 clone with a bunch of mods

                  DonStaufferundefined 1 Reply Last reply Reply Quote 1
                  • DonStaufferundefined
                    DonStauffer @engikeneer
                    last edited by

                    @engikeneer It does not heat up faster or slower. I'm disappointed and I wonder if I can manually tune it better.

                    droftartsundefined 1 Reply Last reply Reply Quote 0
                    • droftartsundefined
                      droftarts administrators @DonStauffer
                      last edited by

                      @DonStauffer said in Bed PID Tuning: How long should it take?:

                      I'm disappointed and I wonder if I can manually tune it better.

                      Insulation won't make it heat up faster, but may/should make it cool down slower. Tuning just means it should hold it's temperature more consistently. If you want to do it manually, see https://docs.duet3d.com/en/User_manual/Connecting_hardware/Heaters_tuning#setting-the-model-parameters-manually

                      Ian

                      Bed-slinger - Mini5+ WiFi/1LC | RRP Fisher v1 - D2 WiFi | Polargraph - D2 WiFi | TronXY X5S - 6HC/Roto | CNC router - 6HC | Tractus3D T1250 - D2 Eth

                      DonStaufferundefined 1 Reply Last reply Reply Quote 0
                      • DonStaufferundefined
                        DonStauffer @droftarts
                        last edited by

                        @droftarts My experience is different, and it makes sense.

                        If you're not losing much heat at all, even while you're heating the bed, it will heat up faster. My experience with aerogel insulation is that it loses very little heat. My other printer will heat up in half the time it did before the insulation.

                        droftartsundefined 1 Reply Last reply Reply Quote 0
                        • droftartsundefined
                          droftarts administrators @DonStauffer
                          last edited by

                          @DonStauffer you said the heater was a Keenovo heated pad. Those have silicone on the back, which provides pretty effective insulation. Was the heater on your other printer the same? Same size and/or power density (ie watts per square cm)? It may have suffered more heat losses if it had a lower power density, that extra insulation improved.

                          PID doesn’t kick in until the target temperature is approached, so the majority of heating from cold will be at full current. The firmware/Duet isn’t limiting this.

                          Ian

                          Bed-slinger - Mini5+ WiFi/1LC | RRP Fisher v1 - D2 WiFi | Polargraph - D2 WiFi | TronXY X5S - 6HC/Roto | CNC router - 6HC | Tractus3D T1250 - D2 Eth

                          1 Reply Last reply Reply Quote 0
                          • DonStaufferundefined
                            DonStauffer
                            last edited by DonStauffer

                            The other was also Keenovo, but 400W @24V rather than 750W @120V. This isn't subtle: the insulation on the other one doubled the heat up speed and also stabilized running temperature. The insulation on the Railcore has no effect on either, it seems.

                            It's not a huge big deal. Everything works. I just expected different behavior, and wondering if I really got a good PID tuning or not.

                            Phaedruxundefined 1 Reply Last reply Reply Quote 0
                            • Phaedruxundefined
                              Phaedrux Moderator @DonStauffer
                              last edited by

                              @DonStauffer said in Bed PID Tuning: How long should it take?:

                              400W @24V rather than 750W @120V

                              The effect would be more noticeable at 24v than 120v.

                              Z-Bot CoreXY Build | Thingiverse Profile

                              DonStaufferundefined 1 Reply Last reply Reply Quote 0
                              • DonStaufferundefined
                                DonStauffer @Phaedrux
                                last edited by

                                @Phaedrux Would it? I would think the wattage might affect it, but I don't see why voltage would.

                                1 Reply Last reply Reply Quote 0
                                • Phaedruxundefined
                                  Phaedrux Moderator
                                  last edited by

                                  Wattage is the result of voltage and current. The lower voltage heating rate would be more affected by the cooling rate of heat dissapating into the air, which is more easily overcome by the higher voltage.

                                  Z-Bot CoreXY Build | Thingiverse Profile

                                  1 Reply Last reply Reply Quote 0
                                  • First post
                                    Last post
                                  Unless otherwise noted, all forum content is licensed under CC-BY-SA