HT Plate Temperature vs Print Surface Temperature

Discussion in 'Tool heads & ToolChanger' started by Spoon Unit, Sep 17, 2019.

  1. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    One of the things I've had for a while in my startup script is a wait step to try to allow the print surface to get up to temperature. This is distinct from the HT Plate, which gets to temperature shockingly quick with the HT bed. Given that I ordered PT100s with the Toolchanger, I had a bunch of spare thermistors lying around so I thought I'd get some data on the performance of my bed (YMMV). I have the PEI-coated Alu plate from Clever 3D in Germany. After a quick test with sellotape, I decided on a bit of crushed Alu foil together with the swiss clip to wedge one of the thermistors on the side of the bed.

    [​IMG]

    Happy to take advise on how to improve the placement of this, but I think the results are useful even if the measurement isn't.

    The PT100s came with nice long connecting leads, so I've trailed this (not entirely perfectly), under the bed and through the hole at the top of the drag-chain mounting. Ideally I'd get it fully into the drag chain.

    [​IMG]

    From here, I'm plugged into the Duex E5 Thermistor connector. The following setup in config.g then reports the temp of the thermistor as 'Chamber' P0 on the DWC:

    Code:
                            ; Actual bed temp thermistor
    M305 S"P0" P6 R4700 T100000 B4388             ; Set thermistor
    M141 H6 P0
    Which then looks like this:

    [​IMG]

    I then ran a number of tests to see, for a given heat plate setting, what the thermistor would eventually settle on, and then graphed the data:

    [​IMG]

    Here's the actual data:

    [​IMG]

    This then gives me a tool that's a little more accurate in order to await bed temp soaking rather than assuming a perfect match between the HT reading and the actual bed. You can see there's quite some lag between the point at which the HT bed reports reaching temp and the point at which the thermistor on the bed reaches it's steady state:

    [​IMG]

    The blue line is the HT bed report, and the black line is the thermistor. So, for a bed set to 115, the HT Bed will report reaching temp in just over 2 mins, but the thermistor sat on the edge of the bed won't report a steady state for about 8 mins, and will land on around 99 ultimately.

    To use this as a tool for starting a print, we can now use M191 to await a chamber temp, but this doesn't work well without first PID tuning the thermistor, which itself is a bit imperfect because it's not directly connected to a heater. In any case, I awaited a cold bed, then set the bed to 60 and immediately ran a PID tuning on H6 to 65 degrees. As soon as the thermistor reached 65, I turned off the bed, and finally got this result saved to config_override, which you could probably just use straight away.

    M307 H6 A150.5 C888.6 D4.0 S1.00 V24.2 B0

    Finally, I can now have a start script that waits for the bed temp to soak in and reach close to a steady state rather than wait for the HT bed to report temp.

    So, if I want the bed set to 80, I can use this now at the start of the print:
    ; Warm bed
    Code:
    ; Warm bed
    M140 S80
    ; Wait for bed thermistor (chamber P0) to report reasonable temp
    M191 P0 S65
     
  2. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    I also did measurements of the Clever3D PEI plate with an infrared thermometer and compared them with the readings you got from the thermistor. Had similar temp offsets, the temperature of the bed thermistor overreads by quite a bit, especially around the usual working temperatures. Hence I had quite some trouble getting some filaments to stick at first, the plate simply was not hot enough, even when leaving it for a while. I messed about with the thermistor setup in config.g and it has gotten a tad better, but nowhere near as accurate as I'd like it to be. It would be damn awesome if there was an option to have a better temp sensor in the high power heatbed and not having to resort to such measures. The bed is pretty expensive, might as well slap some PT100 in there (and in the process get rid of the leftover PT100 amplifier boards that must be laying around in some E3D storage facility).
     
    #2 mhe, Oct 6, 2019
    Last edited: Oct 25, 2019
  3. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    On the plus side though, I can happily send a print request now and get a model printed with zero intervention required if I'm not changing filament. As long as I know it's worked before, I know the pause for temperature via the thermistor/chamber control will be sufficient to get a good first layer and a successful print.
     
  4. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    Yes, I am pretty much in the same ballpark by now, it is becoming almost as reliable as my Bigbox was, I believe once I can put some Hermes on it, it will surpass the Bigbox.
     
    Spoon Unit likes this.
  5. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    Right. The bowden-fed tools are never going to give the same control as a more direct tool. Not clear to me why Hermes over Aero or Titan?
     
  6. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    I had Titans, I had the Titan Aero and I am pretty sure the Hermes will be superior to both. It has a shorter filament path, filament is restricted by 2 powered rolls instead of 1 active and 1 passive (should be awesome for flexibles) and probably be lighter than an Aero. Also it should be cheaper than the Aero according to Sanjay. The fact that it also has a lot less parts might make it even lighter than the Aero.

    Somebody already did the Aeros on the TC: https://www.thingiverse.com/thing:3856096
     
    #6 mhe, Oct 9, 2019
    Last edited: Oct 9, 2019
  7. blarbles

    blarbles Well-Known Member

    Joined:
    Aug 10, 2019
    Messages:
    108
    Likes Received:
    68
    They reason I did not go with this and instead made my almost-direct-drive solution was the part cooling solution for the Aeros relies on you figuring out your own solution to push the air through the tubes. I do not know the effectiveness of just a tube blowing air rather than a wider air distribution solution.
     
    mhe likes this.
  8. Jai Stanley

    Jai Stanley Well-Known Member

    Joined:
    Aug 28, 2019
    Messages:
    115
    Likes Received:
    60
    Thanks for posting your findings. Made me crack out the IR thermometer and do a quick check for all my typical print temperatures. Changed a few material profile settings. My last printer had PEI, then 3M adhesive sheet, then MIC6 milled plate, then 3M adhesive, then Keenovo heater. Of course this gives a different profile and hadn;t condisered it...
     
  9. Winand

    Winand Member

    Joined:
    Jan 6, 2016
    Messages:
    12
    Likes Received:
    0
    I also experience an offset between the measured temp on the heatbed and that what I can measure with an IR laser thermometer.
    This is with the glass bed and I too had some issues with filaments not sticking when I was starting out. It was a nightmare with a HIPS print until I figured it out.
     
  10. dc42

    dc42 Well-Known Member

    Joined:
    Aug 16, 2016
    Messages:
    526
    Likes Received:
    221
    It's normal for a glass bed to have a surface temperature about 5C lower than the bed plate reading at around 60C, and about 10C lower at 100C. Also, when the thermistor is embedded in the heater rather than being attached directly to the bed plate, it will read higher than the bed plate temperature, especially when it has only just reached the set temperature.

    I have a delta printer with a 5mm thick bed plate, mains powered silicone heater, thermistor embedded in the heater, and a 4mm thick glass bed. Instead of worrying about the actual reading, I just use a setting that works well, which is 70C for PLA and 80C for PETG. Also I don't print as soon as the bed temp reading is stable, I run delta calibration first, which takes about 30 seconds and so gives the heat some extra time to percolate through the aluminium and glass.

    You can get a more accurate reading by drilling into the side of the bed plate (if it is thick enough) and installing a thermistor in the hole; but is it worth it?
     
    mhe likes this.
  11. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    True, but still the same would apply to my previous BigBox where I used the same Alu/PEI bed. And the difference wasn't as big there, maybe because it had the old heatbed which would heat substantially slower and therefore the heat would distribute better by the time the set temperature was reached?
     
  12. Winand

    Winand Member

    Joined:
    Jan 6, 2016
    Messages:
    12
    Likes Received:
    0
    I too had minimal difference (few degrees) with just glass on my Bigbox, but with the new heatbed the offset is around 10 degrees. This is also after heating up and letting everything come to temperature, easy to account for once you know, but it took some time to figure it out!
     

Share This Page