Actual print bed temperature off by 25 degrees or so.

Discussion in 'Calibration, Help, and Troubleshooting' started by ndpg, Aug 18, 2016.

  1. ndpg

    ndpg Active Member

    Joined:
    Apr 20, 2016
    Messages:
    31
    Likes Received:
    23
    I've been having adhesion problems with filaments that need a specific bed temperature, ABS etc. so I used a couple of means of measuring the actual temperature at the bed surface and it's way cooler than the display indicates. Set 120 degrees and the actual surface temp is about 95. Given the relatively low thermal conductivity of glass I would expect a gradient between the heater and the glass surface but surely that would have been factored in.

    Do I simply have a bad thermistor? Should I be adding 20 degrees or so to the recommended bed temps? How critical are they anyway?
     
  2. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    You might be suffering from the pretty normal PCB heater sag. It is not touching the glass plate in the middle. Ser if you can press it up with your hand.

    I just found a nice spring or two to prop it up. Or there is a printable "adjuster" somewhere on the forum.
     
  3. ndpg

    ndpg Active Member

    Joined:
    Apr 20, 2016
    Messages:
    31
    Likes Received:
    23
    Printed three of the props but I'm still off by 25 degrees or more. I measured the temperature under the glass and when set to 100 it is actually 85. I would expect a 10 degree or so drop across the glass so to only see 75 on the surface is no surprise. Not even close to what it should be so no wonder ABS doesn't stick! The temperatures at the hot ends are pretty accurate (as is my thermometer which I checked in boiling water). When off (and cooled down) the temperature reading for both hot ends and bed are all within a degree or two. It's as though the calibration is way off. One for the support department I think.
     
  4. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    166
    The best way to measure bed temperature is with an IR thermometer (less than GBP10 from Amazon) but in the absence of that stick your thermometer to the bed with tape and tape a piece of foam/polystyrene over it.
     
  5. elmoret

    elmoret Administrator

    Joined:
    Mar 1, 2014
    Messages:
    586
    Likes Received:
    91
    IR thermometer is not appropriate for reading the bed temperature, because the glass is reflective.

    Regular thermometer on top of the glass is going to register lower temperature than the thermistor embedded in the PCB heater, this is normal just like your core body temperature can be 98.6F while the outside of your jacket is 50F.

    This "gradient" is not factored in, because that would be a guessing game depending on a few different variables.
     
  6. Michael Frampton

    Michael Frampton Well-Known Member

    Joined:
    Aug 5, 2016
    Messages:
    88
    Likes Received:
    7
    I have same issue..measured bed temp with quality IR twin laser.at indicated 115c bed was actually 85 max with 8c under in some places..been told can adjust in fw tho
     
  7. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    166
    Whoops, I have the aluminium bed! :rolleyes:
     
  8. Michael Frampton

    Michael Frampton Well-Known Member

    Joined:
    Aug 5, 2016
    Messages:
    88
    Likes Received:
    7
    Reads it fine. Consistently. Also bed is covered in pva so not shiny
     
  9. ndpg

    ndpg Active Member

    Joined:
    Apr 20, 2016
    Messages:
    31
    Likes Received:
    23
    My thermometer is a thermocouple. I hold it down on the bed with some paper towel to act as an insulator. It's also small enough to go under the glass and that temperature is also much lower than indicated. The Marlin code is either calculating or using a lookup table to convert resistance to temperature. I've no idea whether it has a range of thermistors to choose from or has to be programmed for the particular ones being used. There must be a way of adjusting the code to calibrate the measurement.
    Yes, the temperature drop across the glass will vary. It's a function of the power being generated by the heater and obviously that is less for lower temperatures and will also depend on the ambient temperature. The other parameters are the area of the bed, the thickness of the glass and its thermal conductivity but those are constants.
    At the moment I'm just setting a higher temperature to compensate for the incorrect reading. I suspect that this not uncommon but most people simply don't have the means to measure the actual temperature.
    The hot end temperatures are fairly accurate.
     
    Michael Frampton likes this.
  10. elmoret

    elmoret Administrator

    Joined:
    Mar 1, 2014
    Messages:
    586
    Likes Received:
    91
    Maybe not to visible light, but IR is a different spectrum.

    IR thermometers aren't too great in 3D printing for accurate measurements, be it the hotend (reflectivity and spot size) or the bed (reflectivity).
     
    #11 elmoret, Aug 21, 2016
    Last edited: Aug 21, 2016
  11. elmoret

    elmoret Administrator

    Joined:
    Mar 1, 2014
    Messages:
    586
    Likes Received:
    91
    It is also a function of:

    - How well the heater PCB is bonded to the glass (this depends on assembly, and whether there are spacers under the PCB, and how much the PCB is sagging.
    - The part cooling fan's airflow across the bed, which is a function of fan speed and cooling fan's position.

    It isn't something than can be effectively solved via a formula, there are too many unknowns. Best to just sort out what temperatures work best for your particular printer setup.
     
    Michael Frampton likes this.
  12. Michael Frampton

    Michael Frampton Well-Known Member

    Joined:
    Aug 5, 2016
    Messages:
    88
    Likes Received:
    7
    Is that why ir sensor can be erratic on glass when levelling ?
     
  13. ndpg

    ndpg Active Member

    Joined:
    Apr 20, 2016
    Messages:
    31
    Likes Received:
    23
    That's the plan. It means that hottest I can get the bed to is about 100ยบ but that should cover most requirements.
     
  14. elmoret

    elmoret Administrator

    Joined:
    Mar 1, 2014
    Messages:
    586
    Likes Received:
    91
    I haven't experienced that personally so its hard to say, but it is possible a varying coat of PVA would alter the sensor's readings.
     
  15. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    166
    As @elmoret says, any change to the surface can affect the reflectivity and for auto-levelling the glass should be clean; it is one of the reasons why a lot of us use the manual bed levelling (MBL) process, as this is referenced to the home position, not using the IR after that first check, and provided the home position is clean the results are more reliable.
     

Share This Page