Temperature bias?



  • Hi all,

    TLDR;
    I built my custom delta printer and it works well. However since I have it enclosed and the build chamber heats up to around 45-50°C I noticed the probe on the smart effector has become somewhat unreliable. Is that a known issue?

    So here is what happens:
    I start a print while the smart effector is "cold" (room temperature). Before each print I probe the bed and create a heightmap. First print, while the effector is more or less on room temperature the readings are fine, the bed is flat, first layer looks good.

    After a couple of hours the print finishes and the effector has heated up to whatever temperature is in the chamber, when printing ABS this is around 45-50°C.

    I remove the build surface, pop off the print, put it back in the enclosed printer and start the same gcode again.

    Unfortunately the probe now delivers very inaccurate results, it reports pretty hefty low spots of -0.15 to -0.25mm where there clearly are none. I made sure of that using a straight edge.

    I can reproduce that behaviour every time I let the printer cool down completely, that happens over night usually, and start the first print. Probing is fine and as soon as it soaked enough heat the probe readings are off the charts.

    My current solution is to not probe the bed automatically, that is no big issue from the mechanical/printing perspective since it is flat enough but that also removes the option for me to swap build plates since they have a different thickness.



  • Hi,

    I cannot say anything on the smart-effector, BUT if you have a chamber you DEFINETLY need at least 1 fan + tempsensor that sucks out the hot air on top OUTSIDE of your chamber for the same reason (of course for that you need an inlet at the bottom or side at the bottom) above a certain temperature (for each filament diffrent, define it therfor in filament-config.g)!

    Most things like the avarage-joe stepper-motor, pcb, mosfet, you name it will degrade in powers of ten for every +ten of more heat... so for having long fun with your electronics you should take care 🙂

    E.G. get an additional thermistor from the duet-website/-reseller that you screw inside your chamber at the top where it is most hot and a 80-120mm fan (do not take the small ones take a big/long one with power because you can still drive it with PWM to use less of its power) and 1 cheap dustfilter for inlet, 1 good for outlet (that the air gets a bit cleaned when it gets in and even more cleaned when it gets back into your room)

    In my case for duet2ethernet with RRF3.1.1 an example:

    Than e.g. in your config.g put with M308 a new/higher S_ then what is you have already, in my case:

    M308 S3 P"e1temp" Y"thermistor" T10000 B4274 C0.9018123000e-7 A"Chamber" 	; configure sensor Sx as type"thermistor"/"..." on pin P"..." see https://duet3d.dozuki.com/Wiki/Connecting_thermistors_or_PT1000_temperature_sensors#Section_Thermistor_parameters
    

    and define the fan later, and connect it to the sensor created before, in my case:

    ;chamber fan:
    M950 F2 C"fan2" T3 ;Q500 ;! ;Q25000                  ; create F=fan(x) with "!" to invert, on pin C; opt set frequency(Q) e.g. for most real 4-wire pwm-fans 25000 (max 65535), ordinary 2-wire must be kept low around 100-500, default seems to be 250
    ; set some default vals to start but override it in the print-file with the specific filament-config.g:
    M106 P2 C"chamber fan" B1.0 L0.9 T40:50 H3 ;X1.0          ; P=num; s=set speed 0-255/0.0-1.0; limit L min & X max; H=heater monitored+Thermostatic control, -1=off or Hx(:__...) monitored; B=blip time in ms; if not thermostatic (H-1) define a board-boot-start-up speed S=fan speed;
    

    I am currently in the state of replacing my temp-sensor with a DHT because besides temperature I can track the humidity inside the print-chamber to get to know how it influences print-quality...

    Hope something here is maybe of some use for you, just my experience here... 🙂 Good luck!


Log in to reply