Temp accuracy

Discussion in 'Calibration, Help, and Troubleshooting' started by Dr Jeep, May 6, 2016.

  1. Dr Jeep

    Dr Jeep Well-Known Member

    Joined:
    Sep 12, 2015
    Messages:
    529
    Likes Received:
    169
    Has anyone tested the accuracy of the PT100 and bed thermistor yet against a known good reference (comparing against a known good temp instrument or using one of those calibrated oil baths).

    I only ask because after a day of sitting in my workshop I notice on power on I have pretty interesting behaviour of the temp sensors.

    Workshop thermometer said 20C (I don't trust the accuracy of this)
    On startup my big box said 16-17C for the hot end and 30 ! for the bed

    Within a minute of startup..and with the hot end docked, the bed said 23C

    There is always an offset between the bed and the hot end when the printer is idle...I'd expect them to get pretty close (accepting the hot end responds to drafts etc faster)
     
  2. jfb

    jfb Well-Known Member

    Joined:
    Mar 30, 2016
    Messages:
    169
    Likes Received:
    35
    Not really tested them on the bigbox, but I did test my Prusa i3 a while back using a fairly cheap temp probe just out of curiousity. I found that the temps were off by about a degree or four or so over most of the range. I did find that it was more accurate at higher temperatures - that is, the error was highest at ambient, and was more or less negligible at 180 or so.

    I don't think they're guaranteed accurate - apart from anything else, much depends on the firmware's interpretation of the reading - resistance of 'x' means temperature 'y' and so on - those tables are unlikely to be completely accurate across the whole temperature range, but are probably close enough to work with at working temperature.

    Looking at both printers now, (both idle and cold) - BigBox is showing 20 on the hotends and 22 on the bed. The i3 is showing 23 on the hotend (e3d V6 with PT100) and 22 on the bed. Allowing for some temp variation across the room here, I'd say they're in fairly close agreement.

    Realistically, it doesn't really matter for anything I'm using them for, and once I get it right (or at least good enough) for a given filament it doesn't tend to change. For me, so long as I know to set 'x' for filament 'y' its all good. Judging from what I've seen on line, I'd guess this is pretty much normal and seems to be one of the reasons each printer seems to be slightly different even with the same components.

    I'm a bit surprised at your initial reading of 30 - but I'd consider the 'settled down' figure of 23 to be close enough to work with. I'll have to look at both of mine when I next start them up and see if I get the same behaviour.
     
  3. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    731
    Likes Received:
    93
    When standing in fairly still air mine read within 1oC at the moment the printer is switched on, So 22, 23, bed=22 and remain so continuously (the same) before I apply heat to any of them.

    30oC is well above ambient for most locations (but if you are in the Arizona desert maybe not) and your workshop thermometer is probably right within a few degrees C.

    You have something odd going on.

    You only have sensors, wiring and Rumba inc its firmware involved, is there anything you can swap round like spare connections on the Rumba to use as a test?

    Mine are still at 22, 23, 22 after ten minutes.
     
    #3 Old_Tafr, May 7, 2016
    Last edited: May 7, 2016
  4. jfb

    jfb Well-Known Member

    Joined:
    Mar 30, 2016
    Messages:
    169
    Likes Received:
    35
    Both of mine are still being very consistent this morning - hotends reading 19, bed 21 on both BigBox and i3. I didn't notice any fluctuations on startup either, so I'd agree there seems to be something odd going on with that 30 reading.
     
  5. Dr Jeep

    Dr Jeep Well-Known Member

    Joined:
    Sep 12, 2015
    Messages:
    529
    Likes Received:
    169
    hmm yes I'll look into it I think. The printer is busy at the moment but next time I do a cold start I think I'll take some readings straight off the sensors and log them.
     

Share This Page