SOLVED PT100 Temp fluctuation

Discussion in 'Calibration, Help, and Troubleshooting' started by Nicu Damse, Apr 29, 2016.

  1. Nicu Damse

    Nicu Damse Member

    Joined:
    Sep 5, 2015
    Messages:
    8
    Likes Received:
    0
    I am having some questions about the PT100 temp sensor and maybe somebody here has some answers.

    While the hot end is heating up i have very big temperature fluctuation. Sometimes 10 degree Celsius from one measurement to the next. It jumps from 190 to 180 and back to 191 in 2 seconds. All while the hotend is at max power.
    Similar behavior while cooling down. I noticed this because at 40 degree the fan turns on. When it gets around that temp the fan starts and stops spinning at least 5-6 times during every cool down and the temp fluctuates at least 5 degrees.

    So far in my other printers i only use the old school 100k thermistor (both in a E3D V6 and some other hotend)
    and it is much more stable. It might jump around 1 degree but even that is rare.

    Also i have the profile in S3D setup to wait for the temp to stabilize and it takes quite a bit longer then on a normal thermistor setup. It jumps around like crazy. I did autotune the hotend and i have 2 other E3D hotends so i don't thing it's a tuning issue.

    I don't understand if this is normal behavior or if my setup is faulty.

    I only printer with one hotend atm even thou i have the dual but i noticed similar behavior on the second hotend from time to time even at room temperature although the fluctuations are smaller (3-5 degrees).

    From what i understood the reason for having the PT100 is to get a higher max temperature but also to get more reliable readings. So far i am not impressed...

    Maybe i did something wrong. I just don't know what.

    Any input would be appreciated. Thanks.
     
  2. R Design

    R Design Well-Known Member

    Joined:
    Oct 7, 2015
    Messages:
    918
    Likes Received:
    183
    Mine is much more stable than that. In fact, it's basically stable.

    Very occasionally it can "flash" a rather different temperature (say 5 degees different) but then it returns to being stable.

    Thermocouples generate a very small voltage which is then amplified (and perhaps digitised?) at the dedicated board.

    I wonder if there could be an electrical issue (weak contact) or some electrical interference causing the perceived voltage to fluctuate?
     
  3. Nicu Damse

    Nicu Damse Member

    Joined:
    Sep 5, 2015
    Messages:
    8
    Likes Received:
    0
    So i have to come back to my initial problem.
    Here is a screenshot of the very weird behavior:
    PT100.jpg
    Every fan of the printer is turned off. No stepper is moving. So electrical interference is very unlikely. Also it's dark outside so no sun and no temp fluctuation in the room.
    Also as you can see it's somewhat stable when cooling down then it goes crazy then it's getting somewhat stable again. The spikes are about 9-10 degrees.
    Any suggestions ? Is this normal behavior with the PT100 system ? I didn't start using the second hot end yet but it behaves similar even in idle.

    As i said it's the first time i use a PT100 sensor but it's performance on my printer is much worse then any thermistor i have on my other printers. I am nowhere near stable temp's while printing and unfortunately it's not PID tuning...

    I really hope i can get to the bottom of this problem as i expected much better from this printer.
     
  4. Nicu Damse

    Nicu Damse Member

    Joined:
    Sep 5, 2015
    Messages:
    8
    Likes Received:
    0
    Here is the same problem while printing.
    PT100 HOT.png
    I find nothing that can explain this behavior...
     
  5. Mike Kelly

    Mike Kelly Volunteer

    Joined:
    Mar 7, 2014
    Messages:
    444
    Likes Received:
    89
    I think this has to do with the amplifier board. It might be faulty and/or dying. If it continues to happen be sure to contact E3D support for a replacement.
     
  6. jet

    jet Well-Known Member

    Joined:
    Mar 29, 2016
    Messages:
    199
    Likes Received:
    20
    I'm also wondering how much fluctuation is too much. I'm also doing the "wait for the temp to stabilize" option in S3D and 5-10 minutes will pass by and the print still does not start. A couple of times I've had to power cycle to get it out of "wait" mode.

    temp.jpg
     
  7. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    166
    Have you done the Auto PID routine yet, as that made a lot of difference to my waiting times.
     
  8. jet

    jet Well-Known Member

    Joined:
    Mar 29, 2016
    Messages:
    199
    Likes Received:
    20
    No, I didn't realize Auto PID was still an option until reading this thread.
     
  9. Henry feldman

    Henry feldman Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    969
    Likes Received:
    389
    yeah, makes that totally go away...
     
  10. jet

    jet Well-Known Member

    Joined:
    Mar 29, 2016
    Messages:
    199
    Likes Received:
    20
    [comedic forehead slap] We're running Marlin. Sigh.
     
  11. Stefan

    Stefan Well-Known Member

    Joined:
    Feb 17, 2016
    Messages:
    323
    Likes Received:
    43
    Auto PID tuning is a must I think, but I had an issue which was caused by my diy pulg able cabling.
    the idle temperature (room temp about 20°) was increasing day after day up the 57° then I reconnected the my plug and it went down to ~20° as the 2nd hot end in idle. --> bad connection in the plug.

    But since the problem of @Nicu Damse does happen too while cooling down I think PID tuning will not help. Maybe reconnect all PT100 plugs may help.
     
  12. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    560
    Likes Received:
    75
    It is unlikely that the actual temperature changes by 10oC in a few seconds or less, even if the heatsink fan stopped and the heater was full on, nor change back again (as it's a spike in temp?) in the same few seconds. My heatsink fan does cycle on and off when it gets down to around 39oC as I assume heat soak (with the fan off and heater off) warms the sensor slightly back to 40oC enough to start the fan for a moment (I would say momentarily..........but that has a different meaning in American English)

    Also from the graphs the temperature maintains its level and the spikes go away so both heater and temp sensor (PT100) must then be working again.

    This would point to an intermittent problem (obvious I know) so what could this intermittent problem be? Unless there is a fault on the Rumba board itself then either a problem with the PT100 board or the sensor itself or a bad contact in the wiring/break in the wiring between the board and the heat block.

    I would start with the connections at the PT100 board itself as it's more likely a problem with measuring than with heating.

    As I think you are saying you have a dual (temp varies normally a few degrees on the other extruder) then try swapping just one part at a time to see what changes.

    If the problem is only during printing? then maybe a loose connection exacerbated by the printer and X-Carriage movement.
     
  13. Henry feldman

    Henry feldman Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    969
    Likes Received:
    389
    Mine looked exactly like this prior to PID Autotune. After that, it was razor flat... Not sure why this isn't simply part of commissioning since it seems practically mandatory for proper operation.
     
  14. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    560
    Likes Received:
    75
    So what you are saying is that your temperature graph looked like the graph in the original post prior to the PID Autotune?

    If you accept that in reality the temperature could not fluctuate so quickly then what do you put the rapid variations down to prior to the tuning?

    PS The PT100 temp sensor looks like it has shielded cable as under the clear insulation it looks like a braid. If it needs to be shielded to prevent interference then cutting the cable to add connectors and then not making a (3rd) connection to make the braid (sort of) continuous again will expose the sensor wiring to electrical interference. I have not cut mine yet and I can't remember how the wiring emerges and how many wires there are (looks like two wires and the braid are connected to the PT100 board, this would make sense, two wires for the sensor and braid grounded at one end only i.e. on the PT100 board)

    If the heating uses PWM? (Pulse width modulation) this uses a system powering the heater full on or full off, this (have a look at WiKipedia) is a square wave which has by its very nature of a fast rise time which has high frequency harmonics, these will affect any unshielded wiring, like wiring cut to add connectors and quite likely add spikes. Every time the heater switches on there will be a burst of interference.

    It would be interesting to know if it is only people with the wiring cut to add connectors who get the spikes, although a bad ground connection at the PT100 board could have the same effect.

    What the above would not explain is why the spikes appear when cooling when the heater power is off, but the interference could come from elsewhere, like the power supply. Do we know if the power supply is a switching mode type as this uses a high frequency oscillator.
     
    #14 Old_Tafr, Jun 24, 2016
    Last edited: Jun 24, 2016
  15. Henry feldman

    Henry feldman Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    969
    Likes Received:
    389
    Except PID tuning would have no effect on RFI. I assume it calibrates response to the thermocouple better or whatever.
     
  16. fpex

    fpex Well-Known Member

    Joined:
    Oct 18, 2015
    Messages:
    621
    Likes Received:
    58
    Where and how is this autoPID?
    I am experiencing extrusion problems that seems related to the temperature of the extruder more than anything else
     
  17. Stefan

    Stefan Well-Known Member

    Joined:
    Feb 17, 2016
    Messages:
    323
    Likes Received:
    43
  18. Henry feldman

    Henry feldman Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    969
    Likes Received:
    389
  19. tohara

    tohara Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    128
    Likes Received:
    59
  20. Dr Jeep

    Dr Jeep Well-Known Member

    Joined:
    Sep 12, 2015
    Messages:
    530
    Likes Received:
    169
    I have the same issue, it seems to be intermittent and somewhat random and also happens sometimes during cooldown (which leads me to think it can't be a PID tuning issue). Wiring checked at every point between the PT100, AMP board and AMP board to rumba.

    When I find the proprietary opto lead for my meter I will graph the voltages and then if I can identify the culprit, contact E3D support.
     

Share This Page