Actual heatbed/Extruder temperature vs reported

Hi,

I’m new to this 3D printing and so far have performed 5 prints, of which 1 successfully printed. The rest came away from the bed.
One of the issues I have noticed is that the PCB heat bed does not sit 100% flat, in fact it is raised in the center, so adjusting the heights of the four corners to a 0.25mm gap all looks good until the head is moved to the center, here, it is hard up against the bed and in actual fact depresses the PCB print bed.
So based on various bits of reading, a glass print bed is the best solution for a perfectly flat bed that can be aligned properly.
Whilst I have been reading about these glass print beds, thoughts went through my mind on how much variation in temperature there is between the underside of the PCB bed where the thermistor is and the top side of the PCB where I have trialled tape (currently removed).

Heat Bed
So I decided to take a few measurements using a K thermocouple with the software setting the Heat bed to 60°C:

  1. Reported temperature of the thermistor via the software = 60°C (This you would expect)
  2. Temperature of the thermistor using the K thermocouple = 58°C (Placed the thermocouple onto 1 leg of the thermistor)
  3. Temperature of the top side of the PCB bed = 56°C
  4. Temperature of the top side of a sheet of mirrored glass placed on top of the PCB bed (no heatsink compound) = 53°C
  5. Temperature of the top side of a sheet of mirrored glass placed on top of the PCB bed (with heatsink compound) = Unknown (Need to bye some thermal paste)

Extruder Nozzle
Nozzle temperature in the software set to 190°C:

  1. Reported temperature of the thermistor via the software = 190°C (This you would expect)
  2. Temperature of the Nozzle using the K thermocouple = 215°C

Note - As the temperature of the Nozzle increased, the K thermocouple was >30°C in front of the computer. Controller built in Delay?

Measuring detail-

  1. K Thermocouples have an accuracy of typically +/- 2°C.
  2. Before accepting the temperature reading, I waited for a <2K variation over a 5min period, to ensure the temperature was reasonably accurate.
  3. For the PCB bed, I used a thermocouple embedded into a aluminum 10mm10mm40mm block to ensure a stable temperature on the thermocouple.
  4. For the Nozzle, I placed the K thermocouple directly into the same hole as the thermistor.

When I get some thermal paste, I will update this post.

Has anyone else trialled doing this?

Kind regards,
Neil.

Hi NW27, my bed with glass reads about 10 deg low (but the thermister read 140k when fitted).
The heater has been turned upside down & a small notch taken out of the back of the glass to accomadate the thermister (to improve heat transfer & the accuracy of the sensor). I wouldn’t bother with thermal paste. If anything get a 24V power supply & relay.
The bottom line is to test & find out what works for you, then stick to it. I suspect that there is quite a variation between machines.

For the extruder do something like this youtube.com/watch?v=FSOPsRiiOZk
I do this for each reel of filament, even if it comes from the same supplier & note the results on the side of the spool.

For the bed, again it’s trial & error. Drop the tremerature untill prints dont stick the go back up a little.

You don’t need to use hairspray or anything like that, PLA will stick to glass very well if:
The temeratures are correct & the glass is clean.

If I get a print that fails, it tends to be the first of the day. The second attempt works fine. I have just accepted that it wastes less time & fillament to give the bed & surrounding components time to get to a stable temperature.

I find that after about 20 hours of printing & cleaning the glass with solvent that the prints dont stick even with the same settings. When you get to this point take off the glass & wash it with hot water & a little washing up liquid, rince well dry then hold up to the light.
Decide that it’s not clean & repeat 2 or 3 times. Return to your printer, print & think why didn’t I do this 5 hours ago.

Happy printing.