Temperature readings from thermistors built in to silicone heaters

I've just commissioned my short fat Kossel (a Mini Kossel with the printable bed diameter increased to 300mm). An issue I am seeing is that when the indicated bed temperature is 120C, the temperature read by a thermocouple taped to the top of the glass is only 82C in the centre of the bed, and 60C nearer the edges.

The bed comprises a 330mm diameter 5mm thick aluminium plate, with a 350W silicone heater stuck to the underside and a 4mm thick sheet of glass on top. I know that glass is a poor thermal conductor and I am used to seeing about 10C drop between the thermistor reading and the temperature on top of 3mm glass; but this time it is nearly 40C at the centre and 60C near the edge of the printable area. Has anyone experienced a similar problem? I believe the 100K thermistor built in to the heater has a B value of 3950, and that is the value I configured in the firmware. The temperature reads correctly when the bed is cold. If I hold the thermocouple against the edge of the aluminium plate, it reads 92C.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hi,

Where is the heated bed's thermistor located? Between silicon pad and aluminium sheet or between aluminium sheet and glass?

If it's the former then the 40C or so difference you see in the centre when measuring at the glass surface is believable. I have seen a 15C temperature difference across 2mm thick glass.

As for the edge drop-off, it seems likely to me that the heating elements can't extend to the very edge of the pad due to physical constraints imposed by sealing the edges.

You might have to live with introducing a temperature offset factor into your heated bed setpoint.

Regards,
Neil Darlow

---

I try to write with consideration for all nationalities. Please let me know if something is unclear.
Printing with Mendel90 from fedora 25 using Cura, FreeCAD, MeshLab, OpenSCAD, Skeinforge and Slic3r tools.

Quote
neildarlow
Where is the heated bed's thermistor located? Between silicon pad and aluminium sheet or between aluminium sheet and glass?

Neither, it's buried within the silicone. I am using a 300mm diameter heater with a 330mm alu plate to allow space for bed clips and mounting screws, so some drop off in temperature towards the edge of the 300mm printable area is understandable.

I find that getting good temperature readings on top of an insulating surface like glass is difficult. The glass doesn't conduct heat to the thermocouple well, and the thermocouple wires conduct heat away from the bead.

I finally got good ABS prints by commanding the bed to 140C. This gave a reading on the thermocouple of around 100C on top of the glass, and 120C between the bottom of the heater and the cardboard insulation underneath the heater. So I think part of the problem is that the thermistor is seeing something close to the element temperature, which is about 20C more than the temperature of the alu plate, and I am losing another 20C across the 4mm glass.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

Quote
neildarlow
Where is the heated bed's thermistor located? Between silicon pad and aluminium sheet or between aluminium sheet and glass?

Neither, it's buried within the silicone. I am using a 300mm diameter heater with a 330mm alu plate to allow space for bed clips and mounting screws, so some drop off in temperature towards the edge of the 300mm printable area is understandable.

I find that getting good temperature readings on top of an insulating surface like glass is difficult. The glass doesn't conduct heat to the thermocouple well, and the thermocouple wires conduct heat away from the bead.

I finally got good ABS prints by commanding the bed to 140C. This gave a reading on the thermocouple of around 100C on top of the glass, and 120C between the bottom of the heater and the cardboard insulation underneath the heater. So I think part of the problem is that the thermistor is seeing something close to the element temperature, which is about 20C more than the temperature of the alu plate, and I am losing another 20C across the 4mm glass.

Dave

I am seeing very similar effects on mine still to check on the top of the Glass yet but if you remember I was getting a 20deg difference between the Thermistor reading and a thermocouple on top of the Ally plate wonder actually if 5mm thick Ally might be to thick tho once at temp it does seem to hold it ok.

Doug

I just did the sums. If there were 100W of heat passing through the 150mm radius x 5mm plate, and it were pure aluminium with a thermal conductivity of more than 200 W·m−1·K−1 (from [en.wikipedia.org]), then the temperature drop across it would be:

T = (100 * 0.05)/(pi * 0.15^2 * 200)

which is about 0.35C. The aluminium is probably an alloy with a lower thermal conductivity than 200, OTOH there may be less than 100W actually passing through the plate. So I think the 20C drop is likely occurring between wherever the thermistor is inside the heater, and the bottom of the aluminium.

For glass, the thermal conductivity is about 0.85. If there really was 100W passing through it, the temperature drop across 4mm glass would be around 65C.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Dave

I suspect that the Thermistor is actually under the element ie the element is between the Thermistor and the Plate and this is what is causing the issue.

I had thought about milling a 1mm deep x say 2mm wide groove in the top of the plate and fitting a semitec thermistor between the ally and the Glass will need to find a machins shop that will do this for me or get the CNC machine up and running.

Doug

I find the best way to measure the temperature of bed glass is with an IR thermometer. Glass has good emissivity, unlike metals.

Thermocouples never read accurately pressed against metal. They need to be buried into it or at least well insulated on the side that is not touching. For that reason I always check hot end temperatures with a thermocouple placed inside the barrel. It then agrees with the thermistor buried in the heat block to within a degree or so. When pressed on the outside it can read 30C low.

I don't think there will actually be a lot of drop from one side to the other of 5mm aluminium but when you press a thermistor or thermocouple against it there is not much contact area and a much larger surface area cooled by convection. Covering it with ceramic cloth makes it more accurate.

---

[www.hydraraptor.blogspot.com]

I did wonder whether there might be a spot in the centre of the heater free from wires, where a thermistor hole could be made. Would probably need an X-ray machine to find out though. Another possibility is thermistor attached to the aluminium plate just outside the periphery of the heater. That would probably cause some overshoot, which could be controlled by implementing slow PID for the bed heater.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].