Noisy bed thermistor, what to do?

Hi there,
I have 330mm dia bed, heated by a 220v 300W silicon heater. It seems Ike the thermistor inside the bed has some issues, which causes my temp reading to jump around, causing the FW to report the heater fault and stop the printer.
The exposed thermistor cables seem to be OK, so there is nothing I can fix there.
If I connect a spare thermistor, all seems to work fine. However, I can only mount the thermistor somewhere on the edge of the bed, which will give me quite a big temp difference (up to 20-30 deg at 100C).
The original thermistor is molded into the silicone pad, much closer to the center, which gives more accurate readings.
What are my options now?
I do not think I can separate the silicon heater to exchange the thermistor, and drilling through the heater is probably a not so good idea too. Mounting the thermistor at the edge will not give me a correct temp readings...

I have a 100k "china standard" thermistor with white wires and transparent body (with a little bit of red inside), the legs are covered with teflon tubes. How much temperature the white isolation (extension cable) can withstand? Is it safe to wire it in a direct contact with a 140deg hot bed?

Seems to be a silly questions, but I decided to hear out your opinion...

Were the temp readings bad from the beginning? Maybe it's the 50Hz from the mains voltage that disturbs your sensor. A rectifier to run the bed at 220V DC would help and/or a low-pass-filter in the sensor cable.
Be aware that rectifying mains voltage leads to a higher DC voltage.

If the connection is intermittent you have no choice but to replace the thermistor. If you screw the replacement to the side of the bed there will be a temperature offset, but you only have to calibrate that once. If you have to set it to 80C to get the bed up to 110C then just set it at 80C when you want 110C.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Thermistors embedded in the silicone tend to over-read anyway, because they are too close to the heater elements and not in direct contact with the bed.

---

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].

Yes, the thermistor was jumping from the very beginning... I had the thermistor cables zip tied to the heater cable, and now they are separated, perhaps it will improve now. The bed was run in Bang-Bang mode over the SSR.
My bed thermistor was actually showing only 10C more than the bed itself...

I know one can say that there is actually not much difference if it is showing 10C plus, or 20C minus, it all can be calibrated. but it just doesn't feel right... so I decided to change the thermistor... mounting it on the side will also increase the waiting time before a print can start... perhaps it is a good excuse to get myself a 600W heater

I guess I will mount a second sensor the best I can, and remap it in the FW later if the original one will continue to jump around. Is there a way to offset the readings in Firmware?

Quote
o_lampe
Were the temp readings bad from the beginning? Maybe it's the 50Hz from the mains voltage that disturbs your sensor. A rectifier to run the bed at 220V DC would help and/or a low-pass-filter in the sensor cable.
Be aware that rectifying mains voltage leads to a higher DC voltage.

This might be a good idea, but using a rectifier will give me 310V DC roughly. The bed has an impedance of 154 Ohm, at 310VDC it will deliever 620W of power, running at 2A roughly.
Do you have any experience if this still be OK for the heater? Quite a steep step I would say... rated 320W and 1.4A, and 620W with 2A... or am I completely wrong with my calculations?

The raised current from 1.4A to 2A should not be a problem especially with good contact to the aluminum bed, but don't take me for granted.

Quote

mounting it on the side will also increase the waiting time before a print can start...

Aren't you lying to yourself, when you think so? The "remote" sensor probably shows real world temperatures better than the original.

Yes I know what you are trying to say, but it all depends on how large the print is

I think I have solved the issue. Separating the thermistor wires from the mains did not bring anything in my setup, but since I have a thermal insulation on my bed, I found a spot at about 2/3 of the bed radius and stuck the new thermistor there, just between the insulation and the silicon mat. This gives quite an accurate reading over the whole temperature range (I have an access to FLIR thermal cam). New thermistor is secured with the insulation adhesive, so it seems to have a good and sturdy contact.

I am running the auto tune with my Duet right now, the temp reading is very smooth, so I am quite happy at the moment...

Next time I am going to open the printer - I will try the rectifier mode on the bed, it seems quite a cheap way to get a 600W heater...

To gain significant extra heating power, you will need to use a mains-voltage bridge rectifier and a smoothing capacitor of at least 470uF rated at 380V or more. That's quite a large capacitor.

Bear in mind that if you increase the heating power too much, then if you get an uncontrolled heating fault, the bed heater temperature will rise to a dangerous level.

Edited 1 time(s). Last edit at 12/24/2016 12:43PM by dc42.

---

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].

Good point here.

I am actually quite happy now the way it is. Thanks everybody for the input! And happy holidays