Bed minimum heating time.

Dkos

Hello everyone,

While in process of fixing some other quality issues, i upgraded my Z axis to a dual lead screw setup driven by 1 motor with belts and pulleys/idlers. During this upgrade i purchased a 5mm thick cast aluminum plate with a silicon heater that i attached on the under side.

My concern is that the heater is quite powerful apparently, and since the thermistor is included in the heater by the time the set temperature is reached the top surface of the aluminum is barely warm.

At the moment, i manually set the bed to be heated beforehand and wait a certain amount of time before starting the print. Is it possible to do that with the Duet? Or any other suggestion.

Thanks in advance!

deckingman

@Dkos What I did with mine was drill a hole in the edge of the aluminium, as deep as possible and as close to the top surface as possible, and installed the thermistor in there. As you have observed, the temperature at the junction between the heater and the aluminium rises very rapidly so if the sensor is installed at that point, it will sense the temperature rise, and when it reaches the set point, will turn the heater off. Then it has to wait for the temperature to drop before it turns it back on. This exacerbates the issue the it time time for the heat to conduct through from the bottom of the aluminium to the top. So moving the sensor away from the heater/plate junction and closer to the top edge of the plate helps enormously.

If you are unable or unwilling to do that, then you could use G4 Snnn to your slicer start gcode to instigate a dwell period.

jay_s_uk

Would it be possible to have 2 thermistors? One embedded in the silicone of the heated bed and one attached to the aluminium platform. Then, the one embedded in the silicone heater controls the heater, but printing doesn't start until the temperature has converged (with a couple of degrees) with the one attached to the aluminium.

Maybe this would be possible using the new conditional gcode?

Dkos

@deckingman Thank you for the reply. The drilling seems like the appropriate solution for a closer to reality reading. I will have to wait until a buy a new thermistor, connect it to the Duet and just abandon the embedded one. Meaning the silicon heater has to be detached therefore a 3M adhesive sticker is in order as well. In the meantime i will go with G4, thanks!

@jay_s_uk i would be interested in this one as well.

deckingman

@jay_s_uk said in Bed minimum heating time.:

Would it be possible to have 2 thermistors? One embedded in the silicone of the heated bed and one attached to the aluminium platform. Then, the one embedded in the silicone heater controls the heater, but printing doesn't start until the temperature has converged (with a couple of degrees) with the one attached to the aluminium.

Maybe this would be possible using the new conditional gcode?

I guess that might be possible but it doesn't really address the problem of the heater turning off too soon because of the proximity of the thermistor to the hot junction, rather than the upper surface of the plate. If you attach a second temperature sensor to the aluminium plate, why not just use that to control the heater?

jay_s_uk

@deckingman

If its a thick piece of aluminium, wouldn't using the second thermistor run the risk of the silicone heater being a lot hotter than you actually want?

Dkos

Considering this approaches a bit more, i am wondering on the impact of the delay between the temp change in the heater and the temperature change at the top of the plate. I guess since it is working in an existing setup, this delay is not a problem for an accurate PID control of the bed temps?

deckingman

@jay_s_uk said in Bed minimum heating time.:

@deckingman

If its a thick piece of aluminium, wouldn't using the second thermistor run the risk of the silicone heater being a lot hotter than you actually want?

In theory that might be a problem - in practice my 10mm thick aluminium plate with 800 Watt mains powered heater has been running like that for several years without any issues. Having said that, I do also have semi-rigid insulation under the heater, so if the adhesive did let go, the heater would remain in place.

jay_s_uk

@deckingman

I have 10mm with a 2000w heater underneath so that's more of a concern for me.
I need to get my head round conditional gcode so I think I will add this to my list to investigate

deckingman

@Dkos said in Bed minimum heating time.:

Considering this approaches a bit more, i am wondering on the impact of the delay between the temp change in the heater and the temperature change at the top of the plate. I guess since it is working in an existing setup, this delay is not a problem for an accurate PID control of the bed temps?

It works well for me (and has been doing for a number of years) is all I can say. Once the plate is up to temperature, it has a lot of thermal inertia and I don't every see any oscillations in the temperature.

Dkos

@deckingman Ok, thats great then. Thanks for the feedback.

deckingman

@jay_s_uk said in Bed minimum heating time.:

@deckingman

I have 10mm with a 2000w heater underneath so that's more of a concern for me.
I need to get my head round conditional gcode so I think I will add this to my list to investigate

What size is your bed? Mine is 400mm x 400mm (x10mm thick) and my 800Watt is a bit over powered and could reach 270 deg C if the Mosfet failed and applied full power. So 2000 watt might be a bit too aggressive unless your bed is really big.

deckingman

@Dkos said in Bed minimum heating time.:

@deckingman Ok, thats great then. Thanks for the feedback.

BTW it's Important - that you tune the heater PID if you move the temperature sensor.

jay_s_uk

@deckingman said in Bed minimum heating time.:

@jay_s_uk said in Bed minimum heating time.:

@deckingman

I have 10mm with a 2000w heater underneath so that's more of a concern for me.
I need to get my head round conditional gcode so I think I will add this to my list to investigate

What size is your bed? Mine is 400mm x 400mm (x10mm thick) and my 800Watt is a bit over powered and could reach 270 deg C if the Mosfet failed and applied full power. So 2000 watt might be a bit too aggressive unless your bed is really big.

520 x 520. Apparently it can reach 500 degrees if left on.
It was cheaper to buy the one I got than one from keenovo

Danal

The most important thing for safety is a thermal fuse in the wiring to the heater, attached to the heater.

jay_s_uk

@Danal I have 2. One on each cable

Dkos

@deckingman Yes of course, thanks for the notice.

Phaedrux

As long as the heating pad has a good thermal connection to the aluminum plate the plate should be acting as a sufficient heatsink to keep the heater from overheating even with the thermal lag time between the heater and the thermistor in the plate.

Air gaps between the pad and aluminum could cause potentially dangerous early failure from a coil burning out. And maybe that risk would increase if the heater was being driven harder without a tighter feedback control loop. But we'd need a thermistor on both the heater and the plate to really see what the differential is.

The PID tuning results would likely be more effective for stable bed temps using the plate thermistor and likely allow for a much faster heat up phase.

It seems like this would be a great use of conditional gcode.

dc42

You could use a thermistor embedded in the plate for temperature control, and configure the one in the heater as overheat protection using M143.

Dkos

@deckingman Followed your approach, and I second that it works much better. I drilled a 4mm deep hole with 2.5mm diameter and inserted the thermistor there, leaving 1mm before the top surface. Re-tuned the PID and I am very happy with the result.

Thanks for the tip!