Large bed heating issue

[djeZo888]

There is no documentation regarding max size (or resistance) of the heated bed.

I am currently replacing stock motherboard in Kobra Max with BTT SKR Mini E3 V3. I see that Kobra Max has additional board (NF013 Heating Platform) that contains mosfet WSK220N04 which is used for heating bed. See on the left in the attached picture.

I assume this extra board is used to shift mosfet load from main board (Trigorilla Pro A v1.0.4). In this extra board, there is something more going on - I did not reverse engineer all components - before reaching gate of WSK220N04, I see that MB6S (bridge rectifier) is used - probably to solve issue of possible reverse polarity connection? Or perhaps to utilise diode forward voltage drop and slightly decrease voltage for heating bed? There are few more components... which I don't need to reverse, because I would like to get rid of this extra board. I suspect that this is the culprit in my issues of being unable to perform consistent PID tuning of the bed. Every time I did PID tuning of the bed, I got different results. Consequently, my prints suffered Z banding.

I am not sure if I can connect bed heater directly to BTT SKR Mini E3 V3. The resistance of the bed heater is 1.4 Ohms at rooms temperature. With 24 V that means around 17 A of current. Can SKR Mini take this? I see that there is main fuse on board rated at 20 A. Yes, the bed would be heated using PWM, so the average current would be lower. SKR Mini uses same mosfet for bed heating as Trigorilla - WSK220N04. If I can heat the bed directly with SKR Mini, then why did Anycubic engineers use this extra board? Trigorilla has main fuse 25 A.

Because of all this, I am thinking of using own mosfet (or several in parallel) on special board. I can spread the load, use TH components and mount heatsinks on mosfets.

Let me know your input. Thanks!

EDIT: Thinking some more about it - extra board has power inductor sitting right after 24 V DC input. That is for filtering right? Perhaps this filtering is messing with consistency of PID tuning? It would make sense...

[djeZo888]

Okay, there are modules such as this: https://www.aliexpress.com/item/1005002117470689.html being sold for large beds.

I am wondering why there are som extra components besides power mosfet on this board. What is the purpose of these components.

Anyway, I have created my own solution using IRLP3034 mosfet. Why this one? It is better than WSK220N04 in all important aspects (RDSon, lower gate charge), it is TO-247 package for easier cooling and it is what I had in stock. Mounted large heatsink to it.

Will let you know how it works when I try it.

[djeZo888]

Found some more (including schematic of this extra board): https://3dprinting.stackexchange.com/questions/8553/will-this-mosfet-allow-the-heat-bed-to-run-at-a-different-voltage-than-the-contr

Looks like what is happening on this extra board is power mosfet being driven through optocoupler - for isolation. So, bed could be heated through another PSU, using higher voltage for example. Now, this thing could really mess with input PWM signal, if PWM signal is fast and optocoupler isn't fast enough. This could explain problems with bed PID tuning.