search

makerbase-mks / MKS-Robin-E3-E3D

MKS Robin E3 E3D 32Bit Control Board 3D Printer parts with tmc2209 Uart mode driver For Creality Ender 3 CR-10
https://www.aliexpress.com/item/4000781744682.html
GNU General Public License v3.0
117 stars 99 forks source link

Heatbed FET overheats! - PCB design error - Output FETs GATE voltage only getting 3V3. #166

Open JazzinSA opened 2 years ago

JazzinSA commented 2 years ago

According to the SCH the CPU is powered with 3.3V This means that the FET GATE that drives the heat bed can get a maximum of 3V3. Accorder to the datasheet of the FET - HY1403D - the GATE needs about 4.5V in order for the Drain-Source resistance to be less than 20milli Ohm. Thus if the GATE only gets 3.3V the D-S resistance is I estimate at about 30milli Ohm. Therefore the FET is dissipating almost 6W which is at its absolute limit. This is with a standard 250Watt heat bed.

Have anyone experienced that the FET cannot handle a heat bed at 24V? What is the heat bed current capability of the board since its not specified anyware?

lfreitas34 commented 2 years ago

I have a E3D v1.1, have not connected the heatbed yet, still building the printer. But the heatbed FET should be a different model, here it is marked as G024N03, and the extruder FET is marked as HY1403.

If you have both FETs marked as the same model, perhaps you got a board that was assembled incorrectly?

In the hardware description on the user manual on this repository it says:

HOT-BED use HY3403D, parameters is 30V/100A HOT-END use HY1403D, parameters is 30V/42A

After reading your comment, I would not believe those numbers. They are the maximum rated current for these FETs. Even with enough gate voltage that current would be possible only with very good heat dissipation.

I suppose it is able to handle the heat bed of an Ender 3 printer, as it is designed to be installed on it. Maybe a bit more?

Google says the ender 3 heated bed is 220W, with 24V that is 9.16A peak current.

JazzinSA commented 2 years ago

OK, So a little feedback on this issue. I have changed the Marlin code to PWM maximum to 230 and not 255. Therefore the average current on startup to get the bed on temperature is about 15% less. The average current to keep the bed hot while printing is in any case ~20%..50% of the heat-up current so it should not interfere while printing. I have used my Thermal camera to monitor the heat and it stays well within the limit. (below 60DegC which I feel is a good setpoint). While printing it runs at about 40..45DegC (on a 80DegC bed) I reacon that provides good safety margin. The only issue was to get a good PID setting but the auto PID tuning in Marlin did the trick. Let me just add that I did add a motherboard cooling fan.