Duet Maestro Bed heating uncontrolled

NTX9 · 22 Feb 2020, 11:01

I am running a Crane Quad with a Duet Maestro board. While running a print today the temp started climbing well above the thermister setting. When I reset the printer, the bed started heating as soon as the board was powered. Is this some kind of fault in the board?

A Former User · 22 Feb 2020, 11:14

Without more information it sounds like the bed mosfet might have failed. You could disconnect the bed and measure the voltage on the output with the printer idle.

dc42 · 22 Feb 2020, 11:29

The other likely possibility is a short between the bed heater negative wire and ground.

Bed heater mosfets on genuine Duets are very reliable and will normally fail only if the bed is shorted.

To tell whether the mosfet has failed, disconnect the bed heater form the Duet, then see whether the red LED near that output turns on as soon as you apply power. If it does, then either the mosfet has failed, or the back of the board is shorting against something conductive, or there is an error in the config,g file (but I presume you haven't changed the config.g file).

NTX9 · 22 Feb 2020, 17:12

Thank you for the input. I will check these later this weekend.

NTX9 · 22 Feb 2020, 18:20

@dc42 said in Duet Maestro Bed heating uncontrolled:

The other likely possibility is a short between the bed heater negative wire and ground.

Bed heater mosfets on genuine Duets are very reliable and will normally fail only if the bed is shorted.

To tell whether the mosfet has failed, disconnect the bed heater form the Duet, then see whether the red LED near that output turns on as soon as you apply power. If it does, then either the mosfet has failed, or the back of the board is shorting against something conductive, or there is an error in the config,g file (but I presume you haven't changed the config.g file).

The bed does not appear to be shorted to ground. With the bed disconnected the red led comes on immediately and the board produces 24v across the bed terminals on startup. Reinstalled stock config.g with no change. Will try to remove board to check for shorts on back next.

NTX9 · 22 Feb 2020, 18:25

@dc42

There is no visible short at the back of the board and the red LED lights as soon as power supplied.

NTX9 · 22 Feb 2020, 18:29

Prior to this fault, I had an issue with the hotend thermister. It was throwing inaccurate readings and may have been shorting. Could this have damaged the mosfet?

If the mosfet has failed, this seems like a extremely problematic issue. The amount of power being applied to the bed seems like a tremendous fire hazard.

elmoret · 22 Feb 2020, 22:16

@NTX9 said in Duet Maestro Bed heating uncontrolled:

Prior to this fault, I had an issue with the hotend thermister. It was throwing inaccurate readings and may have been shorting. Could this have damaged the mosfet?

If the mosfet has failed, this seems like a extremely problematic issue. The amount of power being applied to the bed seems like a tremendous fire hazard.

It’s the same amount of power as is applied when normally heating up...?

Most people size their bed heater so the bed cannot exceed safe limits in the event of controller or FET failure. If one chooses not to do this, sound engineering practice calls for a thermal fuse.

NTX9 · 22 Feb 2020, 22:40

@elmoret
In regard to your comment about sound engineering, the engineering was done by M3D for whatever that is worth.

You may be right that the bed may be sized to prevent exceeding safe limits. However, with the mosfet fail, the temps hit greater than 100c and were still climbing. I don't know if I would call that safe...

infiniteloop · 22 Feb 2020, 23:38

@NTX9 said in Duet Maestro Bed heating uncontrolled:

I don't know if I would call that safe…

As we deal with lots of electrons and a lot of heat in our hot ends, the print bed and maybe a heated chamber, safety is definitely a thing we should consider thoroughly. As every single component (such as a mosFET) can fail, a second line of defence will help.

However, there exist already several threads about this topic, so it might be best to have a look at these - they offer a lot of proposals and, to put it in @Danal’s words: „strong opinions“

dc42 · 23 Feb 2020, 17:22

@NTX9 said in Duet Maestro Bed heating uncontrolled:

Prior to this fault, I had an issue with the hotend thermister. It was throwing inaccurate readings and may have been shorting. Could this have damaged the mosfet?

No.

If the mosfet has failed, this seems like a extremely problematic issue. The amount of power being applied to the bed seems like a tremendous fire hazard.

That's why I always advise users to either select a bed heater that can't get to a dangerously high temperature even when run at full power for an extended period of time, or to fit a thermal cutout to the bed in series with the bed heater. I have a Crane Quad too, and the bed heats up slowly enough that I doubt that it can reach a temperature high enough to be a fire risk - but I haven't actually tested this.

But I repeat, it is extremely unusual for a bed heater mosfet to fail unless there is a short circuit across the bed heater. In printers with moving beds, there is always the risk of the bed heater wires chafing unless suitable precautions are taken.

NTX9 · 24 Feb 2020, 15:42

@dc42
I will examine the bed wires to see if that is the case.

NTX9 · 24 Feb 2020, 18:12

@dc42
I submitted a tech support request to M3D but have not seen any response yet. Frankly, I am not confident that M3D will warranty this board.

Is it possible to use an external Mosfet or SSR with the Maestro? Perhaps by connecting to the E1 heater header and re-configuring the firmware? Or by connecting to the Expansion header?

NTX9 · 25 Feb 2020, 02:21

On further investigation it is fairly obvious that the wires shorted against the aluminum bed. Looking at the way this connection was set up by the manufacturer, this short seems like it was inevitable.

Phaedrux · 25 Feb 2020, 03:15

yikes

droftarts · 25 Feb 2020, 17:15

@NTX9 Yikes indeed! Best check yours, @dc42 !

Ian

dc42 · 25 Feb 2020, 22:53

@droftarts said in Duet Maestro Bed heating uncontrolled:

@NTX9 Yikes indeed! Best check yours, @dc42 !

Ian

I will!

NTX9 · 26 Feb 2020, 03:39

For what it is worth, I have two and the newer one is in better shape.

Is the chip next to the LED on the board, the Mosfet that has been damaged? Or is there a diagram that shows the layout of the board.

The leads on that appear substantial enough that I might brave replacing the chip.

M3d customer service is slow at best and I don't have much faith that they will warranty the board.

dc42 · 26 Feb 2020, 09:50

@NTX9 said in Duet Maestro Bed heating uncontrolled:

Is the chip next to the LED on the board, the Mosfet that has been damaged?

Yes, the one marked 036N04L. To replace it you will need a hotplate as well as hot air, because of the amount of heatsinking that the PCB provides.

My Quad is in better shape than your photo, but there is some bare wire very close to the edge of the aluminium. I'm going to put some Kapton tape in the gap and over the edge of the aluminum.

droftarts · 26 Feb 2020, 10:11

@dc42 @NTX9 Those bare wires on the heatbed do look crazily close to the edge. Could you desolder them and resolder the other way around, so they looped back, with insulation over the edge of the bed? Or would you need to heat the bed to 200C first, as it's a massive heatsink?

Ian