Smart Effector hotend fan and Lights

oryhara · 5 Dec 2019, 16:07

At the risk of dredging up a dead thread, i want to give one last update.
I didn't calibrate the MCU temperature sensor, so i'm going to do that and it should fix the issue i mentioned in my previous post.

dc42 · 5 Dec 2019, 21:39

Re the blown fan MOSFET, it's likely that when the fan wire came loose from the connector, it shorted to something else, and that blew the MOSFET. I'm sorry, this sort of thing isn't covered by the warranty. However, Filastuder may be able to put you in touch with our stateside Duet repairer, if you haven't a local hackspace or makerspace where you can get it done. Or you could watch https://youtu.be/LADgxZgyJ04, practice on an old hard disk drive, and if that gives you enough confidence then do it yourself. If you can remove the blown MOSFET without lifting the traces, fitting the new one is easy.

Re CPU temperature, 45C is not unusual. The CPU will function up to at least 85C die temperature.

HTH David

oryhara · 6 Dec 2019, 16:20

thank you. It's worth asking i suppose. I'll stick with the workaround I've found. I rather like the effector lights always on.

My MCU was reading 6 degrees too high, based on the 'calibrate MCU temp' procedure which I executed this morning. Given that it was hovering at 47 during a print before that calibration, it should stay well below the warning threshold now.

dc42 · 6 Dec 2019, 16:35

@oryhara said in Smart Effector hotend fan and Lights:

thank you. It's worth asking i suppose. I'll stick with the workaround I've found. I rather like the effector lights always on.

My MCU was reading 6 degrees too high, based on the 'calibrate MCU temp' procedure which I executed this morning. Given that it was hovering at 47 during a print before that calibration, it should stay well below the warning threshold now.

Is the warning coming from Duet Web Control, or somewhere else? 45C is too low for a MCU temperature warning threshold.

oryhara · 6 Dec 2019, 17:51

It was high temperature on drivers. usually 0 and 1.
I see it in gcode console and a popup on web control and the panelDue.

I thought the warning started at 45 and then they stopped at 65, is that not correct?

dc42 · 6 Dec 2019, 18:44

@oryhara said in Smart Effector hotend fan and Lights:

It was high temperature on drivers. usually 0 and 1.
I see it in gcode console and a popup on web control and the panelDue.

I thought the warning started at 45 and then they stopped at 65, is that not correct?

The drivers warn of over temperature when their temperatures exceed about 100C according to the on-chip sensor. That warning is not related to MCU temperature, except that when the drivers are generating a lot of heat, some of this will feed through to the MCU.

oryhara · 6 Dec 2019, 18:55

I can probably afford to turn down the current.
But with a fan blowing across the back of the board, should they really be getting that hot?

I wonder if there is some other problem.

droftarts · 6 Dec 2019, 21:48

I was doing some testing of the latest firmware (2.0.5RC2) this week, and set up my Maestro with 4x chunky NEMA17 motors, on drives 0, 1, 2 and 3, set stepper current to 1.6A, set to spreadcycle (runs hotter) and 100% idle hold. I created a Gcode file that ran all four motors for 20 minutes, which I ran twice back to back. No fan cooling on the board, ambient temperature was 18C.

Times between high temp reports of different motors:
12/5/2019, 11:50:35 AM Warning: high temperature reported by driver(s) 0 1 2 3 - (3 minutes later)
12/5/2019, 11:47:31 AM Warning: high temperature reported by driver(s) 0 1 2 - (10 seconds later)
12/5/2019, 11:47:21 AM Warning: high temperature reported by driver(s) 0 1 - (9 seconds later)
12/5/2019, 11:47:12 AM Warning: high temperature reported by driver(s) 0 - (33 seconds from start of print)
12/5/2019, 11:46:39 AM M32 “0:/gcodes/StepperTest.g”

I couldn’t get them to 130C, which is when they’d shutdown, so while they all generated high temperature warnings, motor movement continued for the whole test uninterrupted. Motors got pretty hot too! Temperatures dropped below 100C within 30 seconds of the end of the test, despite being on 100% idle hold.

Now I’d be more worried about motors overheating than the stepper drivers!

Ian

dc42 · 7 Dec 2019, 07:21

@oryhara said in Smart Effector hotend fan and Lights:

I can probably afford to turn down the current.
But with a fan blowing across the back of the board, should they really be getting that hot?

I wonder if there is some other problem.

How hot are the drivers getting?

oryhara · 7 Dec 2019, 16:39

@dc42 How can I find that out?
I looked for a gcode but was unable to find one.

oryhara · 7 Dec 2019, 16:50

I think I fixed it.

I repositioned the fan blowing at the MCU so it aims underneath instead of directly at the board from the side(see picture). That fan was standing straight up vertically, now angled down under the maestro.
Haven't seen a warning since.

jay_s_uk · 7 Dec 2019, 17:26

1.8 amps is the full current requirements of the motors fitted to the predator.
Try dropping the current to 1400. That's what I have mine set at.

Phaedrux · 10 Dec 2019, 02:42

@oryhara Would it be possible to mount the board vertically? That would have a big impact on convective cooling.

oryhara · 10 Dec 2019, 14:31

@Phaedrux Not in the predator frame. It's got a triangular pancake space up top. That's where the controller lives. The cover plate is off in that picture, but the height is limited to the space between the outsides of those two 20mm T-slot extrusions visible at the edges of the picture.

Angling the fan downward has fixed the issue. I see no more temp warnings.