Duet Wifi Bricked probably a short 24V F_VIN to 5V

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

USB pin 5

that would imply you had a ground loop causing significant current through the usb ground.

you can replace the connector and run a bodge wire over to the capacitor just next door. (if the trace is fried then the pins in the connector is probably fried as well)

but depending on what caused the ground loop you may find more faults as well

dc42

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

I have looked around a bit on my issues, I think that I have somehow shorted Fan VIN to 5V using an adaptor board with a metalic screwhead touching the back of the board, the thing is I am not sure because where this may have occurred I dont see any scorch marks and I had to push a fair amount to get the screw to touch when I powered things off.

If you did short VIn to 5V then it's likely that multiple components on the Duet will have failed, including the processor, WiFi module and SD card.

The other possibility is that changed the second it happened was I turned on the 24V power supply at the same time I had the duet plugged into the PC which as far as I can tell is not supposed to be an issue.

it is an issue if the 24V negative output was connected to mains ground and either you did not have a good connection between the 24V negative output and the Duet VIN connector, or the PC was plugged into a different mains socket and there is a significant different in ground potential between the two sockets. See https://duet3d.dozuki.com/Wiki/USB_ground_loops.

Anyway I can see visible damage to the U11 chip

That is a common result of a short between +5V and VIN.

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

Update, well I can now see under good magnification and light that the USB pin 5 is toasty

That suggests you had a USB ground loop issue.

ozpete81

Just pulled the usb connector, the pin on the connector is gone, and the trace is lifted and slightly shortend (though possibly serviceable, I guess when you said the capacitor just next door you ment C13, which looks like goes to ground on the schematic) I guess it wont matter which side of cap (I'm a hobbiest with a smd rework station so dumb questions may come up, Im just checking).

Ill go into Jaycar in the morning, but still wondering if this is a pointless exercise, its not cost me to much so far Ive got the board cleaned up with U11, and U2 off. I took U2 off so I could check components like the CPU and wifi for overheating as mentioned in another post, with the 3.3v rail but I dont think I can do that till I get U11 back on but not sure.

If it has shorted to ground at the USB connector, I have to start doubting if I did actually short the 5V to fan_vin as I dont seem to have the same component damage as others with similar shorts.

Maybe it does have to do with something with having the USB connected to computer and turning on 24V, as I mentioned earlier everything was working fine till then (and yes everything was turning on and off perfectly before then with USB and 24V separately, so it wasn't a board wiring issue.) I had it plugged into USB 3.0 in the PC a decent usb type-b cable and into a "Chinese" USB B to micro (the micro end has scorch marks on pin 5 too I can see now.

I have to wonder if it was USB 3.0 current draw and sudden powering up from the 24V VIN that may have overloaded it or something? I think on the schematic the VBUS is protected by a diode so I guess its unlikely. Especially as I believe this is an operation mode the board is supposed to handle.

We will see mid month I guess, when the parts arrive from digikey, I might try plugging in the USB when I get it to see what happens when I get the connector fixed up, see if more magic smoke is still inside this thing.

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

I guess it wont matter which side of cap

furthest from the board edge is gnd

i'd check for a short on between 5v and gnd, replace the USB connector and test with external 3.3v. if the cpu and wifi is okay it will work without u11 (u11 drives the heater mosfets but heaters won't work without Vin anyway)

ozpete81

Can Confirm no short from 5V to gnd, and replaced usb connector I didnt need to run the bypass was able to get pin 5 to solder to the damaged tracing and also replaced the cable (that was burnt too).

PC USB port seems to be functioning fine BTW; but think its highly likely Im dealing with a USB ground loop issue. But not sure if the damaged U11 is nominally a symptom? and Im not trusting of U2 component. I know my connections to VIN negative were good, bootlace ferrels and ring connector at PS end tightened down good.

I connected the duet to computer and its no longer detecting anything but the 5V and USB LED turns on, I assume that this is a good thing as the usb device is driven off the cpu and that is unpowered, and there is a functioning ground pin.

I checked at pads U2 (removed) for 3.3V short to ground and my multi meter is reading 0.03 ohm on centre pin and 7.3 ohm on the outer pin (processor side) i dont know if this is expected but seems like it could be a short, though I checked on a mighty-board FF RevD I had spare at U14 (24V to 3.3V voltage reg.) and that was indicating short on second pin and the board is fine. Also was planning on using the Mighty-board as my 3.3V source but it out puts 3.8V even if I turn the PS to 19V, If I used that will I damage the duet further? (I think this particular voltage reg was changed in later revisions of mightyboard due to failures)

Also as more info that may diagnose the cause I noted that the 24v power supply I was using when the damage occurred (out of the FFCP im converting) was set at 25.1V I wonder if that may have precipitated the incident combined with the USB 3.0 Voltage? Might be barking up wrong tree?

If I cannot use the Mightyboard 3.8V output Ill have to wait for my digikey order to arrive I guess, and suck it and see when I put a new U2 and U11 on.

Cheers Bearer awesome help.

25.1V

https://duet3d.dozuki.com/Wiki/Power_Wiring says max 25v but afaik they are tested at even higher voltages so unlikely to be an issue.

3.8V

both the cpu and wifi is rated for 3.6v max so 3.8v is probably worth waiting for the 3.3v regulator. Or maybe connect a diode or resisor in series to drop 0.5v at 200mA if waiting isn't an option.

7.3 ohm on the outer pin (processor side)

does seem like a bit of a low value indeed; current limited supply is advisable for just this reason. i'd expect to see a couple of kilo ohms or thereabouts.

ozpete81

@bearer said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

both the cpu and wifi is rated for 3.6v max so 3.8v is probably worth waiting for the 3.3v regulator. Or maybe connect a diode or resisor in series to drop 0.5v at 200mA if waiting isn't an option.

I did this worked nicely, seems the wifi module got warm and the blue light was on very dimmly, cpu seemed cold, so maybe a wifi module blown by the look of it. $8 for a replacement seems reasonable so another component came off. (the 3.3 v and diag. LED came on so hopefully this is it fingers crossed, now waiting for more parts.

But I am concerned that the resistance is still a bit low on the pin u2 (on the processor side) I said was 7.8 has increased to 66 Ohm, which I think may indicate something else nasty, Bearer said: i'd expect to see a couple of kilo ohms or thereabouts" Although the cpu with the wifi module taken off is cool with the 3.3V applied.

I think I was reading elsewhere on the forum others suggesting connecting USB at this point, with the 3.3V connected is possible to look at the cpu, is that possible/wise?

Cheers Again Bearer

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

I think I was reading elsewhere on the forum others suggesting connecting USB at this point, with the 3.3V connected is possible to look at the cpu, is that possible/wise?

cpu will act as a usb device with only 3.3v. the diag light being on suggest the firmware may be corrupt or cpu toast. if you don't get a serial port try the erase jumper as described in https://duet3d.dozuki.com/Wiki/Installing_and_Updating_Firmware#Section_Fallback_procedure_Num_3

ozpete81

Well I think this might be the end of the road for this board.

I get 5V 3,3V USB and Diag all on but Windows reporting Unknown USB Device (Device descriptor Request Failed) Error code 43, tried different cables rebooting and different ports It wont recognise the device.

I think that that means the USB device is damaged which means the CPU gone and I think means its uneconomical to fix if all of that has failed. At least the SD cad is ok I guess.

Unless there are any other suggestions, I guess I better start saving my pennies for Christmas.

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

It wont recognise the device.

if the fallback procedure i linked didnt work, then odds are it needs a new cpu afaik.

ozpete81

Thanks very much Bearer for sticking with me on this unfortunately the fallback procedure didn't work, Time to call it. Full Points 10/10 on Duet destruction.

dc42

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

I get 5V 3,3V USB and Diag all on but Windows reporting Unknown USB Device (Device descriptor Request Failed) Error code 43, tried different cables rebooting and different ports It wont recognise the device.

Was that after using the Erase jumper? If it wasn't, try again after fitting the Erase jumper.

ozpete81

@dc42 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

Was that after using the Erase jumper? If it wasn't, try again after fitting the Erase jumper.

Hi David I have version 1.01 hardware I tried the Erase button, didnt seem to help, should I try to hold down while plugging into usb or turning on 3.3Vrail? its probably the only thing I havent tried.

I will probably fit the U2 and U11 later when the parts arrive, and try again but I dont expect it to work I just want to practice with the equipment I borrowed off a friend, not much use for the parts otherwise. Ill keep the board as I will probably order another duet by end of the year if the domestic CFO approves.

But was looking at new slice engineering Mk10 to improve the heat management on the Flashforge hotend as Im wanting to insulate and heat the chamber, plus no more thermocouples and uses e3d nozzles, I have the Bondtech Replicator extruders installed already. Ill pull the Duet board out of my other printer (Hypercube Evo) in mean time to get project moving.

Its really disappointing as I was just commissioning it and everything was working well when this USB goundloop hit. (which doesnt appear in the "How to destroy your Duet" post but is in the Warnings which I dont recall reading about when I was last using the forums couple years ago, I thought I knew everything)

I am pretty sure now that is what it was as I was thinking that the 24V to 5V short must have been it as the screw was sort of close and symptoms were similar but I didnt see how it did it spontaneously as it took significant pressure to get it to touch otherwise, and I didnt know about the ground loop issue earlier.

As they say "**it Happens"

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

I didnt know about the ground loop issue earlier.

it shouldn't be an issue unless there is faulty wiring somewhere; might be worth while looking into.

ozpete81

@bearer wiring on my machine or wiring in my house 240V.

All was well when it was running on 24V and all was well when I had it plugged into USB, the short happened the instant when I had it plugged into USB and then turned on 24V, I know I had a solid connection to VIN, crimped bootlace ferrells tightened down well and crimped ring connector on the Meanwell. It could have been the cheap chinese usb cable or poor grounding, which I read as something wrong with my house, or power boards im using on PC/printer.

I guess there could be other possibilities but I wasn't heating anything, or moving axis and no end stops were triggered, and I had just tested the endstops were working correctly a couple minutes before and I dont have a Z probe. I was having issues connecting to wifi; it doesnt have external antenna (hence why I was connected via USB but I believed that was caused by mounting under the Flashforge all metal frame.

The screw I thought may have been the culprit was grounded to the metal frame, the power supply and the wall socket, maybe that was the route of the short, but again Im not sure the board contacted the screw, may have been close.

Now Im nervous about plugging in the newer duet I have in, I certainly wont be doing both the USB and VIN, but now I am paranoid, but cant think of anything else I can check, I have made absoluetly sure belt and braces that the board cant contact any metal; could it be a dodgy 24V power supply? Its the Mean Well NES-360-24. I have a Mean Well 450 coming to replace anyway.

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

wiring on my machine or wiring in my house 240V.

could be either.

@ozpete81 said in Duet Wifi Bricked probably a short 24V F_VIN to 5V:

I certainly wont be doing both the USB and VIN

that is certainly safer, or if you have a laptop, have it running off batteries, not connected to a charger when using usb. (or get an usb isolator, quite handy)

ozpete81

For posterity and possibly others interest; after hot air soldering on the U2 and U11 chips (without the wifi module installed I tried again and got no different result to before. Unrecognized device. rebooting and trying completely different computer didnt work. I now have extra parts for next time I screw up. Thanks David @dc42 and @bearer for your help.