5V regulator failure (U3)
-
Personally I would try to replace U3, but I have a hot air station. If you don't have experience with QFNs, it is going to be tricky.
Using an external 5V source is an interesting idea. The one possible limitation I see is that you will be limited to 1A instead of 2A on the +5 rail, due to D14 (note on the schematic). Also, if you can't remove U3, you might want to remove L1 and L2 so you don't end up powering or backfeeding the damaged U3. On the other hand, if you already have it running it might not matter.
-
So, based on the advice from this thread:
https://forum.duet3d.com/topic/6462/u3-burned-a-hole-in-itself
Specifically referring to DC42's speculation that if D2 failed open or short, it would likely take the U3 with it. I've checked D2, and it appears shorted. With my (admittedly cheap) DMM in resistance/diode test mode, shows as shorted; same readings with both polarities. Doing a sanity check with another diode, it shows different readings depending on the DMM polarity, so I do at least somewhat trust the DMM reading.
So that somewhat changes the question. With the board isolated,I'm unable to find any wires/stray pieces of metal shorting the +5V circuit to ground, but there is continuity between the 5V pin and ground. I've double and triple checked and can't find any external shorts. So I suspect I can't use external 5V since there is still an internal short somewhere. Any thoughts other than a D2 + U3 replacement? Is there any other component to test? I'm pretty useless at reading circuit diagrams to determine test points.
-
Or rather, I'm dumb and forgot how electricity works. If U3 is internally shorting, that's obviously what a measurement of D2 when it's still attached would show.
-
I guess I was confused because it sounded like you had it up and running on external 5V. You are right, if U3 is slag, it could be pulling down the 5V rail.
If you are not going to replace U3, you should isolate if from the rails. Looking at the actual board and the schematic again, I think the best way would be to remove L3 (not L1 like I originally suggested). While you are at it, remove L2 to isolate it from the input rail. Then check the 5V rail again.
On the other hand, if you want to try to replace U3 and you suspect D2, you would probably do best to desolder it and check it out of circuit. Or you could desolder U3 since you know it is bad, and then check D2 in-circuit.
-
From experience, trying to find shorts on power rails can be a pain, so good luck!
-
If you don't want to attempt to repair the board yourself you can contact sales@duet3d.com and inquire about an exchange for a refurbished board at a reduced cost. Refer to this thread. We'll do our best to get you sorted.
-
Its worth taking into account that the board is 2oz copper (assuming genuine duet), the failed chip is in one of the most densly populated areas with heavy ground planes in 4 layers.
If you understand the implications of the above then go for it, if not I'd go with ext 5v if that is an option, but you might have to remove the failed chip if it has an internal short. Just removing the failed chip is not as difficult as replacing it, but its easy to use too much heat or spend too long resulting in damaging the board and/or surrounding components.
edit: using pre-heater / hot plate helps a lot!
-
@bearer said in 5V regulator failure (U3):
Its worth taking into account that the board is 2oz copper (assuming genuine duet), the failed chip is in one of the most densly populated areas with heavy ground planes in 4 layers.
Yeah, that actually sounds really unpleasant. Removing L3 and L2 would probably be easier, but it is still tight.
-
Just wanted to update everyone and say a huge thanks to everyone that's commented. I'm still weighing my options about going forward and at the very least considering a few options going forward.
1: Depending on cost/lead time for a refurb board, perhaps go that route. I'll be honest about the fact that I'm a bit impatient and I was very excited to have a multi-year long project come together and work (albeit briefly). It's a large format build with a lot of bells and whistles that I don't have on other 3D printers, so I was very excited to see it move and have a decent amount of success printing.
2: Remove U3, and run an external 5V power supply. At this point, I'm somewhat leaning toward this option since I am running a BLTouch probe, and based on what I've read, they can be the cause of 5V failures. Having an external 5V source would likely afford me some more headroom, and additional cooling potential, since it's not located in a heavily populated area on the board. I'm also really not running much off 5V (Just BLTouch) since I have 24V fans. And as everyone has said, removal is generally easier than re-attachment.
3: Remove and replace. I do have some concerns about repair; just due to the board's age and the fact it's a slightly older revision. Looking through some of the revision notes, it seems the newer 1.04b and c revisions did have some improvements to the 5V circuitry, so that slightly pushes me away from the replace U3 option and toward external 5v or refurb. I'm not going to say I'm 100% confident in my ability to remove and replace the chip, but in general I would classify myself as mostly competent with soldering.
At this point, I'd lean toward option 2, but that's without an idea of lead-time and cost of a refurb. I know Duet is based in the UK, and shipping time to the US is either cheap or fast; not usually both.
-
@nightfox939 said in 5V regulator failure (U3):
At this point, I'd lean toward option 2, but that's without an idea of lead-time and cost of a refurb. I know Duet is based in the UK, and shipping time to the US is either cheap or fast; not usually both.
We sometimes have refurb boards in the US, so it might not take too long to get one to you.
