Rock5-ITX suddenly died

Hi all.
I was testing the Rock5-ITX the last weeks.
It was working well all the time. Now was building a fresh Armbian build, not even started to build and it crashed and was turned off completely.
If I now plug in the PSU it does nothing at all. No LED light up, display doesn’t turn on, seems completely dead.
Anything I can try? I’ll see to attach uart ttl. But I fear it’s not gonna give anything.

I had build it in a case. Wasn’t able to get SATA and NVMe to work.

I plugged everything out and tried, but nothing. I hope someone can give a pointer to get it back. Hurts a lot.
Greetings, NicoD

No uart output at all.

I don’t know if you’ve connected the front panel connectors, but if not, you may want to try to make the contact between the “power button” pins with a screw driver. It might be possible that it’s turned off and decides not to automatically start up.

If you’ve placed it in a case, you should also try to pull the board out of it. You may find a pinched wire under a screw, or maybe a screw blocked under the board and making a short somewhere. I find it strange that the board died like this. Your description sounds like a faulty DC-DC regulator. Stupid question, is your 12V PSU strong enough ? Too weak a PSU can see its output voltage drop under load, putting a lot of strain on DC-DC converters and may sometimes exceed their maximum switching current and fry (already faced this situation in my mcbin).

The power button was working before, and it should power on at plugin of PSU. Does nothing.

I’ll take it out of the case.
The PSU is 12V 5A. I did put 3 SATA ssds into it plugged them in but were not recognized.
Plugged them out after the crash and nothing.
Tnx for the help.

12V 5A definitely is perfect for this device. Maybe there’s a bad solder joint under a chip, but that’s not something you’ll easily spot, let alone fix yourself :frowning:

Just thinking, do you have an ATX PSU available ? Or as a fallback, a PicoPSU-like ATX adapter ? And while you’re at it, it could be useful to try another power block (just in case it cuts off for whatever reason), and if you have a voltmeter at hand, checking that the 12V voltage is correctly present on the connector once plugged in could be useful. We could imagine that a fuse or a reverse-diode blew on board and that maybe the PSU just cuts off when facing a short.

1 Like

I would start with multimeter and measuring main components for resistance and voltage as well as visual inspection for anything failed or blow up. Also it’s easy to check if AC is still ok and providing enough power to board.

Will do that now.
Will take it out of the case. Check if I don’t see anything suspicious.
I’ll test my PSU with another board. Then I’ll test the Rock5ITX with an ATX PSU.
Fingers crossed, tho I can’t do the work like that. Thanks

I tested the PSU on another board, it worked.
Took out the board, checked for visible damage, nothing to see.
Tried an atx psu and nothing.
I guess I can declare it officially dead, pass date Friday 12th of July.
I’ll contact Tom to let him know. Thanks for the help.

1 Like

There are chances Tom will be interested in knowing exactly how this happened so that they can possibly spot a weak point on it that can be addressed in a future revision. Let’s hope they at least find something.

I’ll see to send it back so they can debug the issue.
My knowledge is too limited to find that out.

It hurts a lot, I had bought this nice case for it. Wanted it to be my ultimate NAS.
d809825d-2f81-4f4d-91d1-1470840f4515.webp (299.3 KB)
IMG_20240708_181457.webp (220.1 KB)

Thanks

Oh nice indeed! I’ve placed it into my previous Aplus enclosure which is perfectly suited as well, and am still decided on making it my primary server. I’ve been using slackware for ~30 yrs and I noticed that the distro is no longer compatible with small systems due to humongous dependencies (objdump needs libbrotli, vim needs libruby etc) totally crazy things, that once you factor in the fact that they don’t have lib-level dependencies but only full packages, you finally have to install everything to have something usable, and it takes most of the eMMC’s 8GB :frowning: I’ve ordered an SSD adapter for M.2 key A+E to see if that does the trick as a system storage, I’ll report about it later.

This board would benefit from a 32 GB eMMC IMHO due to todays operating systems being quite big for bad reasons. It wouldn’t be much more expensive and would give the board a lot of extra value.

1 Like

@NicoD: I’ve seen just posted crime video about Your installation of this board:
https://www.youtube.com/watch?v=OEXJSwbznSQ

For me it’s obvious that You damaged the board yourself. Exactly as You said - it worked for weeks until You decided to put in radiator and insert it to case, as You can see - board did not survived this.

This is how You installed the plate under the board:

Probably it’s upside down, this white tape probably was there to secure plate-board contacts - to not short anything under the board. Also this plate have some kind of spacers to not contact the board, installing it in reverse caused the opposite,

here on your video it’s easy to see - plate is almost touching board:

image

Here are docs: https://docs.radxa.com/en/rock5/rock5itx/getting-started/assembly-guide and there You can find how plate it’s supposed to be installed:

Did that killed the board? I don’t know, but that’s not everything yet.

In docs there is: Finally, please tighten the screws in diagonal order to secure the heat sink, this will effectively prevent damage to the ROCK 5 ITX motherboard during the securing process.
For sure You need to be very gentle about this proces, You tightened screws very hard one after another next to that. You seem to tight them way too much.

When this was ready You decided to do that:
nico-itx-2

SOC is most sensitive part of the board and radiator is big block that have direct contact with that, all spring screws should keep it at place and leave some space for heat expansion. If it’s tightened too much then cpu will brake. Shaking everything via radiator is completely irresponsible - this will brake some pads under SoC and apply large forces to this delicate component.
You did not stopped here - radiator is not perfect holder for board, You should not use it to insert the board into the case.

After all of those board somehow managed to turn on, but that does not mean that process was successful at all. You clearly seen that it soon died. I don’t know if it was something shorted under board by incorrectly inserted plate, either screews tightened to much (+heat) or soc damaged by incorrect handling, but there are at least three reasons why You may damaged it Yourself.

In future consider leaving such process to someone else, someone more professional and much more careful. You managed to miss m.2 card slot at first, which is easy task, but shows how You are inaccurate. You may just physically damage some small components on board if You don’t pay that much attention. There is much more precautions like ESD, all those sentences in radxa docs to just be careful with electronic components. Here You obviously failed with basics.

Rock5 ITX did not suddenly died, it was murdered by You with cold blood…

That’s not what I said, you change facts to fit them to your narrative. I said it worked a week installed in the case with the fan.

I did place the plate wrong but I’m 99% sure that’s not the cause of the issue.

I don’t even want to read the rest.

Board works. With ATX PSU. I had tried with an old PSU not sure if that was still working. Now used a working PC and it boots fine.

So the issue is probably the 12V barrel jack

1 Like

Ordered this.
https://www.amazon.com.be/gp/product/B005TWE5E6/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1
Should work without needing a fan.

Earlier You posted here:

So it’s something new to the story.
Also You claimed that:

and now:

So You claim that it worked, then died, then worked on different board. This is Schrödinger power supply - working and not working at same time.
Either You are mixing several power sources in this story or you failing with basics.

I don’t know eighter if there is anything to short there by incorrect installation. That was simple mistake to avoid.

As far as I can see on schematic 12V and ATX 12V are wired together. Maybe someone else can measure it and see if there is anything more than on schematics, for now I don’t see a point why it should not work with one and not the other power source.

Im sure that shaking board and testing how much SOC can take such force will not help You and board may be not stable now. In fact You are lucky if it’s still running somehow.

Seems it was a dead psu, must be a reason why it’s not in a pc…