Repairing of X5000 [FIXED]

@All
Some months ago my x5000/020 start to power ups just for second time. I replace of course PSU, gfx, whatever else no luck, after 2-3 weeks it starts only one 4st-5st time, then i put just a mother board + new psu + just memory + gfx on the table, and after few weeks of work, it start to power up only from 10, or 20st time.

In the process i hookup the MCU debug cable, and got that when can't boot:



I do search through manuals, TRM, and watching on motherboard via microscope to find where probably PEX switch are, and which one, and i think (but for now not sure), this is one which have that small black heatsink, with 5 capacitors around. I didn't remove heatsink just yet, so that all wild guess based on TRM which of course not very correctly show where things are.

Currently all i tried it to re-solder those 5 capacitors, but that was false of course, as when i meter them , their capacitance was almost 99%. And i was told by one friend knowing those things better, those capacitors on the plate are of very good quality.

So, next thing i think about is to resolder chip itself (just a little bit giving a hot to see if it will help). Because board mostly non functional anymore anyway, so why not burn it out : )

But maybe some of you have some clue already and maybe some of you already had this kind of issue before ?

Thanks!

@kas1e

This is probably not going to be any help whatsoever, and you are smarter than me so.... But I thought I would mention that my recent problems booting my X5000/040 turned out to be a faulty SATA cable or SATA connection to my optical drive. Symptom was X5000 would boot to splash screen but it would not animate and froze. I never got around to reading the serial output. Eventually found it would boot again after removing and replacing SATA cable. I also replaced the CR2032 battery which showed low voltage, but this did not solve my problem.

Good luck!

@mbrantley
My issue happens early, much. I can even plug off everything, sata, gfx card, even memory, so just PSU , motherboard and debug cable. And when it starts correctly i have:



So for passing this stage, nothing need it to be connected, even memory not need it. But if i can't, then i can previous quoted output.

@kas1e

If you’ve already replaced the CMOS battery and checked all HD cables I say use a heat gun and reflow all chips/connectors. This ‘new’ solder isn’t like the good stuff from 30+ years ago

@328gts
CMOS batter was first changed of course, and as i say it's all happens even without HDD, GFX, memory, whatever. Pure motherboard PSU , etc. And yeah, i thinking about a heat gun and reflow all chips/connectors too.

There was a X5000 that was powering down because one of the voltages was low. Xena voltage I think... So the MCU kept powering down.

It's a bit simple, but if you jumper the power supply to ON, the MCU will be unable to shut you down.

Then at least you can see if it works anyway.

With just a wire jumper, you'll need to power down at the power supply, OR you can add a toggle switch if you want to be a bit more graceful.

I'm sure you can find "power up an ATX power supply" on YouTube or something. For a simple test you can just use a paperclip.

Please let us know if this helps. If this IS the MCU overreacting, at least making it public knowledge will be helpful.

Instructables

@LyleHaze

Interesting & definitely worth a try

@Lyle
Sadly but nope. I made wire jumper, so when i power on from PSU button, it start to works right away. Then i do "power on" on mother board pins so it actually starts, and all what i have is :

When simple power on PSU with wire jumper:



Then, wire "power on" on mother board:



And it same blink leds, just PSU didn't turned on.

If i tried power on via wire-jumper-psu button, and then "reset" on mother board, then:




Tried to put off SD card as well, then:




Not that i expect it to boot with SD card, but, it change the issue.. so maybe somehow related.

@All
Very good news : Thanks to MAX from ACube, he point me out that broken SD-Card can acts exactly like this when died, works-not-works, etc. Also, that boot process (even prior uboot itself), still read from SD card some bytes , so ... I checked my SD card , it was just some crappy "SmartBuy" , i take Kingston one, write an uboot image on it , and IT WORKS !!!

I can't bring all sort of bananas for now, but so far i were able to boot 5 times in a row already.

Will keep testing it.

For me this "ASLEEP" thing wasn't a clue, i were sure its PEX chip issues (even re-soldier capacitors around), but Max is one who know PPC from the roots (of course) , so ASLEEP is thing which tell that there something goes wrong when CPU reading some values from storage memory , which in our case SD card. Damn. DAMN ! BANANA !

@kas1e
Very good news!

@kas1e

That's great news, pleased for you mate.

Cheers

This is great, kas1e! I was a little bummed out about your X5000 and worried for mine. Haha. I remember Max and the ACube guys as being all around great Amigans and very helpful! I still have my little ACube Sam440ep. :)

Also glad no harm was done with the solder stuff there!

@kas1e

always happy and smiles when things get sorted out in AMIGA land

@kas1e

Very good news ! And now come back to develop and release a new improved and bug-fixed version of Mesa for all of us ! TNX

I like to hear that! ☺️👊🏻

@kas1e

Glad you solved!!!
No back to coding work mate!!!

EDIT1: yep K-L is right!! thx Max for pointing where to look.

Great!

And thanks Max! :)

@kas1e

This is great news.
I already keep a bootable flash drive inside each of my Amigas for emergency recovery. After this I think I'll make (and test) a backup SD card for each as well.
Much easier to make it when everything is working. For the cost of a few SD cards (and a CF for the X1000) it seems like cheap insurance.

Also glad for the chance to get the details on forcing the power supply out to the public, in case it helps anyone.

Lyle