A5000 so rotten

discuss the archimedes & risc pc, peripherals and risc os/risc os on pi
Post Reply
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

A5000 so rotten

Post by IanJeffray »

Here we go again - another quick look at the state of a machine and hopeful quick repair. A basic A5000 (ALB12 - IOEB/2MB/25MHz) this time... and from all the horror stories here, I expected ... ah, yes... that...
IMG_E4857.JPG
The floppy and IDE cables are so corroded, no amount of pulling or screwdrivering would release them. I've had that before on A7000. The whole connector has to be desoldered and replaced. Sigh. Anyway... some quick sanding and scrubbing later, the board's revealing its damage more clearly:
IMG_E4862.JPG
and those cables are now gone ... just a quick hack to get me closer to the issues for now... brutal:
IMG_E4863.JPG
After some buzzing, all the usual rot around the battery (and sniffing at C169 made it fall off with its pads!) ... there's going to be quite a bit of work required here and the result's not going to be the prettiest. Ah well. What life do we have? What can POSTbox tell me?

Nothing.

Ah. Great. It's always more 'fun' when we don't even get a POST. Start as usual by checking clocks ... 50MHz input Xtal... 25MHz out of the flipflop ... no ... oh hai problem #1 ! That's a bit surprising - the ACT74 has died. Wow. I'll swap that out before moving on I guess. [Damn, just missed today's Farnell cut-off time] Even if this board's ultimately a scrapper, it'd be useful to use it to verify which of the components can possibly be salvaged.
Last edited by IanJeffray on Thu Apr 22, 2021 1:40 pm, edited 1 time in total.
Simon
Posts: 85
Joined: Sun Apr 12, 2020 9:20 pm
Contact:

Re: A5000 so rotten

Post by Simon »

Wow I thought my board was bad, but that's horrific, poor a5000 :(
Electron - Issue 6 (kernelcrash ram/rom board)
BBC B - Issue 7 (PiTubeDirect, RGBtoHDMI, Watford 1770, Pi1MHz, boobip rom/ram, boobip EEPROM)
Master 512 - Issue 2 (AKF12, External PiTubeDirect, Pi1MHz, RetroHardware ARA III, VideoNuLA)
User avatar
DutchAcorn
Posts: 2425
Joined: Fri Mar 21, 2014 9:56 am
Location: Maarn, Netherlands
Contact:

Re: A5000 so rotten

Post by DutchAcorn »

Seems to have been stored on it’s side. I’ve had one A5000 that had been put away upside down. That kills the IDE drive but at least the pcb is then unaffected. But I guess not many ppl store their old computers upside down.
Paul
User avatar
flaxcottage
Posts: 4579
Joined: Thu Dec 13, 2012 8:46 pm
Location: Derbyshire
Contact:

Re: A5000 so rotten

Post by flaxcottage »

I had one like that. :(

It was working but just before the sale the 'kind' vendor tried to wipe the gunge away and didn't let on. Almost all the components around the battery were wiped away too. No wonder it never worked again!! Still it had some good spares. :lol:
- John

Check out the Educational Software Archive at www.flaxcottage.com
User avatar
a1exh
Posts: 49
Joined: Fri Apr 16, 2021 5:35 pm
Location: Oxfordshire
Contact:

Re: A5000 so rotten

Post by a1exh »

Simon wrote:
Wed Apr 21, 2021 9:17 pm
Wow I thought my board was bad, but that's horrific, poor a5000 :(
Me too. This is much worse than mine.

I'm just starting my A5000 battery repair. The case hasn't rusted on mine but everything is green including under the FP10 socket and under the IDE connector and along this track that runs on the very edge of the board.

If you get it to work there is hope for me.

I'd be very interested to read about your repair, the diagnostic technique... what to measure / scope etc. if (when) it doesn't boot.
Principal ASIC Engineer
RetroGeek
Acorn Electron + A5000
http://thalion.atari.org
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

a1exh wrote:
Thu Apr 22, 2021 12:45 pm
If you get it to work there is hope for me.
I've not been defeated yet. (Famous last words). Have recovered several machines even worse state than this

If you want to see a properly revived trashed A5000 though, there's no comparison to myelin's madness - viewtopic.php?p=260770#p260770 :lol:
wicksee
Posts: 16
Joined: Wed Apr 21, 2021 2:47 pm
Contact:

Re: A5000 so rotten

Post by wicksee »

that's seen better days!!!

good luck mate :?
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

New flipflop installed. ARM now has 25MHz clock. But still not much further sign of life in general. Phooey. No time to look any more today.
IC61
IC61
Last edited by IanJeffray on Fri Apr 23, 2021 9:23 pm, edited 2 times in total.
User avatar
a1exh
Posts: 49
Joined: Fri Apr 16, 2021 5:35 pm
Location: Oxfordshire
Contact:

Re: A5000 so rotten

Post by a1exh »

Good work.

On the A5000 which I'm doing now the tracks that go from the ARM3, alongside and under the FP10 socket (near R291) are the worst affected.

I'm using a tiny water colour brush to apply the white vinegar but R430 and C154 just "washed away" with their pads the moment I touched them (the lightest of light touches) and the neutralisation started taking place.

The tracks going under the IDE socket don't look great either.
Principal ASIC Engineer
RetroGeek
Acorn Electron + A5000
http://thalion.atari.org
thecellartroll
Posts: 354
Joined: Thu Nov 24, 2011 10:43 pm
Contact:

Re: A5000 so rotten

Post by thecellartroll »

I'm surprised that no-one has re-created the A5000 mainboard since the schematic is available. I'm guessing that the ARM3s and MEMCs on these rotten boards mostly work. It would be fun to build an A5000!
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

Another quick bit of poking around -- IC36 and IC46 both appear completely goosed. Using the ARM's address walk under reset, we see the walking address presented at MEMC and the inputs to these three latch chips, but nothing at all on the outputs of IC36 and IC46. IC45 is working as expected. Pin 1s are all connected but ... ah ... Pin 11s are interconnected between the 'faulty' pair, but not connected to the IC45 Pin 11 - so we have our first board-level fault -- must be a broken trace there. I'm going to guess it's nearer to IC46 because that's closest to all the battery rot... let's see where a trace could run from IC45 to IC46 - only one real suspect and it goes to a via right by IC46, which we'll assume is where it then goes onward to IC36 - yep, that via's connected tp pin 11. Scrape a bit of trace nearby and ... it's not connected to the via - aha! - and that trace is indeed connected to IC46 pin 11. We found our first track fault. I always like to "root cause" these things, so I can then apply the smallest possible patch - it would be quicker to avoid this step and just wire the pin 11s together, but single-ended traces r'bad and such a big bodge would look uglier.
IMG_4877.JPG
It's not clear where the track break really is - there's clearly tarnish, but it becomes more obvious when I scrape away all the resist toward the via and flood with solder...
IMG_4881.JPG
This is such a common fault with the more recent Acorn machines -- whether the tension in the track-to-annular-ring joint is greater than other places, or just thinner and more susceptible to rot, I'm not sure. Anyway, it's a quick and easy fix with a tiny wire:
IMG_4883.JPG
Buzz-out afterwards. Yep, all pin 11s connected as required now. Hurrah.

Power-up... still nada. Check for the walking address bus under reset and yep, we've got that at the ROMs now which is good news. Still not enough working for any POST output yet though - hohum. More for another day.
sirbod
Posts: 1216
Joined: Mon Apr 09, 2012 9:44 am
Location: Essex
Contact:

Re: A5000 so rotten

Post by sirbod »

Except for the two Iyonix, which are just plain evil, my A5000 is the only one that has defeated me so far. I’ve been working on it for about 3 years with little success.

In my case the acid ran west along the top side of the board, past the CPU and lifted many of the components off their pads. I’ve buzzed all the traces, vias etc and replaced all the lifted components but still nothing. I need one of those fancy plug-in diagnostic tools to see if it’s the CPU that’s failing, or something else.

Generally when these RO3.1 board fail to do anything at power on, it’s something around the clock/CMOS.

At some point I plan to rewrite the POST code in RO3.2 so it initialises VIDC first and displays a proper POST sequence - flashing lights and failing to boot with no CMOS is just too 80’s for me!
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

sirbod wrote:
Sat Apr 24, 2021 7:07 am
I need one of those fancy plug-in diagnostic tools to see if it’s the CPU that’s failing, or something else.
Do you mean a POSTbox or a scope? POSTbox actually needs the CPU/ROMs/etc to be working as it's "just running code". At this point, a multimeter and scope are the order of the day.
sirbod wrote:
Sat Apr 24, 2021 7:07 am
Generally when these RO3.1 board fail to do anything at power on, it’s something around the clock/CMOS.
Dont' agree about that statement at all. CMOS RAM is one of the last things to bother about IME and is never the reason a machine won't do "anything" at power-up - it's not even addressed at all until well in to POST.
sirbod wrote:
Sat Apr 24, 2021 7:07 am
At some point I plan to rewrite the POST code in RO3.2 so it initialises VIDC first and displays a proper POST sequence - flashing lights and failing to boot with no CMOS is just too 80’s for me!
Certainly there's more that diagnoistics could do. POSTbox output still needs "interpretation". But again, it's still only useful when we're much further down the line than this sort of fail with basic clocks and latches. I'm not even sure I've got code actually running yet. A single busted address or data line to the ROMs can prevent even POST running. I've also had situations where the ROMs themselves were duff, which was fun to diagnose. viewtopic.php?p=288737#p288737
HughF
Posts: 48
Joined: Sat May 01, 2021 7:42 am
Contact:

Re: A5000 so rotten

Post by HughF »

thecellartroll wrote:
Fri Apr 23, 2021 6:15 pm
I'm surprised that no-one has re-created the A5000 mainboard since the schematic is available. I'm guessing that the ARM3s and MEMCs on these rotten boards mostly work. It would be fun to build an A5000!

This would be a cool project, I know the Amiga boards have all been cloned nowadays. Perhaps it would test my kicad skills to the limit :D
thecellartroll
Posts: 354
Joined: Thu Nov 24, 2011 10:43 pm
Contact:

Re: A5000 so rotten

Post by thecellartroll »

HughF wrote:
Tue May 04, 2021 11:34 pm
thecellartroll wrote:
Fri Apr 23, 2021 6:15 pm
I'm surprised that no-one has re-created the A5000 mainboard since the schematic is available. I'm guessing that the ARM3s and MEMCs on these rotten boards mostly work. It would be fun to build an A5000!

This would be a cool project, I know the Amiga boards have all been cloned nowadays. Perhaps it would test my kicad skills to the limit :D
I'm willing to contribute when you put your JLPCB order in :D
HughF
Posts: 48
Joined: Sat May 01, 2021 7:42 am
Contact:

Re: A5000 so rotten

Post by HughF »

thecellartroll wrote:
Wed May 05, 2021 10:52 am
HughF wrote:
Tue May 04, 2021 11:34 pm
thecellartroll wrote:
Fri Apr 23, 2021 6:15 pm
I'm surprised that no-one has re-created the A5000 mainboard since the schematic is available. I'm guessing that the ARM3s and MEMCs on these rotten boards mostly work. It would be fun to build an A5000!

This would be a cool project, I know the Amiga boards have all been cloned nowadays. Perhaps it would test my kicad skills to the limit :D
I'm willing to contribute when you put your JLPCB order in :D
I'd really like to see an open source motherboard/backplane that used a RPi CM4 and gave us a RiscOS compatible mini-atx motherboard. That would be super cool:

EDIT: https://www.hackster.io/news/ross-nicho ... 1884d9dfe2
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

With Adelaide out of the way, I'm back to this. Had to use a lot of heat, a lot of flux and a sigificant amount of violence to remove the remnants of the IDE and floppy connectors. Not really necessary at this point of course, but their disgusting state really annoyed me.
IMG_5122.JPG
More probing of clocks, resets, etc and I feel it's time to break out the logic analyser - I know I'm not in for fun when we're doing this. Start with the LA2 to LA9...
IMG_5123.JPG
Already we can see there's something very odd going on here. It's counting from 1 (not zero?!) to 8... sitting there for longer than expected, then repeating, forever...
image.png
Not starting at zero is "interesting" I'm sure. What's in the ROM at these places anyway?...
Capture.PNG
Ok, so it's not really executing ROM is it. But it's not the ARM's "walk the address but under reset" either, because when I do hold reset, I do see that appear cleanly as expected...
image2.png
So I'm a bit confused now. I'm hoping some smarter Stardotters than me will have an idea?
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

HMm. Actually, sometimes, if I leave it, it drops to this, with ROMCS de-selected...
image.png
Still not really any wiser as to what's going on though :) Probably not thinking straight.
User avatar
IanJeffray
Posts: 1139
Joined: Sat Jun 06, 2020 3:50 pm
Contact:

Re: A5000 so rotten

Post by IanJeffray »

Start up with reset held, then release it, and we get something that looks much more like sense for a while. So maybe the reset circuit's goofed (not holding for long enough on power up) - better look at that, then!
image.png
Post Reply

Return to “32-bit acorn hardware”