Fixed: A3000 black screen after power on.

discuss the archimedes & risc pc, peripherals and risc os/risc os on pi
Post Reply
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Fixed: A3000 black screen after power on.

Post by Nutter »

I have an A3000 which is in generally very good condition. Battery has been replaced and one damaged track has been repaired. Everything else is great shape. It was working fine for many months but when I powered it on just a few nights ago, the only thing I get is a green power LED. No disc LED, no sound, no flashing screen.

I have previously restored an A3000 as I posted on this forum a few months ago. It was almost a write off due to a battery holocaust but came good after a lot of work. I've looked at the manual which points to the ROMs being faulty and I'm looking for a lazy way out here. I'm trying to avoid getting out the scope etc as my desk is cluttered with other stuff which I don't want to disturb.

I've reseated all 4 ROMs but that made no difference. When I removed them all I measured 5.04V at PL3 and PL4. When I insert one of the ROMs (any one) the voltage drops to around 4.8V but it then begins to vary eg 4.6V, 4.75V, 4.58V etc This doesn't seem right but may be a red herring (or poor PSU regulation).

Any suggestions? I'm tempted to try and get a new set of ROMs and try them. That seems like a quick fix (or maybe not). Then if that is unsuccessful, I'll have no option but to clear the desk. Just don't know why the ROMs would die all of a sudden. Is there any way to dump them and compare them to what they should actually be?

Also does anyone know if the MiniPro TL866A programmer can interleave the available ROM images?

Thanks all.
Last edited by Nutter on Fri Jul 26, 2019 2:18 am, edited 1 time in total.
Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: A3000 black screen after power on.

Post by Kazzie »

The A3000 Service Manual indicates that the PSU should give an output voltage between 4.9V and 5.1V, so the voltage drop you're seeing feels a bit off.

I don't suppose the ROMs from the other A3000 would be any easier to dig out you than your scope and desk?
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: A3000 black screen after power on.

Post by Nutter »

I would have used those ROMs except I have only just sold that unit and it's in the post :(

I might end up just buying a new set of ROMs from somewhere. That appears to be the lazy way out here. Anyone here got a set for sale?
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: A3000 black screen after power on.

Post by Nutter »

Well I gave in and cleared some space on the desk. I have removed the 4 ROMs temporarily and ordered 4x 27C4001 just in case they're needed.

Initial tests show I am missing the CAS signals from MEMC at R129-132. I see there is a warning notice in the manual not to keep power on for longer than a few seconds if RAS / CAS are missing.
WARNING: System clock failure
f the computer is powered on for more than a few
seconds and there is no system dock, there is a
risk that all four ARM chips may be damaged, as
well as the RAM, due to lack of refresh on the
DRAM:
1 Check RAS and CAS on R133, R129-132.
2 Check IC41 pin 19 and IC37 pins 1 and 2 for clock
signals.
3 Check for a RAS signal on pin 5 of all the
DRAMS.
The investigation continues.
philpem
Posts: 626
Joined: Fri Apr 04, 2014 6:42 pm
Contact:

Re: A3000 black screen after power on.

Post by philpem »

Nutter wrote:
Wed Jul 24, 2019 5:00 pm
I would have used those ROMs except I have only just sold that unit and it's in the post :(

I might end up just buying a new set of ROMs from somewhere. That appears to be the lazy way out here. Anyone here got a set for sale?
If you fit the ROMs (or just one) -- do they get warm?

If they do, that means they're pulling a lot of current and one or both is probably faulty. If they're pulling a lot of current, that would also explain why your 5V rail is sagging.
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: A3000 black screen after power on.

Post by Nutter »

Thanks for the reply.

No they don't get warm. The voltage starts off at 4.95V then after a few seconds drops off slowly. I'm trying not to keep it switched on too long due to the warning in the manual although the 24Mhz clock is present. I'm a bit more concerned about the missing CAS signals at minute.
Last edited by Nutter on Thu Jul 25, 2019 2:33 am, edited 1 time in total.
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: A3000 black screen after power on.

Post by Nutter »

Success!

I managed to get this fixed after a bit of research (rtfm) and a bit of probing with the scope. The two issues I was seeing were both related: the power fluctuations and the 4 CAS signals missing.

After reading up a bit about the MEMC in VTI ARM databook from the Internet Archive I found that the following case can exist:
If an ABRT is generated during an N-cycle,
the -RAS strobe will be activated as
usual, but the -CAS3 - -CASO signals
are suppressed, effectively disabling
the DRAM cycle.
So I checked the ABRT signal and it was permanently active (high).
When asserted (high) MEMC has detected either an attempted access to a higher privileged area or a non-existent logical page. Both these conditions will cause an abort of the current cycle and exception processing to be invoked by the processor to determine error recovery procedures.
This led me to think there was an issue with the addressing lines. I started probing them and found that IC29 pin12 (A2) was only about 1V pk-pk on the scope whereas all others were around 5V pk-pk. On the other side of resistor R71 (La2) the signal was gone. Removing R71 restored the signal to its correct level so I knew something beyond R71 was affecting the signal. Measuring continuity between 0V and one side of R71 showed a complete short. I now knew it was not the ROMs that were the problem as they had been removed from the board.

Looking at the schematic I followed signal La2 to everywhere it went. I had to assume that the main chips were working so started to focus on the other ones. La2 was connected to pin 13 on the serial interface chip IC1. I removed it from the socket but that made no difference to the signal. However when I started to trace the PCB track from pin 13 back towards the ROM sockets, I noticed it went directly underneath PL3 the PSU 0V connector to the main board!

If I now take you back to my first post in this thread where I mentioned that I had removed a RAM expansion and ZIDE card before starting... well in order to install the ZIDE card I had to slightly bend PL3 and PL4 to allow the card to fit in. In doing so, PL3 had pinched the track for signal La2 and in doing so, it shorted La2 to ground!

I melted the solder on both PL3 and PL4 and pulled them slightly away from the surface of the board and now everything is working fine :D

So it turns out I caused this problem myself :oops: but it was a great learning experience trying to find the fault. Good thing I held off on buying new ROMs!
Attachments
PL3.jpg
Last edited by Nutter on Fri Jul 26, 2019 2:35 am, edited 1 time in total.
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: Fixed: A3000 black screen after power on.

Post by Nutter »

I really should have paid more attention to IanS's warning posted here in the For Sale thread.
Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Fixed: A3000 black screen after power on.

Post by Kazzie »

Nutter wrote:
Fri Jul 26, 2019 2:24 am
I really should have paid more attention to IanS's warning posted here in the For Sale thread.
Warnings or not, you found the fault and fixed it. Well done! =D>
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: Fixed: A3000 black screen after power on.

Post by Nutter »

Thanks Kazzie.

This was such an easy "mistake" to make and if you ask me, it was also a terrible PCB design. But sure... that was then and this is now! I'm so glad to have this machine working again.
User avatar
1024MAK
Posts: 10479
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Fixed: A3000 black screen after power on.

Post by 1024MAK »

More to the point, do you think the warning in the service manual about damaging DRAM chips if they don’t get refreshed is correct?
To the best of my knowledge, if DRAM chips are not refreshed, all that happens is they will after around 50ms, forget any data that they contain. I have not yet come cross any DRAM data sheet that states that loss of refresh will cause damage.

Anyway thanks for posting up your findings! And well done on tracing the fault, even if it was self inflicted :D

Mark
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: Fixed: A3000 black screen after power on.

Post by Nutter »

Mark I thought exactly the same thing but had to take the manual as being the "authority". It's funny you should mention that :lol:

As I know very little about Acorn computers I had to accept that what was stated in the manual was correct. It begs the question, why would they state that a system clock failure would damage most of the ARM chips and the lack of refresh signals would damage the RAM? If that was the case, should they not have designed some sort of fail-safe solution?
User avatar
1024MAK
Posts: 10479
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Fixed: A3000 black screen after power on.

Post by 1024MAK »

In terms of the Arm chips, given their cost at the time and the level of integration possible, why would you design a new silicon integrated circuit that could be damaged if it experienced a loss of / static clock signal?

Only some Acorn documents give the warning about the DRAM not being refreshed. Yet the earlier design of DRAM chips were more fragile than the later types used in these machines...

And in the 8 bit BBC Micro, if a fault prevents the OS from sending set-up data to the 6845 CRTC (which acts as the refresh controller), then there will be NO refresh to any of the DRAM chips. Nearly every Beeb that gets reported as a blank / black screen and a continuous beeee sound is operating without any DRAM refresh. Okay, many of these failures are due to a defective DRAM chip or two. But once the problem is sorted, the remaining original DRAM chips are happy to continue working for many years...

Mark
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: Fixed: A3000 black screen after power on.

Post by Nutter »

So what's the conclusion regarding this? Is it dangerous to run the system without a clock or RAS/CAS? My gut feeling tells me it's not but is there something internal to the chips that may be damaged?? I don't honestly know.
User avatar
1024MAK
Posts: 10479
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Fixed: A3000 black screen after power on.

Post by 1024MAK »

I can’t give you a definitive answer. And this is not the first time that this topic has been discussed on here.

Personally I don’t think there is a problem with running DRAM without /RAS and/or /CAS signals (in other words, without refresh).
Older DRAM chips needed an external -5V supply, as did early microprocessors, this -5V supply was to bias the substrate. But semiconductor manufacturers overcame this issue. Hence later chips only require a single positive supply. Loss of this -5V bias supply if the positive supply rails are present may well cause damage.

The normal things that can kill a chip are: excessive supply voltage, reverse voltage (on either the power pins, or any input or output pin), excessive heat, static discharge and a short circuit on an output pin (some types).

Most chips are far more resilient than people think.

So the only possible thing that I can think of, is that whoever wrote that section in the manual thought that either a bad thing would happen if the frequency of the relevant signals fell below the values in the datasheets, or thought that if needed, the chips internal bias power circuitry depends on an external clock signal.

Mark
Nutter
Posts: 57
Joined: Sun Sep 09, 2018 12:18 am
Contact:

Re: Fixed: A3000 black screen after power on.

Post by Nutter »

Thanks for your input Mark.

I don't think there is an issue running it without RAS or CAS simply because that is the state I found my A3000 in and that was due to the ABRT signal shutting off the four CAS outputs from MEMC, by design. There might well be an issue if the system clock itself fails but I can't reason why that would be. I suppose it's best to err on the side of caution and do as the manual says. It does however, make fault-finding a PITA as you can only keep the computer powered on briefly while probing around.
Post Reply

Return to “32-bit acorn hardware”