DRAM 128K system has stopped working

[CheshireNoir]

Hey Folks,
My DRAM 128K has stopped working. Symptoms are if you turn it on it doesn't even get a video synch. No beep from the speaker. Holding down M on boot doesn't get you into monitor. Holding M and hitting RESET sometimes gets you into monitor but none of the keys respond. Just hitting RESET gets either a garbled screen or a solid cursor. 
I've checked voltages (5.02v on the 5V rail) and the EPROM. (Looked a bit dodge so I re-burned a BN55 ROM image onto the BOOT ROM)
Before I break out the 'scope and the logic probe and start probing around, can anyone suggest where I should start my diagnosis?

Cheers and thanks in advance!

(Of course it failed as I was getting ready to take it to a retro computer meet)

Chesh

[MbeeTech]

Hey Folks,
My DRAM 128K has stopped working. Symptoms are if you turn it on it doesn't even get a video synch. No beep from the speaker. Holding down M on boot doesn't get you into monitor. Holding M and hitting RESET sometimes gets you into monitor but none of the keys respond. Just hitting RESET gets either a garbled screen or a solid cursor. 
I've checked voltages (5.02v on the 5V rail) and the EPROM. (Looked a bit dodge so I re-burned a BN55 ROM image onto the BOOT ROM)
Before I break out the 'scope and the logic probe and start probing around, can anyone suggest where I should start my diagnosis?

Cheers and thanks in advance!

(Of course it failed as I was getting ready to take it to a retro computer meet)

Chesh

First thing to do, as it sometimes boots to a cursor or monitor (by the way if you can get into monitor, it means the M key is detected at some point), is to 
take the coreboard out and 'clean' the 25 way pin strips with a fine file or emery board. Insert & remove the coreboard from the baseboard a few times after this is done
and see if that makes a difference.  I wouldn't bother replacing the boot rom.

[someone]

Do as Ewan suggested, although gold plated pins don't need filing.
The old thicker unplated molex connectors on the ROM coreboards may need filing if they have oxidized.

If that doesn't work, pop out the Boot ROM with a Test ROM.
None were officially published but it's easy to write your own.

[CheshireNoir]

Thanks someone and Ewan,
OK cleaned the contacts with an ink eraser. Also dropped a bit of Deoxit D5 into the socket, taking care to clean up afterwards.
Still no change so I reflowed the edge connector. (It was a bit gnarly and reflowing with fresh flux can't hurt)
Again, no change.
someone, you confidence in me being able to write anything more complex than some basic BASIC for the 'bee is greatly overgifted :-)
I do seem to remember you posted a Test ROM image in a post somewhere, but b&ggered if I can find it, now I need it :-D

Anyway I am off to demo my Educator 32K to a bunch of Amiga fanatics. Should be a good show!

Cheers everyone, and thanks again.

Chesh

[Ernest]

Hi Cesh,
When you get to scope stage, check out your reset circuit operation. 
Give particular attention to Cap C43.
I assume you have not changed to a different power source to what was operational before?
Changing the main input voltage too much can upset the reset.

I can loan you a test ROM if need be after your AMIGA group presentation, just call me.

Ernest

[CheshireNoir]

I blame someone, but now I'm starting to learn Z80 assembler so I can make a simple RAM tester ROM image.
Haven't hit assembler since uni in the early 90s.

Chesh

[ChickenMan]

There is a set of TEST Roms in the Repository, just burn them out

[CheshireNoir]

Awww spoil all my fun!
Googling around suggests I want ut212 so I'll burn that out and give it a go.

Cheers again. Man I am learning from the deep end here.

Chesh

[CheshireNoir]

OK Folks, got a chance to do some more testing.
Now dropping in the ut212 test ROM, I'm still getting no video, same as with the default ROM. Hitting the reset button I get a really flickery version of the boot screen.
https://photos.app.goo.gl/VUnEd4nNjbxNiCqk6 <- Animated! 5MB
or
    <- this is flickering like the display draw is being interrupted. It's more visible inn 
I've done some probing around and I'm now suspicious of the reset circuit. (Thanks Ernest)
I've checked C43 out of circuit, and it's a little on the high side at 715nF, but is within ±20% that caps tend to be rated to.
On the scope on pin 18 of X5 (RESET) I see it only go high as far as I can tell instantaneously (Or at least with a VERY short period). I would have expected it to stay low for a measurable part of a second. 
Should I be seeing more activity on the NMI pin? It stays low unless I press the reset key. Then is goes high and back down again. Nothing else is generating NMI calls?

I have the schematics out and I've been chasing my tail around with this for a bit now. What should I be checking next? (If you can't tell, I'm not exactly a whiz at this level of electronics. I know enough not to hold the burney end of the soldering iron)

Thanks for putting up with my ramblings. Happy to test / desolder and test out of circuit / socket and replace whatever people can suggest.

Cheers!

Chesh

[MbeeTech]

What happens with this cap usually is that it develops a low resistance (comparatively) so the reset cct doesn't end up pulling high enough and stays within the noise margin
of the following gate. Replacing the cap is a great idea even though it might measure within cooee of its rated capacitance.