A3000 Fails Caps-Lock Test, won't boot.

t-bus

Member
Joined
Jun 30, 2013
Posts
77
Country
USA
Region
Pacific Northwest
I picked up a 3K last weekend As-Is for a reasonable price, under the assumption that the chips would provide great spares for my good 3K.

Thing is, I think it might be repairable. I'd love to have turrican in the office!

Boots to a black screen, no disk activity. Fails the Caps-Lock test so the CPU ain't running.

I cleaned up the battery leakage and continuity checked the tracks in the area.
There seemed to be some damage, but the traces beeped-out ok.

Anyone got ideas?

Thanks!
 
Lots of work ahead of you:)...

If you have the proper equipment and a schematic, you can test and find the bad trace (as this is most likely the culprit.

I haven't found a better way to diagnose a bad 3k motherboard.

Sorry I don't have better/encouraging news, but it is lots of work...

You can always remove all the socketed chips, clean the MB with IC cleaner, and reseat then hit solder points. To see if your really lucky...
 
Last edited:
Checked some of the traces by the ricoh and co, seem to be OK.
Reseated and replaced a lot of the custom chips, still nada.

Got the board down next to the Scope... any ideas on where to start probing?

Cheers.
 
Hello,

you can start to check all the clock signals, that mean "Fat Gary" and "Agnus". Then, reseated the ROMs and check if all the pins of the MC68030 are soldered on board.

Regards
Stefano
 
Check if the tick signal is coming from the power supply (move the jumper J350 to the other position) then try to boot again.
 
necromancing this thread-

I've got the bug bad-- I started in December with one A3K, bought a broken 16mhz machine for spares, and then bought *another* 3K.

Well, the two 3K / 25 machines work now (other than the fact that I need to wait ~10m after cold boot for one of the machines to see its hard drive...)

So I've been playing with the last broken machine-- the 16mhz box I picked up for spares.
Me, it, my Tek scope, and a 6-pack of beer had a session today.

I've narrowed down the issue to the _KBRST signal staying low forever... This appears to be Gary's fault. I tried switching gary at the beginning of my efforts on this board, but I suspect BOTH garies I've used in testing may be dead (the spare I had, and the one in this board)-- as odd as that may be.

I will try and hunt down another... KBCLK, Gary's clocks, and VCC / GND are all correct and as expected.

I don't feel like taking apart my known working A3ks to test with... So I'll provide another update to this thread after I manage to purchase another fat gary.

I originally thought U713 went bad and was sinking _KBRST, but I now know that to be incorrect.
 
Last edited:
Well, I swapped in a known good GARY. Even with the good gary, _KBRST is always LOW. WTF... Does anyone know the signals Gary needs to not assert reset continually? the problem must be coming from elsewhere in the system.
 
Start by checking the usual suspect: the 556 timer chip.
 
Output from timer probed at VDE14M and VDE28M look like good and clean timing signals on my scope.

Problem is definitely related to gary keeping _KBRST low, but all the clocks going into gary seem fine.

Any other ideas?

Here's what I know as absolute fact:

* _KBRST is held low keeping the system in RESET
* _FAIL is high, so the system should be attempting to start
* Timing signals from VDE14M, VDE28M, CPUCLKA, CPUCLKB, CLK30, CLK90 seem OK
* No ROM accesses are attempted
* Gary is known good
* U713 has been swapped, not the problem
* CIA, AGNUS swapped
 
Last edited:
Bad via on the 7MHz clock! It was getting generated ok but not making it to Denise or Gary. I didn't notice because when I probed Gary I saw another 7mhz clock- the 90deg phase shifted one, not the one needed for the initialization. I did a bo l dge from the resistor after where the clock comes out of Agnes, into the 7mhz input on Denise. The connection between Denise and Gary for that clock was still good. Now everything works!
 
Back
Top Bottom