Amiga 2000 (6.2) - Repair - Stuck

[misterdid]

Hi,
I'm trying to repair a nice (non-working) Amiga 2000HD I bought on ebay a few days ago...

The battery leaked a little bit, so I removed it.Then I used some vinegar to stop the corrosion, then rinced the board.

First, while reseating all the chips, I discovered that the corrosion totally destroyed some of the 68000 socket pins . I ordered a new socket and changed it.

I tried to fire it up...and I got this very strange screen. Black with some strippes (on the RGB output - I've a VGA LCD compatible with 15Khz, it works fine with my A500 [I bought it in 1988 !] -). The composite output is only black but with a 60Hhz signal.

I did some continuity tests and I discovered that D2 and D4 on the ROM were not connected. Two traces were dissolved between the ROM and the CPU. I replaced them with 2 wires. But the problem is still there...

All the Dn and An lines are now connected (to the ROM).

I think it's very early in the processor startup sequence because if I plug the Vector2030 card I had with it, I've exactly the same screen...I tried to put another 68000P8 and I've the same symptoms.

The processor run, I've power and a clock (I used a logic probe, and the clock is pulsing). The ROM is powered as well. RESET is low and then high when powering up the computer. It's slightly warming.

Any idea ? Hint ?

[Franchute13]

Hi!
welcome!
You have another Amiga? like 500? if YES, check all chip in Amiga 500.
You can check trace using datasheet of Amiga 2000.
https://www.amigawiki.org/dnl/schematics/A2000_R6.pdf
Upload photo in good quality to see the motherboard.
Bye.

[misterdid]

Yeah...that was my next plan for tonight even if I hate the idea of removing the chips of my beloved A500...
And yes, I used the v6 schematics to test most of the CPU pins to Agnus, Gary, ...
It looks like this pattern is “CPU Fault” (black screen and red stripes)...but what fault

I’ll post a picture of both sides later...

[misterdid]

A little update...I disassembled my A500 and tested all chips (except Fat Agnus, I’m really afraid to remove it as it’s so tightly inserted). 100% working (CPU, workbench, Gary, Denise, Paula and the two CIAs).
I’ll post the motherboard pictures tomorrow.

[misterdid]

The pictures of the board...

[Franchute13]

Hi.
What happens with the keyboard caps light? Flashes?
What happens with the power led? Flashes?
What happens if you try to boot from a disk?

[misterdid]

Originally Posted by Franchute13

Hi.
What happens with the keyboard caps light? Flashes?
What happens with the power led? Flashes?
What happens if you try to boot from a disk?

I don't have a keyboard...yet...I'm trying to source one, but the prices are insane right now...I would pay more than for the computer.

The power LED is just green (not dimmed), no flashing.

The drive motor spin a little when inserting a disk or a power-on, but it's not reaching the state where it would look for the boot code...

[Wayne123]

Originally Posted by misterdid

I don't have a keyboard...yet...I'm trying to source one, but the prices are insane right now...I would pay more than for the computer.

Yes some people are asking crazy prices for keyboards, I used a cheap keyboard adapter with a PC keyboard until I found a price I was willing to pay.

I got my 4000 keyboard for $49 and I think my 2000 keyboard was $79, you just have to jump on buy it now deals, I have alerts for anything I am looking for on Ebay.

[misterdid]

Hi again...

I'm still stuck trying to figure out what is preventing my board to start...

Last week, I remove most of the coating around the place where the battery most likely leaked to verify with an USB microscope all the traces. I then replaced my wires for D2 and D4 by a small copper wire soldered on the traces.

I replaced the RP101 resistor network in order to remove the coating under and verify the traces (it was also a little bit corroded).

I replaced the capacitor close to the keyboard because it might had some battery acid as well...

But still no luck.

I also swapped all the ICs from my A500 to the A2000 (except fat agnus, it's really stucked in my A500 even with a PLCC puller). The A500 is booting fine, the A2000 is still not booting.

I checked the signals, HALT is high and imediatly low. I've continuity from the 68K _HLT to _HLT in Gary.

What could happen so early ? Any passive component to check ? (I can't see any other involved in the schematics)

Any idea, please...

[idrougge]

I encountered an A2000 with battery leak recently. The CPU socket had been replaced but the machine still didn't start. I redid the soldering on the socket and it came to life.

[misterdid]

Thanks. Anyway I will have to do it because yesterday I unsoldered the socket I replaced to check the traces under it (what I didn't do at first) and two pins broke ...So I order new machined turned sockets to replace it.

Now, I'm still trying to figure out what should happen at startup (the signal sequences) and what is HALTing the CPU. Because if all the chips are good, I don't understand what I'm missing that prevent it to at least start loading some code and playing with the screen colors at boot like it should...It's driving me crazy

The keyboard is not required, right ? (as it's driving the HALT signal in Gary)

I'm also wondering if the original CPU is required when using a CPU add-on card. Because when I plug the Vector 030 card, I've exactly the same symptoms.

[Dynamic_Computi]

I have a board like your, and did the same repairs and I still get a black screen. I am thinking the acid ate though the green protective coating on the bottom of my board near the ROM socket, as I see bare copper there. I suspect it it shorting a few pins on the ROM. I plan on recoating it (I hear clear nail polish works) and see if that stops the short. Maybe check for shirts there, too?

[Jope]

Just recoating it will not fix the short, it will just coat the short. You really need to debug the problem before reapplying the conformal coating.

[Daedalus]

Originally Posted by misterdid

I also swapped all the ICs from my A500 to the A2000 (except fat agnus, it's really stucked in my A500 even with a PLCC puller). The A500 is booting fine, the A2000 is still not booting.

Yes, be careful with that. I've had one A500 in which the Agnus chip was stuck in the socket by a sticky pad, and I've heard stories of other people finding the same thing. It's not common, but perhaps used by one production line or facility instead of the clip that used to secure the chip on older revisions. Anyway, you could remove the motherboard and poke through the holes under the socket to push the chip out instead, but it will still be quite tricky if a sticky pad was used.

I checked the signals, HALT is high and imediatly low. I've continuity from the 68K _HLT to _HLT in Gary.

What could happen so early ? Any passive component to check ?

Have you checked that the pull-up resistors are actually connected fully? You should be able to measure 2.7K ohms between _HLT on the CPU and +5V. If you don't, there could be a short or open circuit somewhere. But I suspect it's being thrown by the CPU or Gary due to a bus fault of some sort.

You don't need a keyboard, the machine should start to a Kickstart screen with nothing connected to the mainboard but the PSU and display.

[Toni Wilen]

Does CPU generate active HALT signal? If it is CPU, it means CPU detected double fault (which also means something kept BER line active which shouldn't happen normally). Note that HALT being temporarily active probably is normal because 68000 documentation mentions that 68000 external reset should be done by activating both RESET and HALT.

Does screen show same glitch if you remove the CPU? Glitch could be simply normal random Denise internal register contents being visible (but on the other hand they should be random, does the glitch change between power cycles?). Hard reset does not clear most custom registers.

[misterdid]

Originally Posted by Daedalus

Have you checked that the pull-up resistors are actually connected fully? You should be able to measure 2.7K ohms between _HLT on the CPU and +5V. If you don't, there could be a short or open circuit somewhere. But I suspect it's being thrown by the CPU or Gary due to a bus fault of some sort.

Yes, the 2.7K ohms are there.

Originally Posted by Daedalus

You don't need a keyboard, the machine should start to a Kickstart screen with nothing connected to the mainboard but the PSU and display.

Yeah...after writing that, I checked the schematics again and may be U805 could be faulty ? I'm planning (when I'll received the new socket and find some time) to plug the oscilloscope and trace the signals...

Originally Posted by Toni Wilen

Does CPU generate active HALT signal? If it is CPU, it means CPU detected double fault (which also means something kept BER line active which shouldn't happen normally).

I think...all I know with my probe is that HALT is high and them immediately low at power-up. As said above, I'll try to trace all the signals and see if there's something at least starting on the Data lines...But I've the feeling it's in HALT state from the very beginning...

Originally Posted by Toni Wilen

Note that HALT being temporarily active probably is normal because 68000 documentation mentions that 68000 external reset should be done by activating both RESET and HALT.

Yes, I read that. And I think it's what Gary is doing when it receives the KBRST from the LM339 to reset properly the 68K.

Originally Posted by Toni Wilen

Does screen show same glitch if you remove the CPU? Glitch could be simply normal random Denise internal register contents being visible (but on the other hand they should be random, does the glitch change between power cycles?). Hard reset does not clear most custom registers.

Well, It's almost exactly the same pattern with or without CPU (or with/without the vector card). And there's noise. I see a "wave" going down the red lines. And I've the feeling that the frequency of the noise is higher with the vector 030 card...

[misterdid]

Hi,

A little update on my A2000 boad...it works !

First, last week, I replaced both the 68K and the Kickstart socket (there was some corrosion on this one as well) by good quality machine tooled ones.

The symptom changed...I got 2 gray screens and then the video signal stopped...

I spend many time trying to figure out what was going on. There was good signals everywhere, just no video.

And then, yesterday, I tried to boot without my RGB to VGA cable, only composite....It worked !!! The "insert disk" of the kickstart showed up !

Hooking up the boad to the GBS8200 I used in the paste give a decent image

So, my direct connection cable to my 15Khz doesn't work with the 2000 (while it's great with my 500). I read that the 2000 can take this connection for a genlock signal and drop the video...I will make a new cable using a 74SL08N on the Vsync/Hsync and see...

Thanks all for your support !!!