Amiga 3000 help needed - no video output

[nvana]

I just cleaned an Amiga 3000D I received of a friend of mine. Machine looks to have survived the time quite nicely, but the battery had leeked a little bit (see photo).

The machine seems to try to start, it greets me with the dark grey, lighter grey and white screen, and then it stalls No floppy clicks or other activity.

I measured the power supply outputs and they seem good. Kickstart ROMs installed are 390629 and 390630.

Also trying 're-seating' the ZIP memories and swapping FAST & CHIP, but no avail, same white screen greets me.

I've got the feeling it must be the Denise having problems showing a picture, but how to test that ?

Any input greatly appreciated

[Liquidsnake]

The vga-port is maximum 31khz isn't it ? What kind of monitor are you using ? New lcd or old crt ?

[nvana]

I have a Commodore 1942 connected to the 23 pin video port (the monitor works fine with my A1200).

BTW I started measuring tonight with an oscilloscope on the pins of Denise. and all of them do get signals. The RGB output pins do give 5V during a whole scanline, so the white screen actually comes from Denise.

On the 16 bit data bus I do see data flying along all the time. I guess the Denise must be connected to the Data bus of the complete computer and the Register part of the chip via some logic to the Address bus ?

I was even challenged enough tonight to try my luck with A1200 ROMS in the A3000 mainboard, but that didn't improve the situation, it only showed a black screen.

So after testing I can conclude that the A3000 mainboard system 'Denise -> D/A -> Video port' works as far as I can see.

Somehow the system seems to stall in the boot process.

[source]

Dumb question, but do you have the daughter board installed in the motherboard. If I remember correctly, the computer won't work without it installed.

[Shadowfire]

It could be hanging up trying to boot off of the hard disk. Wait for a minute and see if you get any further.
The A3000 will attempt to boot off of scsi device #0.

[Hewitson]

The 74ALS74 and a number of tracks surrounding it don't look too good. Clean the pins on the chip and do some continuity testing.

[nvana]

I had the machine 'hanging' on the white screen for about 15 minutes and still no picture. I concluded that it wouldn't be waiting for the harddisk.

I tested with and without the daughter board installed in the mainboard. Didn't notice any difference.

I will measure the 74ALS74 IC today. Yesterday I did measure the 3 wires right next to the 74ALS74AN, the ones with the vias, and these wires and vias still conduct fully.

Is it a common problem that an A3000 stalls on the white screen in the boot process after battery damage, or should I seek the problem somewhere else ? Maybe one of the CIAs ?

[alexh]

Originally Posted by nvana

Kickstart ROMs installed are 390629 and 390630

Some A3000's have four ROM sockets, if the kickstart chips have been inserted into the wrong sockets OR they require a ROM tower which has been removed then this may be your problem. Which socket is each of your kickstart chips in?

What is the -xx part number of each kickstart ROM? Those numbers could be Kickstart v1.4 boot-roms which load kickstart from floppy disk or on-board SCSI controller hard drive into RAM (a kind of maprom). They only work with the integrated 68030, you cannot have any accelerator fitted.

Have you tried disconnecting the hard drive? With nothing on the SCSI bus it will eventually go to a grey screen with a selection



I think that when the battery dies / is removed that the A3000 battmem defaults to the maximum SCSI timeout so it may take several seconds?

The corrosion doesnt look good in the bottom left-hand side and top middle of the photo you sent.

[nvana]

@axelh: I expected such a Kickstart Menu too, but even after 15m it doesn't show. The ROMs don't say much more than: 'ROM_0 CS=9713' and 'ROM_1 CS=9B21'. Probably the checksums of the ROMs.

[nvana]

BTW, I tried swapping the CIAs but no avail.

[nvana]

Last weekend while cleaning the system I also found this issue on the A3000 board. Is it common to have a blob of tin on 2 of the 68030 wires ? Or is it a last minute fix of Commodore ?

[alexh]

Does your A3000 have four kickstart sockets? (U180, U181, U182, U183)

What sockets are your kickstart ROMs in?

U180 & U181 are for standard (27C400 style pinout) ROMs.

Have you checked the jumpers to make sure they are all set correctly? Maybe this A3000 had an accelerator in the past?

[nvana]

@alexh: This A3000 has 4 ROM sockets. They are in socket 182 & 183.

The ROM type setting jumpers (180&181) are both on 2-3.

The machine actually starts a piece of the kickstart ROM code I think, because the screen goes grey and 1 second later white during the startup process. When removing the ROMs, or swapping them with A1200 ROMs, the screen stays 100% of the time black.

The machine has never had an accelerator or any extensions added except a Piccolo videocard. I could verify that from the dust buildup in the CPU slot ;-)

The machine has been put away in working order so many years ago.

[alexh]

Originally Posted by nvana

This A3000 has 4 ROM sockets. They are in socket 182 & 183.

That seems to hint that they are KS1.4 boot ROMs.

The ROM type setting jumpers (180&181) are both on 2-3.

J180 & J181 need to be both on 2-3. That setting is for 32-bit ROMs in both sockets.

Originally Posted by nvana

swapping them with A1200 ROMs, the screen stays 100% of the time black.

AFAIK A1200 ROMs require a ROM tower to be used on this A3000 via sockets U182 & U183.

[nvana]

Yes, I've seen messages about needing a ROM tower (which I don't have). I guess I should look 27c200 eeproms.

About the blob of tin on the CPU pins; on photos on the net I could not find any other A3000s which also has this. I guess I should remove it ?

[mark_k]

Originally Posted by nvana

About the blob of tin on the CPU pins; on photos on the net I could not find any other A3000s which also has this. I guess I should remove it ?

I don't know who put that blob of solder there, but pins 1 and 2 (the two bridged pins) are both shown as GND in the 68030 User's Manual. So that blob of solder is not doing any harm.

[nvana]

I removed the solder tin of pins 1 and 2. Both are connected to ground indeed so it didn't improve the boot problem of the A3000.

[alexh]

If you remove the chip-ram ZIP's do you get a GREEN screen?

If it does go green, I think that would validate your kickstart ROM and your CPU.

• Dark grey - Hardware working and the registers are readable.
• Light grey- ROM verified.
• Red - Bad kickstart-ROM
• Green - No chip RAM found, or it is damaged
• Blue - Custom chip problem (Denise, Paula, Agnus)
• Yellow - Mostly a bad CPU (no trap routine) or a bad Zorro expansion card. CPU exception error before the "guru meditation" trapping software was enabled.
• Light green - CIA problem
• Light Grey - If it stops at grey, the CIA may be defective
• Black/stripes - ROM or CIA problem
• White - Initialization is alright. Ready to boot.

[nvana]

Thats a good question. I didn't try that upto now. I will try removing the memory tomorrow. Getting out of options to repair this miggy.

I did check the traces around the battery leakage and all are stil connected.

[nvana]

I tested today with the A3000 without any of the memory zips installed.

The result is exactly the same; the machine power on; black -> grey -> white screen and nothing more happens.

I don't get the green 'No Chip RAM found' screen. Somehow the Kickstart ROM doesn't even notice the memory is gone.