Troubleshooting STFM with white screen

[Slartibartfast]

Hi,

I started this thread in the Introduction section after saying hello for the first time but thought it best to continue in the correct forum.

I'm new to the Atari ST having spent most of my youth using 8 bit machines like the C64 and BBC B but I have recently obtained a 520STFM in pretty good condition apart from the fact it doesn't start.

Apart from the power LED being on, all I get out of the machine is a white screen. There is no floppy activity even with a disc inserted and the reset button seems to have no effect.

I've checked the power supply and it's outputting 5.127V & 11.8V. The 5V rail does have around 100mV of noise on it so I will try and boot up using an external bench PSU in case that's the issue.

The motherboard version is C103414-001 revision 1.1 so all the chips apart from the 6 ROMs are soldered in making chip swapping to diagnose rather difficult.

I've checked that the CPU is receiving the 8MHz clock and that all the ROMs have 5V and ground. I also have reseated them all.

I've identified that the CPU HALT pin is low so the CPU is stopped and I've also identified that whilst the BERR line to the CPU is high, it does show a single low pulse when turning the machine on which I think indicates a Glue chip issue.

I plan to check over the RESET circuit next but it's a bit difficult as all the schematics I have are for different boards and seem quite different to mine in layout and component identification. Does anybody know if a schematic for this particular board revision exists?

Given the overall 'dead' nature of the machine, would a diagnostics cartridge help at this stage?

Any other ideas on what to try?

TIA, Brian

[Badwolf]

Probably best to start with what tools you have available for the diagnosis.

Logic probe you've mentioned. I assume multimeter. Scope? Logic analyser?

Does the ROM CE line ever flash low after reset?

BW

[Slartibartfast]

I’ll check the ROM CE line.

Yes I have a multimeter & 4 channel scope. I also have access to an old logic analyser if it comes to it…

[Badwolf]

I’ll check the ROM CE line.
Yes I have a multimeter & 4 channel scope. I also have access to an old logic analyser if it comes to it…

The first eight bus accesses after the reset line goes high are all to ROM.

Specifically the 68000 reads four longwords starting at 0x000000. These four longwords are mapped to ROM by the GLUE asserting the ROM CE line.

Therefore, with a four channel scope, I'd probe the following:

1) RESET on the CPU. Trigger on positive edge
2) AS on the CPU. This is the address strobe line and the most obvious indication of a bus cycle starting
3) ROMCE, from either of the ROMs, switching if one seems fine. This may be called ROM1 and ROM2 on the schematic -- memory is hazy with the ST
4) DTACK on the CPU.

You should see about 400ms, IIRC, from RESET going high to AS starting to go low. AS will go low, shortly followed by the ROMCE line, then around 125ns (IIRC) later by DTACK.

This should repeat at least 8 times. If those all work, you'll see more AS and DTACK accesses but ROMCE will become intermittent as some will be to RAM.

If that's the pattern you get, great. Change the ROMCE line to BERR, RESET to HALT and see how many cycles you get from HALT going high to BERR strobing and HALT going low again.

If not, however, then you could have a problem with GLUE or a trace.

BW

[rubber_jonnie]

Though I answered over on your other post, the diag cart can be very useful, but it also can leave you dead in the water.

I have had machines that have a white screen, but the diag cart does not work at all, even via the serial port, so there are no diagnostics to be had.

However if you intend on having multiple machines that you may need to repair, it's definitely a bonus to have one, and it's useful for post repair testing.

I've used them for finding RAM issues and diagnosing floppy issues too, so for the price they are definitely very useful.

[stephen_usher]

Indeed. The diag ROM requires the machine to be able to access the ROM and run the first few instructions. If the MMU isn't enabling the ROM then you get no-where, but this also points you at the first fault to look at, so it's part of the useful diagnosis process.

Seeing the equipment you have though you can get a long way before needing the diag ROM just by seeing if the ROM gets selected (logic probe) and looking at the address and data buses when powering on to see if there's any activity (oscilloscope). No need for a logic analyser unless you really get down the rabbit holes.

Oh, and a desoldering station and RAM tester are always good things to have ready... :-)

[Slartibartfast]

Thanks for the advice folks.

Unfortunately my investigation will have to on hold for a few days as I’ve ended up getting Covid and it’s laid me out in bed at present.

I’ll let you know how things progress once I’m better.

Apologies for asking the ROM cartridge question twice, you’re right, it was answered in the other thread.

[stephen_usher]

Get well soon. There's a lot of COVID-19 about at the moment and hospitalizations are rising fast. "But it's all over".

[Badwolf]

Aye.

I had it a couple of weeks back. I'm sure I must have copped it before (symptoms were basically that of a head cold), but that was the first time I'd come up postiive.

Felt better after about four days. Clear in a week.

Good luck.

BW

[rubber_jonnie]

Hope you feel better soon.