ICL OPD 3 vertical white stripes fault

[Chr$]

SRAM issue I found on my machine was the lack of the POWER-GOOD from the power supply, listed as "shutdown" on the pinout.

Check VCC (from the 3V battery) and /CE (from PSU via MOSFET on the left of the board) on the SRAM chip.

Thanks. I'm not sure if my NOS board 3v battery is properly connected, so if it needs that it's very likely not getting it. That'll be easy to remedy!

I don't use the original monitor though, mine (both) have the video mod and are hooked up to an ATX PSU with breakout scart cable. And as it seems to and always has powered the other one fine, I suppose it must be doing thatg correctly.

[Chr$]

Back to the one with the bootloop...

I discovered that the bootloop is the same whether the ROM pack is plugged in or not, so does that mean it's somehow not communicating with it? i.e. that it's died. I've re-touched all the solder joints and short of removing and testing all the 4 LS244/LS245 I'm not sure what I should do now. I could just buy another and try it, but that pesky battery acid has me in 2 minds.

[Chr$]

SRAM issue I found on my machine was the lack of the POWER-GOOD from the power supply, listed as "shutdown" on the pinout.

Check VCC (from the 3V battery) and /CE (from PSU via MOSFET on the left of the board) on the SRAM chip.

The SRAM chip has battery 3v on pin 24. There are 2 /CE pins. One of them has battery voltage when the computer is off, and that one goes low when switched on. The other /CE pin goes high and stays high when the computer is turned on. The connection to the other side of the board, the MOSFET and then up to a pin on the power connector seemed fine. Compared to the other OPD that all seems normal.

I also swapped the TC5516AP SRAM chip and that made no difference.

Back to the other OPD - It really doesn't seem happy with certain combinations of ROMs (EPROM or PROM) despite the same part numbers. A 'mismatch' results in 3 white stripes or just a black screen. I wonder if there were changes to the code and that perhaps they don't all work with each other. It's odd that with the original ROMs I get 3 stripes and with 2 borrowed in positions 6 and 7 I get it almost to boot (with the bootloop). I'd love to obtain bin files of a working suite of daughter board ROMs (and also the ROM pack ROMs). Just to rule those out and to make known working spares. I know there is a checksum check, but it just seems odd that the original ROMs stopped working and that it almost boots with the replacements.

[Pr0f]

The ROMS are I think in 3 parts - the first 64K are a combined OS ROM - so that's ROM A and B

ROM C has a header like other application ROM's - so I wonder if this OS Extension is handled as an application ROM - ROM D seems to be the one that may have different code in it and seems to be mostly provided as an EPROM.

Although looking at the start of ROM C - it is configured as a 64K ROM - the D ROM seems to have a lot of references to Microdrives in it.

[Chr$]

The ROMS are I think it 3 parts - the first 64K are a combined OS ROM - so that's ROM A and B

ROM C has a header like other application ROM's - so I wonder if this OS Extension is handled as an application ROM - ROM D seems to be the one that may have different code in it and seems to be mostly provided as an EPROM.

My almost working ROM D (IC7) is a mask ROM. Originally it was an EPROM but that doesn't seem to work now - with the original back in I get the 3 white stripes. Interesting that ICs 6 and 7 look like application ROMs then and that they don't seem to cooperate with the ROM pack plugged in, i.e symptoms unchanged. Hmmmm. I've ordered a replacement ROM pack from RWAP in case I have killed mine with ESD or some other freak accident - I did have it out while I was assembling the C1 disk interface, just to make sure my bodge connector would fit through the standard pack hole etc.

[stephen_usher]

If the ROMs were bad you'd get four stripes as they'd fail their CRC checks.

I wonder if a capacitor has gone bad and you're getting noise on the power rail. Something's very screwy.

[stephen_usher]

By the way, if the second /CE stays high then whatever is supposed to bring it low so that the CPU can read it isn't doing its job. Time to trace that back on the board to find where it goes.

[Pr0f]

I think that comes from the drain of FT5 - controlled by the powerdown on the 15 way connector, that FET will ground the /CE input when powerdown is at 5v (this signal goes to multiple places) - the drain has a load resistor connected to the diode junction from 3v lithium, and diode from 5v, so either of these will hold this /CE high unless powerdown is at 5V, causing the FET to ground it.

[stephen_usher]

I thought that he meant the other one. Anyway, SRAM error is 5 stripes.

[Pr0f]

The other one comes from the NAND gate - with FC2 and the CMOS RAM enable from the ICL ULA as the inputs to the gate - so the RAM can only be accessed in supervisor mode.