MTX 512 Black screen (of death?)

[JonB]

Thanks Martin

So something seriously hokey is going on here.... .. or the ICE is duff (possible, I've had issues with it before).

I have another truly ancient device that might prove it one way or another. A Zicon "logic analyser" which cna monitor the pins of the Z80 and capture its instructions as they are executed. It can show a disassembly... I haven't used it in years, but I'll try it.

[Martin A]

If you have a external rom board, or can knock up one, or a ROM to EPROM adaptor, then there's the diagnostic rom.

viewtopic.php?f=17&t=420 The "final" binaries are on page 3

That works without ram, though since it replaces ROM A in diagnostic mode it can't check that.

[JonB]

That's jolly useful! Can I put the 8k ROM image directly in ROM A's slot and run the diagnostic tests without building a test board? Reading the thread it appears so. But you mention and adapter, so what is the schematic for it?

And secondly, what types of ROM chip are acceptable to the MTX?

And thirdly, are there any MTX 512 schematics? I can work it out from those.

[Dave]

The MTX schematics are in the manual, available here
http://www.primrosebank.net/computers/ ... anuals.htm

Or get just the schematics here http://www.primrosebank.net/computers/ ... hPages.pdf

(There is no schematic available for the 4000-05/06 boards)

[JonB]

Super, thanks Dave

There's a definite pattern in my ROM:

Code: Select all

OK ===> d 0

0000   F3 AF 21 00 40 C3 94 01-5E 23 56 23 C9 FF FF FF    ..!.@...^#V#....
0010   E3 F5 7E FE 40 C3 FA 06-C3 74 3B D7 2D 0A C9 00    ....@....t;.-...
0020   F3 AF 21 00 40 C3 94 01-5E 23 56 23 C9 FF FF FF    ..!.@...^#V#....
0030   E3 F5 7E FE 40 C3 FA 06-C3 74 3B D7 2D 0A C9 00    ....@....t;.-...
0040   F3 AF 21 00 40 C3 94 01-5E 23 56 23 C9 FF FF FF    ..!.@...^#V#....
0050   E3 F5 7E FE 40 C3 FA 06-C3 74 3B D7 2D 0A C9 00    ....@....t;.-...
0060   F3 AF 21 00 40 C3 94 01-5E 23 56 23 C9 FF FF FF    ..!.@...^#V#....
0070   E3 F5 7E FE 40 C3 FA 06-C3 74 3B D7 2D 0A C9 00    ....@....t;.-...

The first 32 bytes are correct, they match the normal ROM, per Martin's ROM dumps. But thereafter these bytes just repeat themselves until 0100h; whereupon the same thing happens with 16 bytes from 0100h.

So what this suggests to me is "stuck address lines". But I'm going to check with the Zicon anyway, because my Z80-ICE is flaky sometimes.

[Dave]

Although there is no schematic for the ‘05 board, you might find some useful info here http://www.primrosebank.net/computers/m ... eplace.htm

[Martin A]

The schematic is printed in the back of the user manual. The rom wiring is one Pg214 of the original or pg226 of the Phoenix version.The 2 rom 04 board roms are detailed in both, only the older version has the 3 rom 05 too, nether has the -06.

The techincal manual however suggests the -06 is the same as the -05 as the MTX512 uses the same red dot PAL.

As Dave points out, both user manuals and the technical manual are on Dave's site if you don't have the paper version.

The boards are wired for mask roms, so won't take a an 8k (e)eprom directly. The onlythe 04 has links so can be tweaked to take 8,16 or 32k(e)eprom in the c settings.

The 05 can'tbe modded without cutting tracks.

[Dave]

The 05 can'tbe modded without cutting tracks.

Which is described in the link that I posted about (about replacing the obsolete mask ROMs with EPROMs)

I had forgotten that the older (rubbish) manual had the detail of the ‘05 board ROMs in it

[JonB]

Yes, it's an 06 with the red dot PAL.

I've connected the Zicon up (what ablast from the past) and amazingly it is still working.

The Zicon connects to the Z80 with a 40 way IC clip. Then a box, which is an adpter that allows it to work with Z80, 6502, 6809 or 6800 processors. On the other end is a large box that looks a bit like an oscilloscope. It's got a 9" CRT and some buttons.

IMG_8001.jpg (100.43 KiB) Viewed 17606 times

I set it to trigger when the address line is 0000h, so it captures events from that point. Then applied power to the MTX.

Disassembly of the start of the boot code.

IMG_7999.jpg (77.91 KiB) Viewed 17606 times

The Zicon reconstructs the running code (really, it is recording the buses and disassembling the instructions) and displays it. We can see from this that the issue detected by the ICE is still present; in other words, the ICE is not giving a false reading.

[Dave]

Wow!

That’s a really cool tool