MTX 512 Black screen (of death?)

[JonB]

Well, I'm having a less than brilliant day.

Preparing an MTX for sale and have just discovered it's not working properly. It previously had a Rememorizer attached and was fine - now I've pulled the RM roms etc and refitted the originals. I'm getting a black screen on the composite video out - there is a signal, my monitor is detecting it, just no "Ready" prompt. I have not yet tried CTRL-G (that makes it beep via audio, right?) or checked voltages (but I expect these to be OK).

Where to start?

[Dave]

You said it yourself

You need to connect audio and try Control-G to see if the system is running (it does not sound like it is though)

[JonB]

Doesn't work Dave. I get a sort of constsnt tone from the sound chip, like it's uninitialised.

Have you got a picture of the ROM chips, please? I need to be sure I inserted them in the right positions. (Actually no need, I see one on primrosebank.net).

[Dave]

http://primrosebank.net/computers/mtx/p ... 5_2000.jpg

if it's a 3 ROM board, B, A,, C reading from top to bottom

Yes, "constant tone" is the usual machine dead symptom, could be anything, but the ROMs is a good start as you've had them out.

[JonB]

It’s been so long since I did anything with this machine.

The ROMs look to be fitted correctly. It’s a 4000-06 board. Looks like I’ll have to connect the Z80-ICE and check to see if there is any RAM error. Is the RAM all accessible when it boots, or does it switch? What I need is an initial memory map.

[Dave]

You did not fit a new PAL did you?

Andy sometimes supplied replacement PALs that changed the memory MAP

(Is it a MTX500 or a 512?)

[JonB]

It's a 512. I think Andy gave me a PAL for the RS232 card which mine does not have. That went with the Rememorizer to Lardo Boffin. But I was wondering about the PAL and whether or not I changed it. I don't think so... how best to tell? Mine has a red label on it (like a disk). It looks the same as the one pictured on your web site.

[Dave]

The Red dot label sounds original.

The only obvious quick check is to remove/reseat the ROMs, looking for bent pins and making sure that they are sat to the right hand side of the sockets, rather than at the left. If that is OK, then try reseating the other socketed chips.

Then you're into checking voltages, looking for the clock etc. Martin has looked at the ROM code previously and silencing the sound chip happens really early in the boot sequence. Bad RAM could cause this, but so could many other things. I don't think that the RAM "switches" during the boot - other than the normal paging in and out. The early RAM check probably checks everything that's there.

Martin will probably be able to shed more light on the boot sequence

[JonB]

No bent pins on the ROM chips.

I've tried reseating them all but it's not made a difference.

Is there a ROM disassembly anywhere?

Looks like it's got a full memory map:

Code: Select all

OK ===> md

0000000000000000111111111111111122222222222222223333333333333333
0123456789ABCDEF0123456789ABCDEF0123456789ABCDEF0123456789ABCDEF
RRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRR
OOOOOOOOOOOOOOOOWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWW
0000000000000000111111111111111122222222222222223333333333333333
0022446688AACCEE0022446688AACCEE0022446688AACCEE0022446688AACCEE

That's a good start. I can see some code at 0000:

Code: Select all

0000 F3             DI
0001 AF             XOR A
0002 21 00 40       LD HL,4000
0005 C3 94 01       JP 0194

Then..

Code: Select all

0194 2B             DEC HL
0195 11 BC 23       LD DE,23BC
0198 B9             CP C
0199 12             LD (DE),A
019A 00             NOP
019B 20 83          JR NZ,0120
019D 0E C1          LD C,C1
019F 13             INC DE
01A0 D1             POP DE
01A1 12             LD (DE),A
01A2 91             SUB C
01A3 23             INC HL
01A4 9E             SBC A,(HL)
01A5 12             LD (DE),A
01A6 86             ADD A,(HL)
01A7 12             LD (DE),A
01A8 83             ADD A,E
01A9 11 A5 23       LD DE,23A5
01AC 0B             DEC BC

..and so on. Not entirely sure this is correct, it does not look as if it makes too much sense - CP C without checking the result at 0198H for example.

Wonder if Martin's about...?

[Martin A]

He is, and that looks nothing like the rom code that should be at 0194

Code: Select all

                            .initial
0194        32 D2 FA        LD   (PAGE),A
0197        D3 00           OUT  (&00),A
0199        08              EX   AF,AF'
019A        06 08           LD   B,&08
                            .fill
019C        70              LD   (HL),B
019D        23              INC  HL
019E        10 FC           DJNZ fill
01A0        01 01 08        LD   BC,&0801
                            .test
01A3        2B              DEC  HL
01A4        7E              LD   A,(HL)
01A5        A9              XOR  C
01A6        20 09           JR   NZ,zeroz
01A8        0C              INC  C
01A9        10 F8           DJNZ test
01AB        08              EX   AF,AF'
01AC        3C              INC  A
01AD        FE 10           CP   &10
01AF        20 E3           JR   NZ,initial
                            .zeroz
01B1        21 00 C0        LD   HL,&C000
01B4        11 01 C0        LD   DE,&C001
01B7        01 FF 3F        LD   BC,&3FFF
01BA        AF              XOR  A
01BB        77              LD   (HL),A
01BC        ED B0           LDIR

Full source build enclosed I can't see any obvious pattern of stuck bits etc.

RomBuild.zip

Full ROM Source

(160.46 KiB) Downloaded 978 times