Busted CX5MII mapper

Ascended (9580)

24-05-2014, 20:26

Hey,

I accidentally left my CX5MII on and unattended for an hour or two, and when I got back it was in full sunlight and had gotten really hot. So I quickly shut it off.

However when I restarted it a little later, DMS2 no longer worked! It gave weird sounds when booting and crashes. Also when I unloaded DMS2 from the MegaFlashROM SCC+ SD 512k, it seems to work fine at first but booted in MSX-DOS 1 in stead of MSX-DOS 2. And when I tried _MUSIC it gives me a blue screen with full white background and then crashes, and _FMV just crashes without any feedback.

Because MegaFlashROM SD didn’t boot in MSX-DOS 2 mode any more, my first suspect was the memory mapper. And indeed, this is the result of a test:

10 ? PEEK(&H8800)
20 POKE &HC800,47
30 ? PEEK(&H8800)
run
255
47
Ok

Also after OUT &HFE,1 or any other value, the result is the same. And when I e.g. do OUT &HFF,3, it should crash because I swap out the system memory, but nothing happens.

Finally, the result of INP(&HA8) is 240 (11110000B), so the internal RAM in slot 3 is selected, apparently the CX5MII’s (modified) MSX1 BIOS does not search for RAM otherwise it would’ve picked the RAM in the MegaFlashROM.

So now that we’ve established the internal memory mapper is busted, can someone help me figure out how to find out which component is damaged and how to replace it?

p.s. The CX5MII service manual is here if you need to take a peek.

By maxis

Champion (512)

24-05-2014, 21:56

Hello Grauw,

Which version of the motherboard do you have? Is it a German version (no RTC clock routed) or universal version of the M/B (503/II, CX5MII, CX7)?

Also do you have any test equipment, like a simple scope or a logic probe?

Yamaha machines can support quite an extreme heat and these are virtually undestructable. You have already tried to re-seat the memory mapper daughter board, right?

By Grauw

Ascended (9580)

25-05-2014, 00:19

Hey, thanks for helping out . It’s the CX-5MIIB, motherboard says YIS 604, CX7, CX7M, LC56770, YAMAHA.

I have a multimeter and I know how to handle a soldering iron, but I have little actual clue about electronics.

I just took off the RAM board and put it back on, but it didn’t fix the problem.

Pics: 1, 2, 3, 4, 5, 6, 7, 8 (sorry bit blurry ’cause it’s night).

By maxis

Champion (512)

25-05-2014, 00:58

1. Can you r/w the mapper ports, for the exception of the 0xFF port? Do they follow your command?
2. If the mapper r/w follows your command, then IC403 (74LS670 - mapper register file), IC405 (74LS125 - read buffer) are ok. But the problem still could be in the IC 401 and IC404 (bit 14/15 multiplexer and MSB address multiplexer) with processing bit14.
Do you have around a spare 74ls157 or 74HC157 or 74AC157?

Please, publish the results.

P.S. BTW, cool pics. I've grown up on 503/IIR :-)

By Grauw

Ascended (9580)

25-05-2014, 00:58

Hey Maxis, yup 64K version.

Do the tests in my original post cover it? If I write to FFH, nothing happens (no crash). If I write to FEH, nothing changes either, 8000H-BFFFH still mirrors C000H-FFFFH.

Basically, all mapper pages seem to be set to the same value, mirroring the same mapper page in every page, and writing values to the mapper ports seem to have no effect.

I only tested mapper ports FEH and FFH, harder to try the other two from Basic but I could make an assembly test program if needed.

By maxis

Champion (512)

25-05-2014, 00:59

No, no, try just to write values to 0xFC, for example and read them back

By Grauw

Ascended (9580)

25-05-2014, 00:59

maxis wrote:

P.S. BTW, cool pics. I've grown up on 503/IIR :-)

That must be pretty much exactly the same machine, right? I can’t understand if they used the CX7M board, why they didn’t go all the way and just produce an MSX2…

By maxis

Champion (512)

25-05-2014, 01:02

If you can write to 0xFC port the values 0xFF, 0xFE, 0xFD, 0xFC and read the same back, then IC403 and IC405 are OK, which are the good news already.

By maxis

Champion (512)

25-05-2014, 01:10

Well, IMHO, they didn't use 503/III board for example, since all the music SW was originally written for the MSX1 style GUI. And 503/III didn't have the ROM cartridge connector populated. Still don't know whether CX7M was sold in Germany, due to absolutely crazy EMC requirements.
Also Yamaha produced quite a few completely different models in the same plastic body.

By Grauw

Ascended (9580)

25-05-2014, 01:10

maxis wrote:

No, no, try just to write values to 0xFC, for example and read them back

maxis wrote:

If you can write to 0xFC port the values 0xFF, 0xFE, 0xFD, 0xFC and read the same back, then IC403 and IC405 are OK, which are the good news already.

Got it! Whatever value I write to ports FCH-FFH, they always return 255.

maxis wrote:

Do you have around a spare 74LS157 or 74HC157 or 74AC157?

No…

By Grauw

Ascended (9580)

25-05-2014, 01:12

maxis wrote:

Still don't know whether CX7M was sold in Germany, due to absolutely crazy EMC requirements.
Also Yamaha produced quite a few completely different models in the same plastic body.

I think the CX7M was only produced in Japan? It’s a nice body, btw .