NMS 8245 black screen repair

Page 2/3
1 | | 3

By marcoexo

Resident (62)

marcoexo's picture

06-12-2017, 08:37

Thanks for the test Pentarou! I was just reaching the same conclusions on a faulty RTC but I didn't know booting without RTC was not possible!...
It looks just as if the boot program is stuck in an endless loop, and the RTC was my first suspected part,
I already checked the RTC address lines and they all remain high, I initially thought that an RTC fail could probably badly affect data lines preventing normal program flow, but if you tell me that booting is not possibile without rtc responding I strongly believe in something like corrosion near it.

Tonight I'll check, Stay tuned ;-)

Edit : can someone confirm there's a loop waiting for RTC In the msx2 boot rom ?

By zeilemaker54

Master (213)

zeilemaker54's picture

06-12-2017, 18:05

marcoexo wrote:

Edit : can someone confirm there's a loop waiting for RTC In the msx2 boot rom ?

Yep:
J03DF: CALL C0431 ; initalize RP-501
J03E2: CALL C1CF7 ; select RP-501 bank 2
LD B,3
CALL C1CD2 ; read RP-501 register 3
AND 01H
PUSH AF
CALL C1CDB ; read byte from RP-501 registers 4/5
POP BC
CP 80+1 ; screenwidth >80 ?
JP NC,J03DF ; yep, initialize
OR A ; screenwidth <1 ?
JP Z,J03DF ; yep, initialize
DEC B ; screenmode 1 ?
JR NZ,J0402 ; nope,
CP 32+1 ; screenwidth >32 ?
JP NC,J03DF ; yep, initialize

RP-501 is reinitialized in a infinitive loop when a illegal value is read. This would be the case when no RP-501 is connected to the B4/B5 ports

By marcoexo

Resident (62)

marcoexo's picture

07-12-2017, 01:04

Thank you zeilemaker54 for pointing out the source code... Very good job on that disasm

Here is actual status :
I checked all RTC pins, continuity seem to be ok for all signals and I have no corrosion around it.

Summing up the overall behaviour I see the folllowing :

- RTC Clock is ok
- No CS select is evver risen for the RTC (pin 1), wether at start or at reset /CS is always high
- All RTC address lines are high at startup, they go low after a reset and never go up until restart

I know that CS is the result from a decoded address together (and) with the /IORQ signal, but I have no /IORQ at CPU, it stays high, this is the reason why I think no RTC CS is produced...

M1 shows activity, and also data and address lines do, so I still believe CPU is processing data, must be stuck before accessing RTC ... but where, what are the suspect components that can halt boot phase ?

..still mumbling on the issue...

By marcoexo

Resident (62)

marcoexo's picture

07-12-2017, 02:11

...Update : I tried to diagnose the board using the diagnostic cartridge rom by writing to an eprom the file... Suprisingly a blue screen now appears showing up with an "Error 241"... (I started a new topic on this matter on hardware forum)

By marcoexo

Resident (62)

marcoexo's picture

08-12-2017, 01:10

As pointed out by Pentarou, service repair manual contains Error Codes : http://msx.hansotten.com/uploads/files/vg8235repairmethod02.pdf

Being the NMS8245 technically similar to the VG8235 I tried to check the suspicious components indicated in the manual... Many TTLs are replaced by a unique gatearray in the 8245 but schematic is substantially the same.

Unfortunately I could not find anything else that could lead to the error (incorrect slot selection).
It appears that the problem arises when the program try to write to PPI register A8 . I think that the PPI 8255 is actually encapsulated into the S3527 MSX Controller... so This whole chip might be faulty... :o

By marcoexo

Resident (62)

marcoexo's picture

08-12-2017, 12:36

We have a GO! Smile Happy to say the 241 problem has been solved. It happened to be an S3527 pin (A3 pin) partially desoldered. I carefully resoldered it using a drop of smd soldering liquid.

Now using the diagnostic Eprom I've got a blue screen only, but replacing with original rom.. BUM! It worked.

I don't know why diagnostic rom fails to run, it appears that the ROM enable signal disappears after a while, freezing progrma flow, maybe it's not designed to be run on internal socket instead of specially designed cartridge

By gdx

Paragon (1684)

gdx's picture

08-12-2017, 13:19

Congratulations!
Welds weaken over time. Look closely the board, resolve what seems suspicious and straighten the pins of the cartridge slot connector. Do not mix solder with and without lead.

By Alexey

Paragon (1259)

Alexey's picture

08-12-2017, 16:48

Nice!

By marcoexo

Resident (62)

marcoexo's picture

08-12-2017, 22:08

Alexey, we should take into consideration the diagnostic rom from Philips. I guess it contains most of the functions we are looking for in msx repairing. It would be great if once disassembled could be used as a starting point for a good diagnostic tool to be used by replacing the bios roms....

By Alexey

Paragon (1259)

Alexey's picture

09-12-2017, 10:37

Could someone please send that diagnostics ROM to me? Thanks. If it uses a mapper, I have the cartridge that I could try it on. It's our commercial platform with multi-mapper support.

Page 2/3
1 | | 3
My MSX profile