What's new
By:

MV1AX. Z80->68k comm issue (ack)

K

kissedin

Student
Joined
Jan 28, 2023
Messages
40
Reaction score
31
Location
Sweden
Got a MV1AX. Been running diagrom with m1 cart. Can't figure out how to fix this error.

Z80->68k comm issue (ack)

Symptoms:
No sound is heard during sound test in original bios menu. Cartridges boot but samples don't play - only synthesizer music. When playing KOF'99 system randomly play the neo-geo jingle during fights.

Testing so far:
I've been probing z80 pins. Adress lines are pulsing. Data lines are pulsing. Pin 20 (IORQ) is pulsing during startup and during playback from original bios test menu.

Running diagrom w. no cart and probing NEO-D0: Pin 41 SDZ80R is High. Pin 43 SDZ80W is High. Compared with working board and this seems to be an ok behaviour.

Checked for continuity between Z80 and NEO-BUF @ GA9. Checked for continuity between NEO-BUF @ GA9 and HC273 @ U11. Tested ok.

(Here is a good link that describes some of the architecture behind the MV1A: https://wiki.neogeodev.org/index.php?title=MV1A)

Tried shotgun approach and replaced NEO D0 from known working board. Replaced NEO BUF @ GA9 with furrtek repro. Replaced Z80 with new replacement.

Suspicious findings:
There's a HC273A instead of a HC273F, as written on the board, mounted at position U11 and U10.

Any insight regarding the hc273 or other things I can check is greatly appreciated.
 
Last edited:
kissedin said:
Got a MV1AX. Been running diagrom with m1 cart. Can't figure out how to fix this error.

Z80->68k comm issue (ack)

Symptoms:
No sound is heard during sound test in original bios menu. Cartridges boot but samples don't play - only synthesizer music. When playing KOF'99 system randomly play the neo-geo jingle during fights.

Testing so far:
I've been probing z80 pins. Adress lines are pulsing. Data lines are pulsing. Pin 20 (IORQ) is pulsing during startup and during playback from original bios test menu.

Running diagrom w. no cart and probing NEO-D0: Pin 41 SDZ80R is High. Pin 43 SDZ80W is High. Compared with working board and this seems to be an ok behaviour.

Checked for continuity between Z80 and NEO-BUF @ GA9. Checked for continuity between NEO-BUF @ GA9 and HC273 @ U11. Tested ok.

Tried shotgun approach and replaced NEO D0 from known working board. Replaced NEO BUF @ GA9 with furrtek repro. Replaced Z80 with new replacement.

Suspicious findings:
There's a HC273A instead of a HC273F, as written on the board, mounted at position U11. (This chip use the SDZ80W as a clock signal and output data to Z80 through NEO-BUF @ GA9. That is why it has caught my attention.)

Any insight regarding the hc273 or other things I can check is greatly appreciated.
Click to expand...

Had a similar issue but on an MV1FZS slot, which does not have BUFs.

Z80->68k COMM ISSUE (Hello) error was due to a defective 6126 ram. After swapping it sound came back but weak. Completed the repair by swapping the Y3016.

Hope this will help you.
 
Thank you for your advice! My sound ram is a 6116 and it is already socketed and changed (no improvement). The sound I get (synthesizer ) is loud and clear so i think the y3016 is ok. But I will look into it anyway.
 
Last edited:
kissedin said:
Thank you for your advice! My sound ram is a 6116 and it is already socketed and changed (no improvement). The sound I get (synthesizer ) is loud and clear so i think the y3016 is ok. But I will look into it anyway.
Click to expand...

A potential stupid question: dif you try another game besides kof99? Just asking because i had a game with corrupt M1 that was working fine just days before the issue occure.

Besides that i jsut remembered i had a sound issue were NeoGeo jingle would play at charachter selection screen whatever game it is, and almost drove le crazy replacing chips, and it was the NeoC1 in another MV1FZ.
 
Last edited:
Retro Mecano said:
A potential stupid question: dif you try another game besides kof99? Just asking because i had a game with corrupt M1 that was working fine just days before the issue occure.

Besides that i jsut remembered i had a sound issue were NeoGeo jingle would play at charachter selection screen whatever game it is, and almost drove le crazy replacing chips, and it was the NeoC1 in another MV1FZ.
Click to expand...
Lol. Yes I have tried multiple games. In Puzzle bobble synthesizer music sounds good but all samples are corrupted. In KOF99 there is no music or sound at all (except the jingle).

Yes, I noticed your other thread! Most interesting! I'll update this thread when I've got the time to check all good info you got there. MV1AX use chips like Neo buf and hc273 instead of neo c1 but the signals they transfer seem to be the same.
 
So a small update. Been reading the guide at github more thoroughly. In PRINT I get the error message "Z80->68k comm issue (ack)". But the SOUND code is suggesting it's a "68k->Z80 COMM ISSUE (HANDSHAKE)".

(Found some good info in this thread from user ACK: https://www.arcade-projects.com/threads/mv1fz-sound-error-solved-neo-c1.32653/")

The MV1AX doesn't have a C1 chip. Instead it uses NEO-BUF and HC273 for the Z80<->68k communication.

This is how I think it might work. Please take it with a grain (shovels) of salt.

D8-D15 from the 68k is connected to the data inputs (D0-D7) on HC273 @ U10.
HC273 output these input signals to its Q0-Q7-pins.
These Q0-Q7 are connected to B0-B7 on the NEO-BUF @ GA9 (But because of the HC273 the data is only transfered in one direction).
The NEO-BUF is uni directional so B0-B7 is transfered to the A0-A7 on the same chip.
A0-A7 on the NEO-BUF @GA9 is connected to Z80 and its RAM SM1, D0-D7.


HC273 @ U10 is somewhat suspicious...
I'll check its CP pin ("SDW" from PAL-chip SNK1A (pin 22) and NEO-D0 pin 33 ) and its MR pin ("SDZCLR" from NEO-D0). Also might replace it.
 
Last edited:
Great pinning layout.

Do th pinning you traced from 68k and z80 is the same if it was to a NeoC1, instead of BuF+273?

If yes , you are right it might be one of both. Even The BUF seems to be failing regulary.
 
Well ... It seems there's one hc273 "per direction*. One that handles z80->68k. And another one for z80<-68k. They both use NEO-BUF at ga9.

But again. This may be accurate or it may be not. I'm no expert. I just wanna fix my games.
 
ack said:
Its pretty common for NEO-BUF's to fail
Click to expand...
... Well, but like i wrote in my first post this one at GA9 (which seems to handle z80-68k communication )is a brand new repro from furrtek.

But ok. First I'll try to replace u11. If that doesn't work maybe I should try with another repro.
 
kissedin said:
... Well, but like i wrote in my first post this one at GA9 (which seems to handle z80-68k communication )is a brand new repro from furrtek.

But ok. First I'll try to replace u11. If that doesn't work maybe I should try with another repro.
Click to expand...

I am currently fixing an MV1A slot for different problems, but during diagnose i was swapping almost all bufs, and i got 2 time z80-<>68k communication error in Diag, one time removing buf in GA9, and anoyher time in GA8.

When removing the HC273, i got the following error:
Hope this helps also.

20250425_164628.jpg
 
Yes. This is really helpful. Because this indicates that more BUF:s than the one at GA9 is involved.

When doing the pin out I noticed that the A-side of NEO-BUF at GA7, GA8 and GA11 are in parallell on the B-side with the NEO-BUF @GA9.

Here's an example:
68k D0 ->GA9 A0 <->GA9 B0 <-> GA7,8&11 B0

I thought this was just a way to distribute the data signals to other operations around the board. But maybe they're involved in the z80-68k communication. I'll try to change more BUF:s. And maybe go deeper into the NEO-BUF maze.
 
Last edited:
One mor
kissedin said:
Yes. This is really helpful. Because this indicates that more BUF:s than the one at GA9 is involved.

When doing the pin out I noticed that the A-side of NEO-BUF at GA7, GA8 and GA11 are in parallell on the B-side of the NEO-BUF @GA9.

Here's an example:
68k D0 ->GA9 A0 -> GA7,8&11 B0

I thought this was just a way to distribute the data signals to other operations around the board. But maybe they're involved in the z80-68k communication. I'll try to change more BUF:s. And maybe go deeper into the NEO-BUF maze.
Click to expand...

Another thing g to keep in mind, found out that BuF @ GA8 and GA9 have also to controllet input. So if you have trouble in some buttons behaviour, than i am quite sure is one of them.
 
Retro Mecano said:
One mor

Another thing g to keep in mind, found out that BuF @ GA8 and GA9 have also to controllet input. So if you have trouble in some buttons behaviour, than i am quite sure is one of them.
Click to expand...
That is true. I have already changed GA8 because of this. But thing is I had to change it twice because the first replacement (an SNK original) did only hold up for like 5 minutes.

Will use repros this time.
 
Some progress. Changed all stock NEO-BUF to new repros. No changes until NEO-BUF at GA4.

After replacement of this chip I got beep code "001101".

IMG_20250427_094034.jpg

Headline says z80->68k. Copy says 68k ->z80. I guess it is the latter.

I know the clear command is a dedicated pin from NEO-D0 that is connected to pin 1 (MR) on hc273 at U10. Changed this chip. No improvement. I can make this signal go from high to low - and back to high - by pressing A during stock bios sound test...

Next step is to see whether this one is trigged during the actual diagrom test.


Regarding GA4: This chip has its B8-B15 in parallell with GA9's B-side.The A-side of GA4 goes to RAM 8, palette RAM. I can't really figure out why this caused the 68k-z80 handshake to malfunction.)
 
Last edited:
The Z80 has the ability to clear the latched data that is waiting for it from the 68k. The error is indicating the z80 tried to clear the latch, but when it re-read the latch afterwards it wasn't 0x00.
 
You must log in or register to reply here.
Back
Top