marcao_cfh
Student
mv1fz video ram error
- Thread starter marcao_cfh
- Start date
heh, I was after images of the screen when its showing the error message not the ram chips. The wrong tiles being displayed can provide info about what is wrong, but I can't make out what tiles they in your original screen shot of the error.
marcao_cfh
Student
Ah my bad, I thought you wanted a photo of the board. Anyways, I went ahead and resoldered both 32K VRAMs and now image is back to normal!
But replacing the 2K VRAM didn't solved the issue. I'm getting:
Video RAM Error
Address 00008000
Write 0000
Read 0002
And using a diagnostics bios:
VAM Address (A0-A7)
Address: 008000
Actual: 0202
Expected: 0000
And I tested the RAM using my T48 and it says the RAM is good.
Any clues on what to check now?
So the address lines are shared between the 2x 2k ram chips and come from the LSPC chip. The error is indicating you have an issue with one of those address lines (likely A1 or A2). Its effecting both chips so you probably have a bridge to another pin/trace or a cut trace.
xodaraP
Legendary
VRAM error at $8000 boundary is a fast VRAM error, not the standard VRAM
Usually Sony chips, 6264 slim type DIP ICs even on the 1FZ
Unless you’re replacing with the exact brand/speed, replace both
The tile error you’re getting is probably related to the RAM you removed/replaced. Check your soldering
Usually Sony chips, 6264 slim type DIP ICs even on the 1FZ
Unless you’re replacing with the exact brand/speed, replace both
The tile error you’re getting is probably related to the RAM you removed/replaced. Check your soldering
marcao_cfh
Student
I checked the traces with a multimeter, and indeed the A1 line doesn't have continuity to the LSPC2 IC. That was amazing! I'll solder a jumper wire tomorrow and hope this finally brings this board back to life.
I replaced it with one of the exact brand/speed. It's a Sony CXK5814P-35L. The tile error was solved by resoldering the other RAMs around the one I replaced, it's likely I had a cold solder somewhere. Now for the new error, I checked and A1 trace have a cut somewhere, so I'll solder a jumper wire tomorrow.
arcadecollector
Enthusiast
had the same "stuck at 1" at VRAM read error on a MV-1C, changed RAM5 chip with a one from a PGM mainboard 62256 (SOIC2
type and worked again!
type and worked again!marcao_cfh
Student
Ok, so good news bad news.
Good news is soldering a jumper wire fixing the A1 trace solved the issue and now I have image on the small screen I use for testing purposes (it's a car rear camera screen).
My setup for testing purposes is mvs board connected to a openmvs board kit to handle all signals, and since this small screen expects a composite signal and not a RGBS signal, then I take the R, G, B and S lines and combine them into a single line using a 70r resistor per line. This was a suggestion somebody gave me some time ago, and gives me a black and white signal that's compatible with composite and good enough just to check there's image. So far, all good when testing this board.
Next step is to pick everything, remove the resistors and connect the mvs board and openmvs to a homemade gbs-c so I can test sound and everything else. Now things got weird. My gbs-c doesn't recognize the signal from this board. I tested the same gbs-c, the same scart cable and the same openmvs board kit but using a couple different mvs boards and they all worked ok but for some reason this one doesn't, so the issue is something related to this board. I also tested like 3 different openmvs boards I have here, and no video. I did got the gbs-c to recognize sound from this board once, but it didn't recognized video a single time.
Any clues on what to check now?
Good news is soldering a jumper wire fixing the A1 trace solved the issue and now I have image on the small screen I use for testing purposes (it's a car rear camera screen).
My setup for testing purposes is mvs board connected to a openmvs board kit to handle all signals, and since this small screen expects a composite signal and not a RGBS signal, then I take the R, G, B and S lines and combine them into a single line using a 70r resistor per line. This was a suggestion somebody gave me some time ago, and gives me a black and white signal that's compatible with composite and good enough just to check there's image. So far, all good when testing this board.
Next step is to pick everything, remove the resistors and connect the mvs board and openmvs to a homemade gbs-c so I can test sound and everything else. Now things got weird. My gbs-c doesn't recognize the signal from this board. I tested the same gbs-c, the same scart cable and the same openmvs board kit but using a couple different mvs boards and they all worked ok but for some reason this one doesn't, so the issue is something related to this board. I also tested like 3 different openmvs boards I have here, and no video. I did got the gbs-c to recognize sound from this board once, but it didn't recognized video a single time.
Any clues on what to check now?
marcao_cfh
Student
I've had MVS boards that don't get any picture on the Framemeister, due to the sync signal. It could just be an incompatibility with the GBS-C, and you'd need to use something else.
Kujako
Champion
On the GBS-C, depending on the firmware its running I've had very mixed results. The one I built in the before times had to run an old firmware to work with a Neo Geo. Seems there's a bug in the code that keeps getting fixed and coming back. Most of the newer prebuilt ones seem OK (not great) but it's possible that the unit in question isn't compatible.