Neo Geo MV-1A COLOR RAM BANK 0 ERROR

[DaMerl]

Hi folks,

My MV-1A board gives the following error:

COLOR RAM BANK 0 ERROR

Address: 00400000

Write: 0000

Read: 1000

http://www.imgbox.de/show/up/Merl/P6020518.JPG

The red screen is black now, I soldered the pins of the RAM's again. But the error is still there.

The wiring diagrams for the MV-1F board http://wiki.neogeodev.org/index.php?title=MVS_schematics don't fit, so the question what is to do?

Or is there somebody who can send me the right wiring diagrams?

The traces are looking well, but under the RAM chips, there are vias, so I can't follow the traces exactly.

Hoping for quick responses.

Merl

[channelmaniac]

Easiest thing to do if you do not have the test equipment (scope or logic probe) to check the data lines on the RAMs would be to carefully desolder them, socket them, then swap 'em to see if the error moves. If the error stays then you have a problem with something other than the RAM.

[DaMerl]

Hi,

My logic probe is on the way to me. But is it possible for you to give me instructions to use the logic probe correctly? That would be great!

With greetings from the Baltic Sea Coast!

Merl

[channelmaniac]

Grab the data sheets for the RAMs and check for stuck pins. Check the binary values of the error messages to see what bits you are looking for problems on.

The hard part is that the Color RAM tends to have bits stuck high or low so it's hard to determine just from the values listed. That's why I suggest swapping the chips and see if the problem follows the chip.

[DaMerl]

Ah, okay.

Only for the right understanding:

When I've

WRITE 0000 and READ 1000 than I know that one RAM is odd. This is the first pair of the code: 00 vs. 10

In binary form it means:

00 = 0000 0000
10 = 0001 0000

And from right to left it should be data line 5, which is odd.

Is that right?


Greets!

Merl

[DaMerl]

I ordered two IC's from Hong Kong, because in Germany you don't get the needed one.

The MV-1A board needs LH5168N-80L but I ordered LH5168N-10L. Both chips have the same datasheet: http://www.datasheetcatalog.org/datasheets2/43/431798_2.pdf

So it must be pin 16, which is to be tested, right?

At VCC I should get a HI signal and at GND a LOW signal with the logic probe. But when I tested the chip with a logic probe, the board must be connected at the power supply and switched on. Can you confirm this?

By the way...your Repair-Almanach is awesome. 


Greets!

Merlin

[DaMerl]

Okay,

The Ram's are swaped and th error is still there, but wit other code:

COLOR RAM BANK 0 ERROR

Address: 00400000

Write: AAAA

Read: AAAF

It's the other chip 

AA = 1010 1010
AF = 1010 1111

It's dataline 1 and 3 which should be tested, isn't it?

Furchtbar!

Merl

[channelmaniac]

Sounds like the replacement chip is either bad or you pulled a trace. Check the connections to the chip. Could also be that the -10 (100ns) is too slow to replace an 80ns chip in that function.

RJ

[DaMerl]

Moin!

Yes, the 100ns could be too slow, so I'll replace them.

Because of the numbers under the RAM's...

a colleague from me, recommended these guys here:

http://www.reichelt.de/Drams-Srams/6264-70-M/3//index.html?ACTION=3&GROUPID=2954&ARTICLE=40085&SHOW=1&START=0&OFFSET=16&&SID=10T@MeWn8AAAIAAEzQQbU6e025399112ba792d1710d04d77204a6&LANGUAGE=EN

Don't wonder about the missing pads, the pins are already bridged with Kynar-wire.

Furthermore I have got my logic probe now. Can you explain me how I can test the RAM?

Thank you in advance and best regards!

Merl

[DaMerl]

Hi folks,

Now I’ve replaced the RAM’s, but now I’m getting this message here:

READ 5555    = 0101 0101 0101 0101
WRITE 51CF = 0101 0001 1100 1111

Now I’ve got the following I/O’s to check:

For RAM 7: I/O2
For RAM 8: I/O1, I/O3, I/O4, I/O7

I’ve tested all there with my logic probe. My Hereby the results. The settings are listened in the ( ):

For the HM6264 IC’s (CMOS and “pulse�)


And for the LS273 (TTL and “pulse�)


U9 is connected with RAM8 and U8 is connected with RAM7, the connections are good. VCC has 5.12V at every IC.




But now I’m not able to read the logic probe result correctly. Is this good or not? 
How the error message code is related to the logic probe result? I can’t see it.
At the yellow marked pins in the drawings, there I all three LED’s (green, red and yellow) lightened up. Is this correct? I don’t think so.
And…at the MV-1A board there isn’t a NEO-GO chip, what for a chip is used instead of the NEO-GO?


And the main question: Why in hell is a pentagram there under the socketed BIOS?


Awaiting you soonest reply and best regards!

Merl

[DaMerl]

Hi folks,

I've discovered NeoGeodev Wiki and now I've more questions then before.

My MV-1A board has not NEO-B1, not LSPC2-A2 and not the NEO E0 chips. Instead of this, the graphic chip is a NEO-GRC, which is normally found in CD systems.

http://wiki.neogeodev.org/index.php?title=NEO-GRC

Furthermore I've 5 times the NEO-BUF on the board and the CD unit interface chip NEO-MGA-T. Now I'm very confused.

http://wiki.neogeodev.org/index.php?title=NEO-GRC

http://wiki.neogeodev.org/index.php?title=NEO-MGA


However....I won't put the board into the junk, so maybe somebody can help me??? 


Greets!

Merl