Maker Pro
Maker Pro

Ram Error

U

Uriah

Jan 1, 1970
0
This is maybe not a basic question but I guess it is all relative. I
am working on a arcade game board that is operated by a M68000. (Neo
Geo 4 slot). It comes up with a "Work Ram Error" I am pretty sure the
ram is ok. I have compared it many times with a few Huntron style
testers. There are about 6 SMD RAM chips. It shows an address and then
what was written AAAA and what was read AAEA. Can someone provide me
with some tips on how to track down the fault? The 68000 is tied to
three custom 100+ pin smd chips and it is real hard to follow the
traces to be sure because the board is not translucent and many of the
traces go under chips. But what is the process generally used to wire
up RAM and what could cause this read error? Are RAM chips generally
wired directly to the M68000? Or do they have to go through some type
of controller? Can a read error like this be caused by an open or
short in the trace or is this something that is only caused by faulty
logic in the RAM or faulty logic in another chip. It seems there are
no short cuts but there is also no documents on this board. No
schematics, no pinouts on the custom chips and no theory of operation.
I am sure someone understands this more then I do.
Thanks again
Russ
 
L

Lord Garth

Jan 1, 1970
0
Uriah said:
This is maybe not a basic question but I guess it is all relative. I
am working on a arcade game board that is operated by a M68000. (Neo
Geo 4 slot). It comes up with a "Work Ram Error" I am pretty sure the
ram is ok. I have compared it many times with a few Huntron style
testers. There are about 6 SMD RAM chips. It shows an address and then
what was written AAAA and what was read AAEA. Can someone provide me
with some tips on how to track down the fault? The 68000 is tied to
three custom 100+ pin smd chips and it is real hard to follow the
traces to be sure because the board is not translucent and many of the
traces go under chips. But what is the process generally used to wire
up RAM and what could cause this read error? Are RAM chips generally
wired directly to the M68000? Or do they have to go through some type
of controller? Can a read error like this be caused by an open or
short in the trace or is this something that is only caused by faulty
logic in the RAM or faulty logic in another chip. It seems there are
no short cuts but there is also no documents on this board. No
schematics, no pinouts on the custom chips and no theory of operation.
I am sure someone understands this more then I do.
Thanks again
Russ

This could easily be a bus buffer that has gone bad... 1010 1010 written,
1110 0000 read therefore bits 9, 11 and 14 are in error.

Isn't the Huntron tracker a simple short / open / resistive tester that
displays
the I /V curve on your scope?

Too bad you have no docs...if the chips wern't SMDs I'd suggest swapping
their position to see if the error follows the chip. You might need to
figure
a way to use ram from a piggybacked board in order to use know good memory.
Otherwise, see if you can isolate these bits and check for ajacent shorts.

Good luck
 
J

John Fields

Jan 1, 1970
0
This is maybe not a basic question but I guess it is all relative. I
am working on a arcade game board that is operated by a M68000. (Neo
Geo 4 slot). It comes up with a "Work Ram Error" I am pretty sure the
ram is ok. I have compared it many times with a few Huntron style
testers. There are about 6 SMD RAM chips. It shows an address and then
what was written AAAA and what was read AAEA. Can someone provide me
with some tips on how to track down the fault? The 68000 is tied to
three custom 100+ pin smd chips and it is real hard to follow the
traces to be sure because the board is not translucent and many of the
traces go under chips. But what is the process generally used to wire
up RAM and what could cause this read error? Are RAM chips generally
wired directly to the M68000? Or do they have to go through some type
of controller? Can a read error like this be caused by an open or
short in the trace or is this something that is only caused by faulty
logic in the RAM or faulty logic in another chip. It seems there are
no short cuts but there is also no documents on this board. No
schematics, no pinouts on the custom chips and no theory of operation.
I am sure someone understands this more then I do.
 
R

Robin

Jan 1, 1970
0
Uriah said:
This is maybe not a basic question but I guess it is all relative. I
am working on a arcade game board that is operated by a M68000. (Neo
Geo 4 slot). It comes up with a "Work Ram Error" I am pretty sure the
ram is ok. I have compared it many times with a few Huntron style
testers. There are about 6 SMD RAM chips. It shows an address and then
what was written AAAA and what was read AAEA. Can someone provide me
with some tips on how to track down the fault? The 68000 is tied to
three custom 100+ pin smd chips and it is real hard to follow the
traces to be sure because the board is not translucent and many of the
traces go under chips. But what is the process generally used to wire
up RAM and what could cause this read error? Are RAM chips generally
wired directly to the M68000? Or do they have to go through some type
of controller? Can a read error like this be caused by an open or
short in the trace or is this something that is only caused by faulty
logic in the RAM or faulty logic in another chip. It seems there are
no short cuts but there is also no documents on this board. No
schematics, no pinouts on the custom chips and no theory of operation.
I am sure someone understands this more then I do.
Thanks again
Russ

While cycling the test, spray freezer on half the board and then the
other half, gradually localising it the while.

If you are lucky you might be able to repeatedly make the problem worse
or better when only one chip only is frozen, if so change it.

Cheers
Robin
 
Top