Difference between revisions of "PCB Repair Logs Dragonball Z"
m |
m |
||
Line 20: | Line 20: | ||
<tr class=""> | <tr class=""> | ||
<th scope="row" style="text-align:left; white-space: nowrap;">Pin Out</th> | <th scope="row" style="text-align:left; white-space: nowrap;">Pin Out</th> | ||
<td class="" style=""> | <td class="" style="">[http://wiki.aussiearcade.com.au/index.php/File:Jamma_pinout.png Jamma Pinout]</td> | ||
</tr> | </tr> | ||
</table></p> | </table></p> |
Latest revision as of 11:12, 5 February 2013
Dragonball Z
Manufacturer | Banpresto |
---|---|
Year | 1993 |
PCB Image | Reserved |
Pin Out | Jamma Pinout |
Repairer: cal2
Forum Thread: Dragonball Z PCB Repair
Symptoms
Repeating graphics on first scroll plane:
Repair
Based on the mame driver, I found that the graphics for this plane come from 222a09.25l (K053536 PSAC-2 #2).
The board passes the PSAC2 rom test, so I assumed the mask rom is good. Note that the pinout is compatible with M27C160.
I reproduced the problem on mame by changing the way the pixels bits are fetched from this eprom.
Normaly it should be:
0 1 2 3 4 5 6 7 8 9 a b c d e f
But is displays like this:
0 1 2 3 0 1 2 3 8 9 a b 8 9 a b
In binary this means:
3210 3210 0000 0000 0001 0001 0010 0010 0011 0011 0100 0000 0101 0001 0110 0010 0111 0011 1000 1000 1001 1001 1010 1010 1011 1011 1100 1000 1101 1001 1110 1010 1111 1011
So it means bit 2 is likely stuck to 1 or 0.
Since the tiles are 16x16 in 4bpp, this means b0 isn't used. So, if b1 = A0 it means b2 = A1 on the mask rom.
Looking at A1 using a probe, i found it was bad compared to the other signals. It should be a squared signal of 1/2 frequency of A0. Instead, it is pulsing very quickly.
So A1 is bad. A1 is connected to O3 of 27F374 @ 24M.
I changed the IC to see but it didn't make changes.
After taking time tracing the board, I found that a pin of the custom chip 053936 @ J16 was desoldered and made contact to an other pin.
Fixing this pin, also fixed the graphic problem