Raph_friend
Grand Master
The client alleged that the PCB had problems with the graphics and no sound.
I tested the board and it just show me static garbage screen, for this only reason I continued the repair, as it means that something still works.
I separated A and B boards and noticed that the cases had a liquid on it, just like when B board has a leaking battery. I did check it and there was no leakage. Tested on another A board and everything was fine, excluding problems on B board. I removed the battery just in case.
Going for the A board restoration, its case was also dirty with liquid, looking like leaked lithium, however, was alot, it even looked like water. I unscrewed the case, I came across this:
Holy mother of Jesus! I ALWAYS see PCBs that come with batteries installed in place of the SUPERCAP, but I didn't imagine they would reach this point. Until then, I was used to see something like this:
Okay, let's see what can be done.
I removed the initial layer of dirty and dried the liquid that was spread over the board and installed the supercap, which should have been done since the previous "maintenance".
When I see garbage graphics problems on A boards, I always replace the LS157 below the CUSTOM DL-0921 one by one (7U, 7V and 7W) starting at 7W. On the first, game has already started to present graphics but no colors, I changed the other two and the color palettes came back, but still with graphics problems.
Inspected the PCB again, and there was some strange traces near the CN4 socket and one going towards the 3W LS157 was damaged. When restoring, a few more data lines were appeared on the screen.
I left it testing for a few minutes to see if something new was happening. The board started to rebooting.
After another inspection, I noticed some broken tantalum capacitors and the large capacitor in the 12V line of the sound circuit was a little “fat”, I decided to replace these.
The reboot problem was solved, so I back for the rest of the graphics problems.
On the solder side, a broken trace was found near a capacitor's pin solder point, and, to prevent a short, I decided to leave a jumper simple jumper (I could leave it straight, but prefer to keep the original trace form).
Testing a game (Marvel Super Heroes) it show perfect graphics, now, time for the sound circuit, which was totally non-functional.
I admit that A board was still unknown to me. A few years ago I studied it a lot, but without a schematic and block diagram, that we have today, it was really hard to learn. In fact, I thank to Loïc * WydD * Petit, for his wonderful service, which can be seen at https://gitlab.com/loic.petit/cps2-reverse.
First, volume and test switches, it didn't work. I opened everything, cleaned it, and closed them again, perfect.
The continuity of the switch signal was reaching the kick socket, but not the volume control IC.
There's something in common on these PCB that are not so well taken care of, which is this specific damage at the edges of the board and it brings bad traces. After repair, the volume, at least, increased and decreased, but still without sound.
After analysis, I decided to replace the ICs related to the reception and direction of the sound signal to QSOUND. Starting with 2P LS374, which, in resume, is a buffer for the high / low impedance signal coming from board B. The sound came back, but all strange, a little metallic, so I replaced the neighbor 2R LS374 and the sound normalized, but after a few seconds, it stop and repeat the last note played. Looking the documents again, I replaced the 5S LS245, since the function of this CI (in its project) is to manage data communication at a logical level, it was probably responsible for the QSOUND signals reaching the end correctly, and, for my surprise, it worked. The sound was perfect.
I tested another game, ZERO 3, also perfect.
I left testing a few more hours and playing at the same time, but the sound died on me.
We will investigate again.
I was sure it was not QSOUND, as ZERO 3 does not start INTRO if it is not communicating well, and neither does JUKEBOX, and both were perfect.
I think that this board was lay aside for years and suffered a lot of degradation, even more with the giant battery leaking, maybe the most sensitive ICs were not withstanding the current flowing, and also the sound impedance, so I started for the LS374 next to the last replaced LS245. With the 4S LS374 it didn't work, but when replacing the 3S LS374, the sound came back more beautiful than ever.
Until then, the sound was from JAMMA, in mono, I went to test the stereo, nothing.
By the diagrams, I saw that the sound signal leaves the QSOUND and passes through the 6P F04, goes to another logic IC, until it reaches the 10J TDA 1543, which receives the signal and directs the R and L signals to the LM833, until it reaches the 10E TC9176P, which will send the signal to the amplifier and then to the outputs. I checked the continuity, it was ok.
Then I changed the line ICs one by one, when I arrived at the TDA1543, the stereo sound was magnificent!
It was already a day in testing and the card was working correctly, so I decided to switch to another game, this time X-men vs Street Fighter, and, sadly, again graphic problems, and this time very strange. As you can see in the video, after inserting credits, the game normalized (???)
I did not want to replace the DL-0921, as I was “almost sure” that it was not him, since I did some tests with instrumentation on all pins and did not present wrong comparative values, so it could only be some CI still defective or bad traces.
I went back to studying the documentation. There are 5 pins responsible for the management of BUS address, my intuition said that that was it.
In a continuity test between these pins with signals that come out from DL-0921 and go to CN1, I identified that the signal from pin 81 was not reaching the socket pin. Following the trace, I found the culprit.
Repair done.
Game working like new \ o /
In addition to the help of the documentation already mentioned, I also used a semi-decapped board for continuity comparison.
It was a few days of work on this board, hours of study, but a lifelong learning experience. I am very satisfied, since until then, I had never had many results on A boards with these kind of problems (mainly sound), simply abandoned the service or if it was mine, sent it to the scrap box.
Sorry for the long text, but I hope it can help and encourage the community to not let their PCBs die.
Any problems, doubts, suggestions or criticisms, get in touch.
I tested the board and it just show me static garbage screen, for this only reason I continued the repair, as it means that something still works.
I separated A and B boards and noticed that the cases had a liquid on it, just like when B board has a leaking battery. I did check it and there was no leakage. Tested on another A board and everything was fine, excluding problems on B board. I removed the battery just in case.
Going for the A board restoration, its case was also dirty with liquid, looking like leaked lithium, however, was alot, it even looked like water. I unscrewed the case, I came across this:
Holy mother of Jesus! I ALWAYS see PCBs that come with batteries installed in place of the SUPERCAP, but I didn't imagine they would reach this point. Until then, I was used to see something like this:
Okay, let's see what can be done.
I removed the initial layer of dirty and dried the liquid that was spread over the board and installed the supercap, which should have been done since the previous "maintenance".
When I see garbage graphics problems on A boards, I always replace the LS157 below the CUSTOM DL-0921 one by one (7U, 7V and 7W) starting at 7W. On the first, game has already started to present graphics but no colors, I changed the other two and the color palettes came back, but still with graphics problems.
Inspected the PCB again, and there was some strange traces near the CN4 socket and one going towards the 3W LS157 was damaged. When restoring, a few more data lines were appeared on the screen.
I left it testing for a few minutes to see if something new was happening. The board started to rebooting.
After another inspection, I noticed some broken tantalum capacitors and the large capacitor in the 12V line of the sound circuit was a little “fat”, I decided to replace these.
The reboot problem was solved, so I back for the rest of the graphics problems.
On the solder side, a broken trace was found near a capacitor's pin solder point, and, to prevent a short, I decided to leave a jumper simple jumper (I could leave it straight, but prefer to keep the original trace form).
Testing a game (Marvel Super Heroes) it show perfect graphics, now, time for the sound circuit, which was totally non-functional.
I admit that A board was still unknown to me. A few years ago I studied it a lot, but without a schematic and block diagram, that we have today, it was really hard to learn. In fact, I thank to Loïc * WydD * Petit, for his wonderful service, which can be seen at https://gitlab.com/loic.petit/cps2-reverse.
First, volume and test switches, it didn't work. I opened everything, cleaned it, and closed them again, perfect.
The continuity of the switch signal was reaching the kick socket, but not the volume control IC.
There's something in common on these PCB that are not so well taken care of, which is this specific damage at the edges of the board and it brings bad traces. After repair, the volume, at least, increased and decreased, but still without sound.
After analysis, I decided to replace the ICs related to the reception and direction of the sound signal to QSOUND. Starting with 2P LS374, which, in resume, is a buffer for the high / low impedance signal coming from board B. The sound came back, but all strange, a little metallic, so I replaced the neighbor 2R LS374 and the sound normalized, but after a few seconds, it stop and repeat the last note played. Looking the documents again, I replaced the 5S LS245, since the function of this CI (in its project) is to manage data communication at a logical level, it was probably responsible for the QSOUND signals reaching the end correctly, and, for my surprise, it worked. The sound was perfect.
I tested another game, ZERO 3, also perfect.
I left testing a few more hours and playing at the same time, but the sound died on me.
We will investigate again.
I was sure it was not QSOUND, as ZERO 3 does not start INTRO if it is not communicating well, and neither does JUKEBOX, and both were perfect.
I think that this board was lay aside for years and suffered a lot of degradation, even more with the giant battery leaking, maybe the most sensitive ICs were not withstanding the current flowing, and also the sound impedance, so I started for the LS374 next to the last replaced LS245. With the 4S LS374 it didn't work, but when replacing the 3S LS374, the sound came back more beautiful than ever.
Until then, the sound was from JAMMA, in mono, I went to test the stereo, nothing.
By the diagrams, I saw that the sound signal leaves the QSOUND and passes through the 6P F04, goes to another logic IC, until it reaches the 10J TDA 1543, which receives the signal and directs the R and L signals to the LM833, until it reaches the 10E TC9176P, which will send the signal to the amplifier and then to the outputs. I checked the continuity, it was ok.
Then I changed the line ICs one by one, when I arrived at the TDA1543, the stereo sound was magnificent!
It was already a day in testing and the card was working correctly, so I decided to switch to another game, this time X-men vs Street Fighter, and, sadly, again graphic problems, and this time very strange. As you can see in the video, after inserting credits, the game normalized (???)
I did not want to replace the DL-0921, as I was “almost sure” that it was not him, since I did some tests with instrumentation on all pins and did not present wrong comparative values, so it could only be some CI still defective or bad traces.
I went back to studying the documentation. There are 5 pins responsible for the management of BUS address, my intuition said that that was it.
In a continuity test between these pins with signals that come out from DL-0921 and go to CN1, I identified that the signal from pin 81 was not reaching the socket pin. Following the trace, I found the culprit.
Repair done.
Game working like new \ o /
In addition to the help of the documentation already mentioned, I also used a semi-decapped board for continuity comparison.
It was a few days of work on this board, hours of study, but a lifelong learning experience. I am very satisfied, since until then, I had never had many results on A boards with these kind of problems (mainly sound), simply abandoned the service or if it was mine, sent it to the scrap box.
Sorry for the long text, but I hope it can help and encourage the community to not let their PCBs die.
Any problems, doubts, suggestions or criticisms, get in touch.