First what I tried to do. I desoldered the SKA2195D and swapped in a CXA2075M with the appropriate 3.32Kohm 1% resistor. The Genny 2 worked fine but the composite output was still crappy on my LCD TV, anyway I was after SVideo so I couldn't care less although I expected some improvement. Then I noticed that after the swap the audio after say 2 minutes of playing fades to zero, the voltage regulator is very hot and then I can no longer power the console. If I let it cool down (say for an hous) I can repeat the process anew, play 2 minutes then the audio volume goes to zero and then can no longer power the console on (reset works). I took out the CXA2075 and run with no video encoder whatsoever, same results, so it can;t be the video encoder chip as there was none at all at this point. So I decided to perform a complete recap (using the proper capacitor kit at console5.com) but that didn't do much, same issue, so I also replaced the voltage regulator (7805 again bought at console5.com) and again nothing, the same problem reappears. Anyone know how to diagnose/fix the issue? I repeat every single electrolytic capacitor has been replaced and a new 7805 has been put in. Also even without any video encoder (no chip at all) it still runs for 2 minutes fine [I can hear the music] and then the audio volume fades to zero at which point if I shut it down it won't power up for some time. The voltage regulator sink is also very hot at this point so I normally don't wait to see what happen because I am afraid that I'd fry the regulator or the parts around it. Help please. BTW: I did build the simple circuit to get SVideo out from the CXA2075 (jut 2 75ohm resistora inseries with 2 220ohm electrolytic capacitors) nd the graphic does look sharp/crips on my LCD TV, reason that I swapped out the SKA2195D in the first place as it doesn't support any SVideo output. So now I know it is a worthy mod. Also I already bought a replacement console, another Genny 2 with a VA3 motherboard to which I did the same process, swap out the SKA2195D and put in a CXA2075M (I had to buy 5 from Ebay anyway so I do have a few to play with). It works perfectly on this second console so the point for me is to try to "fix" the other one just because it is fun and because being a 1.8 I can attempt the CCAM to try to get the best audio froma a genny.
Couldn't that be worked around with a voltage regulator capable of handling a higher current output as well as a superior power brick? Or would the PCB be unable to handle the additional draw?
Yes, but that costs money and hence why SEGA didn't want to follow that route. The CXA2075 eats so much power because it's RGB buffer can do resolutions up to 1024x768... O_O
That would explain why the other MD2 that I modded is heating up a little. I am planning to buy a second 7805 with heatsink and connect it in parallel so they share the load, and see what I can do on the AC/DC transformer side. But back to my MD2 VA1.8, it overheats even without CXA2075. No Video Encoder at all. I took it out thinking it was the culprit but it didn't help. So I put it back in to complete the SVideo mod and see the results. Thanks to that I decided to replace the SKA2195D on other MD2 (the VA3 that nows gets a little warmer). Anyway from the data sheet the CXA2075M uses max 780mW, which is the same max as the CXA1645M, and Sega did use that on the MD2 VA4, I guess they fixed the voltage regulation section. I do also have an MD1 VA3 that I plan to mod by installing a parallel CXA2075M (don't want to take out the 1145P) and those boards came with 2 7805 already so hopefully the extra needed juice should be available without much issues (will see to that). Back to my MD2 VA1.8 that overheats and cuts the audio after 2 minutes, any suggestions on where I should look for problems? I am afraid one of the SMD component (either R or C) has gone bad or a trace has deteriorated, or worse a splash of soldering material wedged itself somehwere that I can't see and is "shorting". Any past experiences on where to look at? I only have available a multimeter (mainly useful for continuity testing) and of course a soldering iron, no oscilloscope or logical analyzer or even a signal generator (to try to inject a sinus pulse and see where it gets absorbed). Thanks for any help that can be provided.
I know this a late reply but I can't think of anything to cause this problem. If the problem still exists then do an eagle eye scan of the board for anything strange. Perhaps a line got crossed somewhere. There are places where the + and - lines run right next to each other. Also do what you said and check for bridged stuff. If all that checks out, then there is a faulty chip, resistor, or SMD device in there somewhere.
