Sega Megadrive (French RGB PAL) black screen [SOLVED/not SHORTS & CAPS but 1145p Chip]

[Retro Mecano]

Ok this one is getting me nuts.

After a long diagnostic and cleaning, i suspected that the black screen issue on my megadrive model 1 might be the CXA1145P video chip (under the heat sink).

I went ahead and removed it, put on a chip socket , and replaced it. Still a black screen.

By coincidence i hit it with the multimiter and it came of the socket slightly, and miracle, video came back briefly. I thought it was a back solder or contact, and after checking and wiggling, discovered that the video will only function if i keep the half right of the chip put of socket. Yes you hear it right. See photo below. I tried another chip, exactly same thing.

Other odd thing is that if i seat it completely, its a black screen, but if i put my finger in one of the rgb data pins, image is nack with waving color of that pin.

Now wtf?

 

[Fluffy]

I'd try to find which pin exactly needs to be disconnected for it to work, then check for short-circuits on that pin.

 

[Retro Mecano]

I'd try to find which pin exactly needs to be disconnected for it to work, then check for short-circuits on that pin.

Exactly.I was having the same thinking.
Looking at schematics, i found out that the culprit was Pin12 (VCC1) that has something to do with audio. Now it's a struggle to go back and find out why it is behaving like that..

 

[Nic034]

Hello

Genesis/megadrive is notorious known to have bad cxa1145p or video issue with black screen/no video output sometimes. Just perform a an RGB bypass and issue will will be solved.

I’ve had this issue several times in the past, but ne suspect bad chip or something merde and it was fixed with a small, well-placed bypass.

If you find the root cause I will be interested in that for sure

Bye

 

[Retro Mecano]

Hello

Genesis/megadrive is notorious known to have bad cxa1145p or video issue with black screen/no video output sometimes. Just perform a an RGB bypass and issue will will be solved.

I’ve had this issue several times in the past, but ne suspect bad chip or something merde and it was fixed with a small, well-placed bypass.

If you find the root cause I will be interested in that for sure

Bye

I could do that also, but my purpose is purely technical, just to underwtand what would be causing it.

I will keep you suggestion as last resort if anything else would fail.

 

[Fluffy]

Exactly.I was having the same thinking.
Looking at schematics, i found out that the culprit was Pin12 (VCC1) that has something to do with audio.

VCC1 should be positive voltage. (I think what usually happens if VCC is disconnected is that power goes from the pins through the ESD protection diodes, which results in glitchy operation.)

 

[Nic034]

Yes of course, interesting

The datasheet indicates that VCC on pin 12 is connected to the sync via capacitors, and the PAL implementation diagram is different from the recommended test circuit in terms of components.

Are the capacitors on the 5V line going to pin 12 okay? did the amp on this line ok ?

Indeed, VCC on pin 12 is dedicated to powering the buffered audio section and everything that is not related to RGB. Possible shortcut on the audio side ?

I’ll take a look on my end, starting from the audio input section up to the audio amp/DAC, there’s probably an issue somewhere along that path.

In any case, you need to isolate the fault, does it improve if you momentary « isolate » audio input to the encoder?

Note : power consumption seems to be at 20ma average referring to datasheet.

 

[Retro Mecano]

Yes of course, interesting

The datasheet indicates that VCC on pin 12 is connected to the sync via capacitors, and the PAL implementation diagram is different from the recommended test circuit in terms of components.

Are the capacitors on the 5V line going to pin 12 okay? did the amp on this line ok ?

Indeed, VCC on pin 12 is dedicated to powering the buffered audio section and everything that is not related to RGB. Possible shortcut on the audio side ?

I’ll take a look on my end, starting from the audio input section up to the audio amp/DAC, there’s probably an issue somewhere along that path.

In any case, you need to isolate the fault, does it improve if you momentary « isolate » audio input to the encoder?

Note : power consumption seems to be at 20ma average referring to datasheet.

Well that was a looong shot on desoldering almost the whole audio area.. even put the ICs on sockets to swap.

Finally, i found two caps that were way out of range (c24 and c1). Swapping them fixed the issue. I am not sure how this has things to do with audio shorting. Or may be putting ICs on sockets removed some short... i donnow to be honest and i don't want to find out. it already took almost a week.

Thanks guys, and specially @Nic034 for the PAL schematics.

 

[Nic034]

You’re welcome man ! Don’t hesitate ton contact me directly on X.
Thanks for providing feedback on this ! Very interesting results.

 

[Retro Mecano]

You’re welcome man ! Don’t hesitate ton contact me directly on X.
Thanks for providing feedback on this ! Very interesting results.

Now thinking back, i think i know what the issue was. Starting the repair, i fixed what i thought was a cut trace near the video connector.

After that i started swapping the video chip, and encountred that pin short. I found out that original chip was bad, but fixing the trace caused the issue: it was factory made in every f*** french RGB megadrive. I just opened 2 other and they had EXACTLY the same cut. So i removed the patch in the middle of repair when i searched for the short, and forgot about it... later this week, video came back with audio (thinking it was capacitors...)

Checked another PAL G megadrive with RF only, the trace was normal straight to connector.

It seems they did it in purpose to isolate the VCC audio from sync signal.

Now it was officialy a bad 1145P chip

What are the odds to find such engineering marvels...

 

[Nic034]

Bad 1145p