Japanese Mega Drive not outputting RGB/component

[Nemoide]

So earlier this year I got a Japanese (model 1) Mega Drive to expand my collecting into the world of Japanese games. It seems to work perfectly with composite video, but when I use my HD Retrovision component cables, it doesn't seem to work. When I hook it up, it sends a nice crisp image for a few seconds and then cuts out to nothing.

The HD Retrovision website mentions that sometimes it doesn't work properly because one of the pins in the AV port isn't sending the right signal and that reflowing the solder should fix it. I got a friend to help with the solder and... it's still not working.

Does anyone have any ideas as to what the problem might be and how it fixable it is? Or does anyone know of a repair service that would be willing to work with a Japanese retro console?
I can live with composite video if I have to but I'd really prefer to use my component cables.

[Ziggy587]

Are the HD Retrovision cables confirmed working? Can you use the RGB output of the Mega Drive in any way for testing?

[pierrot]

Is it an early revision of the Model 1 Mega Drive? I have a pretty early revision (not the earliest, but still non-TMSS), and it has always had an interesting quirk where it locks up after a time with Landstalker, Ragnacenty, and Legend of Heroes 1 and 2. It would not surprise me if early JP Mega Drives have some slightly different specs in places. The video output should be standardized, but it could be doing something to throw off the Retrovision cables.

What pin specifically did you reflow? What kind of solder was used? Specifically, was it lead-free? Is there any audio when the video drops out?

My guess is that there's potentially an issue with the sync signal, 5v line, or a grounding issue. All of these things can be tested, but you won't be able to tell much about the integrity of the sync signal without an oscilloscope. If you have a multimeter, it might be worth checking the voltage on the 5v pin (depends a little on what kind of devices the cable is using, but if it's out by more than +/- 5%, it could be an issue), as well as the resistance from the ground pin to the ground plane (should be very small).

[Nemoide]

I've used the cables with my American model 2 Genesis, so they work. Although thinking about it, even though I've used the adapter for model 1/SMS with my SMS, I wonder if that might be the culprit; I'll test it with a multimeter. There is still audio output when the video cuts out, which makes me think it's MOSTLY working fine.
The unit is currently at a friend's house who has more soldering/electronics skills than I have; they did the work and while I'm not sure what solder they used I trust it was something appropriate.

I'll also try checking the AC adapter; I feel like this is a long shot, but maybe my American model 1 AC adapter will provide better results for some reason.

[pierrot]

There is still audio output when the video cuts out, which makes me think it's MOSTLY working fine.

Yeah, so I would have said that potentially increases the probability that it's a sync issue, but it occurred to me that some of the component signals are going to be superpositions of the RGB signals, and the circuitry might actually lose video output with a problem on one of the RGB signals, if it's a mixed signal process. It's probably not going to be an issue with 5v or ground, but you could test it anyway, just to be sure. The RGB and sync outputs would really require an oscilloscope to verify any sort of serious signal distortion, or external coupling from other noisy sources.

The unit is currently at a friend's house who has more soldering/electronics skills than I have; they did the work and while I'm not sure what solder they used I trust it was something appropriate.

Yeah, the only potential issue here is that mixing alloys can potentially alter the heat profile, and might reduce the integrity of any rework. It's probably not a concern in general, but lead solder mixed with unleaded solder would be a pretty dramatic mixture, and I know some people who just refuse to use lead solder.

I'll also try checking the AC adapter; I feel like this is a long shot, but maybe my American model 1 AC adapter will provide better results for some reason.

US power supply would technically be better, but Japanese power is nominally ~17% out of US specifications, and most of the critical components in the Mega Drive are given a regulated voltage, internally. That said, the unregulated DC off the power supply is used for the rails on a lot of the analog audio and video circuitry. The only real concern is that the rails might be reduced enough by the difference in transformer winding ratio that the signals would start to clip, but I doubt it because those signals probably aren't greater than 3.25 V in peak magnitude. You may as well use the US adapter, anyway, though. There's at least a possibility that it would solve the issue.

[Nemoide]

EPILOGUE:
I was mistaken earlier when I said the audio was fine (I didn't actually test it before posting); the audio goes out with the video. Some games work for a couple seconds, others won't go past the TMSS screen.
I went on ebay to get another Mega Drive... and it has a slightly different RGB/component issue in which some games work fine but certain parts will cause the signal to drop out (ie during the "Sega" logo screen or during the title screen but not during the game itself). And other games won't work at all with the component cables.

But even though this was a longshot, I decided to learn how to up my soldering skills, got a capacitor replacement kit from Console5 and recapped that original system! Since I had a backup MD that worked better, I felt like if I destroyed the original it would be at least a good learning experience.
But at least I feel like I've improved my soldering ability and prolonged the life of the system; I'm glad I didn't destroy it. Plus I have a good SD CRT now so playing with composite really isn't so bad.
(...but it still bugs me a little. I *may* attempt the triple-bypass mod at some point since I have another MD for composite-video.)

[pierrot]

Congrats on your first recap, Nemoide! If you don't happen to have professional tools, that's quite a feat, at almost any skill level. I sometimes have had to add batches of little 0402 surface mount caps to boards by hand, for RF filtering on noisy traces, but even though it's a bunch of through hole parts, full board recaps on old consoles is still a really daunting thought for me.

It's too bad it sounds like it didn't solve the issue. Board recaps aren't necessarily the panacea they're sometimes made out to be. The sound dropping out is interesting to me. It seems like that would point to a power issue. I suppose there could still just be something wrong with the RGB lines. That technically might not be ruled out. I'm not sure what kind of TV you're using, but it might drop everything if it loses the video signal. If you're checking out console5 now, you might want to try locating the pins on the processor that go to the DIN port on the back, and just check continuity (/resistance) between them with an ohmmeter.

[Nemoide]

I'm not sure where that is on the site and if I found something that wasn't working, I'm not sure I could figure out how to fix it. I can follow directions but I'm not sure I really understand the workings of the console! Although I know HD Retrovision has test software... if I someday get a Mega Everdrive, I might try that.

But I feel like my *actual* best option is to let that sleeping dog lie and just say the MD only gets played on the CRT!

[pierrot]

Oh, yeah. Sorry. I guess I take reading a schematic for granted sometimes. They take the RGB signals off of the composite encoder to the connector, anyway, and I’d have to look at a data sheet to see if it’s a simple passthrough or not. There ‘s a couple components in between the processor and the encoder as well, which would get in the way of a continuity check.

If you wanted, and had a picture of you board handy, I could probably give you an inspection diagram for where to meter. Something about the implementations of those encoders causes jailbars even on RGB output, over time. It’s conceivable that it might create other integrity issues too.

If you just don’t want to think about it anymore, though, I get that.