PCEGT voltage issues on battery

[blazingtaserface]

So a little background, I purchased a PCEGT on eBay that already had a composite video mod. It seemed like a good deal at the time (it wasn't). When I got it, the video was terrible to the extent I considered it unplayable. While I might have considered trying an RGB mod myself, I was too intimidated to try undoing the composite video mod as there seemed to be a lot of strange goings-on on the boards that I didn't understand. So with that said, I spent even more money and shipped the unit off to someone on eBay to have them do the conversion. After about 3 months, they got the unit back to me. Now I've got more problems I can't figure out.

To be fair, I can't say the person who did the RGB mod caused any of these issues. The issues may have been, and probably were, pre-existing. That said, the first problem I noticed was IO errors on my Everdrive when running on battery. There were no issues with a power adapter. Since I have multiple Everdrives and PCE-GT's, I swapped things around, different batteries, different Everdrives, just to rule out those being issues. I also did a diagnostic in the everdrive menu, and on battery, the read I/O would sometimes fail. It showed the 4.5V check point as only like 3.1V when on battery. With the AC adapter, it would get more like 3.8-3.9V. So knowing that I had no desire to ship this unit back to the person who did the mod, I decided to open it up and explore.

At the power switch when off, I measured my Eneloop battery voltage at 8.x. When I'd power it on, the voltage at the battery bank itself would drop to the high 7's, BUT the voltage at the switch would drop to 4.x, and the 5v regulator was only putting out about 3.5-3.6V.

However, on the AC adapter (Triad Magnetics 7.5V unit), when the unit was powered on, I would measure about 6.9-7.0V at the switch. And the 5V regulator was putting out  ~4.5V.

So why such a massive voltage drop at the switch only on battery? At this point, I decided to test the battery bank through the ac adapter port. It worked fine, just as if an ac-adapter was present. No weird voltage drop. So there's no problem with batteries, their ability to provide current, or the unit trying to pull too much current.

I'm really not sure what to make of this. It's got to be some issue with the current path from the battery input on the main board. Some component the current goes through for battery that it doesn't for ac adapter. I don't know. In any case, photos are below. Ignore the 4001 diode in the picture that I bridged near the ac port as I was just temporarily testing something. Also, the RGB mod guy did have two wires running to two points on the brightness control, but I removed them because the brightness control did nothing.

I'm thinking I may just run the battery bank wires directly to the ac-input board with a diode and see how that does. Maybe skip over whatever is the problem. But if anyone has any insight, please let me know.

Also, maybe unrelated, but are all those pin-lifts and bridges around the cpu an issue? Remnants of the old composite mod.

Thanks.

Using Imgur link since files wouldn't attach for size:

https://imgur.com/a/ihRubQr

[Keith Courage]

All the pins lifts on the CPU are fine. Those are just there to enable the old composite video mod. They don't change or effect anything else.

There are some inductors known to go bad for the Battery input but they normally go bad altogether not letting any current through, not limitting it.

The regulator puting out about 3.5-3.6V while on batteries is indeed a little too low. Are your batteries maybe just old? Oh, maybe some of the battery terminals are dirty or need replacement. Try spinning the batteries and see if the voltage goes up or down.

[blazingtaserface]

Thanks for the confirmation on the pin lifts. Would it be a bad idea to just cover the whole chip with liquid electrical tape to protect it, make sure lifted pins don't accidentally break off or somethin?

So a little update from yesterday: To test things, I unplugged the battery connector from the main board, then ran wires from the battery compartment directly to the solder points of the ac-input jack. All problems resolved. So whatever the issue is, it's definitely somewhere between the battery connector port on the main board and the power switch. I went ahead a put a 4001 diode on the positive lead from the battery compartment just to protect the batteries in case an ac adapter is plugged in. For all intents and purposes, everything functions fine for my uses now. But I'm aware I just went around the problem, and I still have no idea what it actually is.

Does a hardware maintenance manual for the TG16/PCEGT exist anywhere?

[Keith Courage]

These are the inductors that sometimes go bad. https://console5.com/store/turbo-express-pc-engine-gt-smd-inductor-coils-l501-l502.html

They are labeleb on the board as L501 and L502. They are underneathe the gray stuff on the board which you can just carefully scrape away. Maybe check those to see the voltages off of each end. I'm just guessing one of them could be the problem since they are known to go bad.

[blazingtaserface]

Thanks much for the info and link. Will check voltages there and see what's what.