Hi. I have a Neo Geo AES3-4 that I am trying to fix.
I was told the fault was "only works 1 in 100 times" before it came to me. It's a 5V model, no step-down regulator. I powered it up using a bench power supply, and observed around 700mA draw. Purple screen with graphical corruption.
I measured 5V on the jack side of the T1 common mode choke, but only 3.4V on the switch side. I removed and cleaned the switch, and tried bypassing it, no change. I removed all three transistors that are part of the over-voltage protection circuit, they all tested fine and bypassing them didn't have any effect. I also tried injecting 5V after the under-voltage protection, and it still drew around 650mA and dropped to 3.4V at the 68000 CPU.
I noticed some caps looked like they had leaked, or it might have just been flux and dirty, but I removed them anyway. I'll test them at some point, but removing all the electrolytic ones them didn't fix the low voltage issue. I left the ceramics. I couldn't find anything getting really hot either.
I measured around 400Ω between the 5V and GND pins of the 68k, which seems fine. I probed around, the right side of the board is a little higher by about 10Ω, but that's the only variation I could find. I'm wondering if the multimeter doesn't use a high enough voltage to trigger the issue, i.e. it's some kind of PN junction failure that needs ~3.4V to start conducting.
At this point the only thing I can think of is to start removing chips to try to find which one has the issue, but that's not a lot of fun even with a decent desoldering gun.
Does anyone have any ideas for other things to try? There aren't many components left that can't be ruled out, beyond the chips.
I was told the fault was "only works 1 in 100 times" before it came to me. It's a 5V model, no step-down regulator. I powered it up using a bench power supply, and observed around 700mA draw. Purple screen with graphical corruption.
I measured 5V on the jack side of the T1 common mode choke, but only 3.4V on the switch side. I removed and cleaned the switch, and tried bypassing it, no change. I removed all three transistors that are part of the over-voltage protection circuit, they all tested fine and bypassing them didn't have any effect. I also tried injecting 5V after the under-voltage protection, and it still drew around 650mA and dropped to 3.4V at the 68000 CPU.
I noticed some caps looked like they had leaked, or it might have just been flux and dirty, but I removed them anyway. I'll test them at some point, but removing all the electrolytic ones them didn't fix the low voltage issue. I left the ceramics. I couldn't find anything getting really hot either.
I measured around 400Ω between the 5V and GND pins of the 68k, which seems fine. I probed around, the right side of the board is a little higher by about 10Ω, but that's the only variation I could find. I'm wondering if the multimeter doesn't use a high enough voltage to trigger the issue, i.e. it's some kind of PN junction failure that needs ~3.4V to start conducting.
At this point the only thing I can think of is to start removing chips to try to find which one has the issue, but that's not a lot of fun even with a decent desoldering gun.
Does anyone have any ideas for other things to try? There aren't many components left that can't be ruled out, beyond the chips.