At this point, all the surviving hardware modules from the 90's and early 2000's could use complete rebuilds with full electrical component, physical switch, and interface port replacements. From those recordings, it definitely sounds like you've got some old, glitchy hardware component that needs to be examined.

well, i have now tested it with a generic 9V switched adapter. playing some more complex MIDIs, right after powering on, it played fine for a couple minutes, then severe bouts of clipping/dropouts would occur, after which it would play on. this at least rules out the power adapter as a cause. also tried the headphone jack, with same results. it can suddenly fall into this"clipping" state for a couple of seconds and then play normally again, so the intermittent nature of this is really weird to me.

this unit also has been repaired previously, as the rear output jacks had come off during shipping. as can be seen in the picture, they are soldered to L19/L20 inductors - there's no SC-55ST service manual around that i'm aware of, but in the SC-55mkii manual it also mentions the outputs being connected to inductors, so this should be correct. there's also a via connected there, which i'm not 100% sure about, but it does seem to be in that same path anyway. ground is connected elsewhere on the back of the board, by the way.

i'll try to recap this thing sometime as a last resort, though i'm not very hopeful that will fix the issue.

i've redone the soldering on the aforementioned repair and since have not been able to reproduce this fault. now, i don't even understand what the issue was in the first place, as the only thing i did differently is to try and not solder into the via, which shouldn't even make a difference as the vias and SMD inductors are definitely connected by traces.

also at what seems to be the GS ROM chip (see picture), it totally seems like three traces have burned up, but those actually connect fine. i've also done some in-circuit tests on all the capacitors, and just about everything that it was able to test seemed to test out fine, so i don't think the capacitors are an issue here.

well, i spoke too soon, the issue is not fixed, but it primarily seems to happen after multiple hour long playback. the only thing i can really add is that power cycling the unit at least sometimes won't clear it up immediately, though leaving it off for a certain time will make it go back to normal again. sometimes it can also spring back from distorted to normal output within the same track.

i might try a recap to give it a final shot, though again everything tested out well except a couple of caps that show "in-circuit/leaky", but that's kind of expected. but maybe the caps can drift during usage. i was not able to get any kind of reading on the L78MR05 regulator for some reason.

if there's no ideas i might just have to accept this unit is going to be flaky. since a couple of vias at the rear and front output jacks did get ripped out from the shipping damage i guess there is a chance there's a missing connection somewhere, especially if this board actually has an invisible middle layer. lastly this unit also has a replacement volume pot+MIDI switch that i got from roland, which look and work like the original parts, so i wouldn't think those would be any issue. of course i tested the soldering for shorts a million times already.

so, it looks like i finally identified the cause of this issue. after sitting powered on for a few hours, it started clipping again, but i finally realized the pattern here: it's just select instruments that are broken, most noticeably all of the GS drumkits, so a MIDI that doesn't have any of those could play absolutely fine. i tested the aforementioned pins for continuity again and lo and behold, now the middle of the three is severed, and pushing down a bit on it with the multimeter probe didn't help either. as expected, the unit is in this "clipping" state now.

as to why this was manifesting so sporadically, i can only assume that either heat from the regulator or minor vibrations from the table were enough to sever and re-connect that solder connection again. i have no idea whether this is some sort of corrosion issue from the factory or something hit those pins while flying around in the case during shipment. upon close inspection it seems like the little solder blobs on the end of those three pads are totally missing so maybe it is the latter.

i guess i'll see if this is fixable with some flux and fresh solder, or if scraping off the solder mask will be needed.

Nice sleuthing! Hope you manage to get it fixed without too much trouble.

well, as i had kind of suspected, those pads were practically totally gone to the point i'm surprised there even were some connections at all. anyway the result (after some retries) doesn't look pretty but the unit is working again.