Neo Cup 98 - Blank screen

Hi, anyone seen a blank screen on a cart before? I am thinking either the 273 has gone completely, or maybe the 74LS174. I've got a 174 on order, but just wanted to float this past people.

The sound is there, and the game is working, just nothing on the screen. It's all black for the Neo splash, then goes all white for the game. If you press something you can get a different coloured box to appear at the top. If I boot into unibios and try do a checksum etc there's no text at all - just a black screen whilst it does the checksum.

EDIT: It doesn't need cleaning (again) or any reflowing etc. All perfect and I've reflowed everything for good measure. It's annoying I cannot probe the board whilst its in the MVS... God I hope its not the 273!

EDIT: I've ordered a 74LS139 as well just to be sure. From what I understand, the 139 and 174 are used to select the C chips?

EDIT: It's not the 139 - just swapped that out. I forced 5v to pin 11 of the C ROMS, and then gnd to pin 11 of the C ROMS - nothing still. Would that suggest a dead 273?

I am thinking it can only be the neo 273. Connectivity between cart and 273 is fine. C ROM data bus connects straight to the cart. The only other thing at play here is the chip select which seemingly is controlled by the 139 and 174, and since I've forced the chip select constantly enabled and still got no output it must be the 273 I guess. If there was a problem with either (or both) of the C ROMS, wouldn't I still get text out of the S ROM and 273 were working?

EDIT: That sort of makes sense now. Since the 273 latches both the C and S addressing, that's why there's zip coming out on the displays. It's clearly not addressing the ROMs. I will try and find a doner 273 from a boot or something.

Something interesting I noticed just now when testing - If I press the reset button whilst the screen is black I get a quick glimpse of the display - eg. on the Neo splash I can see the 'Neo Geo' text as its supposed to be there but just for a split second. This makes me think the addressing is correct? Don't quite understand how that's happening. Surely someone has seen behaviour like this before? Also wondering if the CHA512Y is compatible with all single slots?

Just an update for anyone coming across the same issue - It was the 273. I did some experimenting by forcing chip selects on various chips and that made no difference. It was as if it was latching on initial addressing, then failing to do so after that first hit. Just to be completely sure I removed each C chip, and the S chip and retested at each stage. Connectivity all OK, just still a black or white screen.