STV - Sprite Layer Missing?

[tika]

I purchased a STV with a Columns 97 cart and it's displaying visual artefacts when running with a game cart inserted or without. The sound and inputs work fine, it's just visually that things are off.

The ROM/RAM test completes, with IC29, IC30 and IC33 coming up bad. It looks like a sprite or background layer is missing from all of the video - see the pics.

I'm looking to replace these 3x ICs on the mobo, although it's odd that 3x IC's (all of the same model) have all failed and was hoping that someone had come across this issue/similar issue previously and could provide some insight (perhaps a memory controller for the VRAM or similar is borked?).

I've read that the Saturn VA 0.5 schematics is close to the ST-V, but after looking at these schematics, IC29, 39 & 33 are different IC brand and models, so I'm not convinced that the VA 0.5 schematics are super useful.

Any help would be appreciated!

 

[MetalliC]

it smells like VDP1 has been died, or something wrong with its connection to SCU chip, so, all the 3 RAM chips connected through it doesn't pass RAM test. not sure which exactly chip it is on ST-V, it might be 315-5689 or 315-5883

 

[ShootTheCore]

Try reflowing the custom chips first before you replace them. Sega didn't use much solder on that board and the flex when inserting cartridges which can break the connections between chip legs and pads.

 

[tika]

Thanks for the info! I remembered there was a diag menu on ST-V boards when holding down SW1 and powering up the board, and ran a bunch of tests. Most tests pass, although the VDP RAM checks fail. Running the VDP RAM checks multiple times, the VDP 1 FRAME_BUFFER_0 and FRAME_BUFFER_1 checks alternate in which IC fails between each run - the VRAM always succeeds.

The CRT test also looks off - I suspect the whole CRT test screen should have solid thatched lines without any holes in it - perhaps it's due to the bad VDP 1 RAM?

I reflowed both frame buffer ICs, which has made a difference with the sprite/background layers being shown now, but still showing visual artefacts and the diag tests are still failing, so I'm thinking the RAM IC's could be bad and replacing them might be the next course of action

 

[ShootTheCore]

Have you reflowed the VDPs?

 

[tika]

I reflowed the VDP 2 and IC 33, and the sprite layer has come back - thanks for the suggestion! The CRT test looks fine now, with no holes in the thatched display/output.

I also reflowed VDP 1 and the associated frame buffer IC's (IC 29 and IC 30), but they're still failing the diag/mem tests (and alternating failures between the two IC's between diag runs). I guess the only thing to do now is replace the VDP 1 IC?

Would the SMPC or SMC controllers have something to do with the VDP 1 IC's throwing errors (ie: not synchronizing the VDP 1 with other parts of the system)?

 

[ShootTheCore]

I’m not an STV hardware expert, but since you’ve already had some success from reflowing parts, you may as well reflow the SMPC and SMC - as you say, it’s a complex platform with a lot of interdependencies.

If it still flunks the diagnostics after that, swapping the VDP1 with a donor from a Saturn makes sense to me.

 

[tika]

Yeah I'll find a broken Saturn in my collection and replace the VDP 1 on the STV - hopefully this fixes it up. Thanks for your help!