I remember that I started using it but during the gameplay the image kept blinking, until one day there was no image, there was no sound, the disk was spinning, the VMU beeped but the display was off, at the time I left it aside and bought a VA1 ( the same as in the other topic, about GDROM).
After solving the VA1 reading problem, I decided to focus on this VA2.1 and started analyzing:
1- Power Supply OK!
2- Replacement of some electrolytic capacitors, mainly the one that forms part of the fan signal line. (Photo1)
3- measurements of important points with a multimeter/oscilloscope
In one of these measurements I realized that there was no 2.5v power supply for the Holly (GPU), but nothing indicated a short circuit, with the multimeter I realized that 2 points were not connected, the problem is that these 2 points are just below the Holly, I I would have to remove it to see what happened, beforehand I made a jumper connecting the 2 points (photo 2). After the jumper the 2.5v was activated. The TV appeared to receive some signal, but 10 seconds later it reported that it had no signal. The Holly kept getting hot, but in my opinion it was normal heating and didn't indicate that it was shorted.
I decided to remove the Holly with my hot-gun, during the heating process there was something coming out from under the GPU, but I don't remember if I added some flux when I did the reflow years ago or if it was some attempt by the previous owner. When removing the GPU, only pad 116, on the board, came out. The GPU pads were intact (Photo 8). On VA2.1 board, the flux looked like mud.
The 2.5v voltage section was actually broken into two parts (photo 3), probably due to oxidation and/or temperature stress caused by Holly, as the points are well below Holly's DIE.
I cleaned the pads on both the GPU and the board and tested all 500 pads on the board twice, the first time was with the oscilloscope checking the voltage from each pad and the second test was with the multimeter to see if each pad was connected to its destination (I had help from the Dreamcast VA1 diagram on the internet), in the tests I discovered that pad 379, 2.5v, was also broken. I remade pads 116 and 379, in addition to remaking the contacts of the 2.5v section that are below the Holly, I put UV paint to protect the electrical parts that were exposed (photos 4, 5 and 6).
My next step is to test the GPU pads and it will only be with the multimeter on the continuity scale, the GPU has an oxidation appearance on the paint that protects the tracks (photo 7), probably caused by the flux that was there.
Photo 1
Photo 2
Photo 3
Photo 4
Photo 5
Photo 6
Photo 7
Photo 8
