Tips for fixing Wersi LiveStyle TGM1 (faulty WIC906)

Moderator: happyfreddy

zjanosy
Beiträge: 14
Registriert: 8. Jul 2021, 20:22
Kontaktdaten:

Tips for fixing Wersi LiveStyle TGM1 (faulty WIC906)

Beitrag von zjanosy »

Dear Friends,

I have spent quite some time and effort on repairing two TGM1 boards. In both cases the culprit was the WIC906 timing generator chip.

I had one board, which worked well without the EFM1 effect board, but gave nasty feedback noise with the EFM1 board installed. The other board had a strange problem: some of the partials seemed to be played back at a wrong frequency, causing distorted, heavily modulated sound. This was especially noticeable on patches with lot of high-frequency content, like e.g. the harpsichord. Furthermore It had strange release sample problems, retriggering some sounds when a chord was held and a new note was played.

My first guess was that either the EPROM or the MEM2 board were faulty, because these store the sampled sound data, but I was wrong. Exchanging these components between the "almost good" board and the bad board did not make any difference.

Next I tried to reverse engineer the board, while looking for faulty components or traces. I've started with the WIC906 chip, because there have been several reports of failed LiveStyle boards that have been fixed by replacing this chip. I have measured all the pins and compared them on the two boards. I have found two differences: on the "almost good" board the clock signal going to the EFM1 board was missing, while on the bad board there was another clock which did not work.

Since many signals seemed to be available on several pins, I have tried to find a "replacement" pin. I was lucky to find out, that the failing clock of the "almost working" board was indeed available on another pin as well. I've disconnected the bad pin, and connected the coupling resistor to the other pin, and voila, it worked!

On the bad board I could not find a replacement clock signal on another pin, only the inverted signal was available. I could have built a simple inverter, but since I already had a non-working TGM3 board, which I bought for parts, it seemed to be easier to replace the WIC906 chip. However, with the "new" chip the board did not work at all. I figured out that now another clock signal was missing. However, this time I was lucky, because this signal was available on another pin. I've connected them, and this board got fixed, too.

I have made a table describing the various signals on the WIC906 chip in order to make troubleshooting easier. There are quite a few signals that are available on several pins, so with luck you can also fix the board by wiring another pin to the missing signal. Of course I cannot be not sure that the signals are always the same, and do not change according to patches. It is also possible, that on other boards the chip is configured differently, but at least this method worked for me.

I've attached the table. I also made high-resolution scans of the boards. I've also created a (rather incomplete) schematics, mainly focusing on the connections between the timing generator chip and the other ASICs. The json file can be opened with EasyEDA, the svg in Microsoft Edge or in any vector graphics program.

I hope this helps someone to fix their LiveStyle board.
Dateianhänge
Wersi TGM1_MEM2D_schematics.zip
(218.35 KiB) 44-mal heruntergeladen
Wersi_MEM2D_scans.zip
(1.24 MiB) 37-mal heruntergeladen
Wersi_TGM1_scans.zip
(7.95 MiB) 38-mal heruntergeladen
Wersi_TGM1_WIC906_signals.zip
(405 KiB) 39-mal heruntergeladen

Wer ist online?

Mitglieder in diesem Forum: 0 Mitglieder und 2 Gäste