Hi,

I got a broken V3 3000 that doesn't output any signal.
I tried the trick with another PCI graphics card to test if it would be detected on Win98SE, but no way.
The Avenger chip remains dead cold when powered up which is not a good sign... No clock signal either on the crystal.
The original heatink has been removed by the previous owner to repaste it. I know that the heatsink removing could be kind of fatal for the BGA chip if not performed properly.

I also desoldered the BIOS chip to backup it (just in case).

I finally decided to desolder the Avenger chip to reball it.
During the desoldering process, 2 pads on the PCB were pulled off. I don't know exactly if it was due to the process or if they were already weakened. Fortunately I fixed them with new copper trace/pad and UV solder mask.

I made some measurements on the bare Avenger chip :
Firstly I measured the resistance between Vcore and GND pins : very high or open circuit which is not a good sign
Then I measured between these two terminals but in diode mode and reversed (+ on GND, - on Vcore) : I get 0.45V which is a diode behaviour. Good sign! It's not in open circuit but not shorted either.
I finally applied some voltage up to 2.7V between Vcore and GND : the current is very low, barely 1mA. Maybe the chip need some king of other "enable" signal to start drawing current...

On the board, without the Avenger chip I measured about 260 ohms between Vcore and GND (across a tantalum capacitor). A friend of mine measured on the same place, on a working V3, 250 ohms.

I have now to resolder the chip on the board...

What do you think about that ? Cold Avenger = Dead Avenger ?

Cheers,
JB