Dead VIC resurrection

[topcat]

My first Vic has been dead with a black screen since the end of the 80s but recently I've used the 'dead test' mode of the Penultimate+ cartridge http://blog.tynemouthsoftware.co.uk/201 ... -plus.html and amazingly, the test runs, and produces a test result showing a fault in 'RAM 0' bits 4-7. Just like the TOP LINE of this image (but IGNORE the KERNAL line as this is a sample image from the manufacturer's website) :-


It's an early VIC (not a CR version) and has 11 chips of HM472114P-3... I think the 'dead test' result means that the fault is one of these 11 RAM ICs, which I think means the pair of RAM chips covering the VIC's 6502 address ranges 0-1023.
But now I'm stuck... which physical chip is this on the PCB?
Do any of the experts here know which IC I need to change???

[mrr19121970]

Interesting.... I thought the concept of an external kernel was not possible

[srowe]

Assuming it's an original PAL then the section of the circuit diagram of interest is http://www.zimmers.net/anonftp/pub/cbm/ ... 4001_3.gif
The diagnostic seems to indicate the fault memory is bits 4-7 of RAM0 so the chip to replace should be UE2.

[srowe]

Interesting.... I thought the concept of an external kernel was not possible

I'm guessing that the image was created on a system when the original system ROM was physically replaced by the diagnostic image.

[topcat]

Assuming it's an original PAL then the section of the circuit diagram of interest is http://www.zimmers.net/anonftp/pub/cbm/ ... 4001_3.gif
The diagnostic seems to indicate the fault memory is bits 4-7 of RAM0 so the chip to replace should be UE2.

Many thanks! I agree this suggests this IC. I think I''ll also check the CS line on an oscilloscope, and solder in a socket for UE2.

[topcat]

I have good news to report!
The Chip Select line looked messy on a 'scope, but compared with a working VIC it looked very similar. That could easily have led me down the wrong path. I bit the bullet and snipped off each leg of UE2, and then de-soldered each leg individually. I cleaned up and soldered in a dip18 socket. I inserted an HM472114AP-2 into the socket as I couldn't source an original HM472114P-3. Looking at the datasheet, the AP-2 appears to be a slightly faster version of the original Commodore fitted P-3.
And... my VIC works again! 3583 BYTES FREE !!!
Thanks to user 'srowe' who helped locate RAM0!And thanks to Tynemouth software for implementing their 'dead test' mode!