Off course if you'd done the tests in "correct" order you'd have saved yourself a few minutes, but ok that's enough of me being a smart arse 
and I promise this the last time I'll mention this!
I'm not familiar with the Rottendog boards, but I'd guess the switch area is fairly similar to the originals. They probably use different row switching transistors to dump the 5W resistors and the predrive transistors (this is a fairly standard upgrade even for old driver boards), but the circuit behind this will probably be the same. If the chips are socketed you could try and swap a few around to sort out if the fault is chip or board / track related. The switch PIA and associated buffer chips are the ones I'd swap if I could. In the original WMS design they are IC10 the PIA, IC13 & IC14 the col buffers and IC19 & IC12 the row buffers. It's probable the buffers aren't socketed so you may not have that option.
Do you have a schematic for the board and can you post up a snippet showing this area?
Ok Have now fixed the problem, Will get to that in a minute.
Yep, swapped the 6821 PIA's around - made no difference, and they were the only socketed chips on the board too.
Secondly I did the diagnostics test with the alligator lead and all the switch numbers were coming up fine until I got to pin6 on the 2J2 header.
Before then just 1 number would appear and then sequentially decrease as I moved down the header as it should indicating that particular switch number, but shorting out pin6 on the 2J2 header with any pin on the 2J3 header would give all sorts of odd numbers , but then move the alligator lead down to pin 7 on the 2J2 header and then the sequential decreasing of switch numbers would continue again all the way down to 1.
So , it would appear to be a fault with the board.
A visual inspection came up all ok, no broken tracks or shorts on the board.
Mate, pin 6 is the key pin on 2J2 on the Rottendog board as far as I can tell.
I assume it shouldn't be tested as it would give strange readings.
Yeah it is probably pin 4 then. Not sure wether you read pin # from top to bottom or vise versa, but either way the key pin was cut off anyway.
I will replace the molex connectors too just in case of any future problems.
Ok, I emailed Jim at Rottendog who was very helpful and offered to refund or repair the MPu, and suggested it was probably a faulty IC17 ( 74HCT240n).
I didn't want to waste maybe $80 to send board back to USA then back here again, so I opted to change ic 17 which in fact turned out to be the problem.
All switches in that column/row now work as they should.
I can only assume the chip was always faulty- certainly possible, or maybe there was a playfield switch problem which stuffed that chip in the process, so I guess fingers crossed it doesn't happen again.
Now I have just got to work out why a few resistors on the display driver board are burning and hopefully we'll be in business.
Thanks to everyone for their help on this, much appreciated.
Also, could not find any 74hct240n at the usual places and then I remembered A guy who used to have a store locally but now operates out of his garage not far from me.
Here is his website
http://members.iinet.net.au/~worcom/ .
He carries a lot of ICs and sends stuff all around Australia and takes Paypal and credit cards over the phone, so give him a call if you cant find that part you need at your local place.
Chris
