2600 4 switch rev 12 - Paddle 3 acts odd. Ideas?

[+-^CrossBow^-]

Yes I'm reaching out to see if anyone else has seen anything similar and to document what I'm finding on a strange issue I've got with the player 3 paddle control only a revision 12 4 switch I'm working on. Here is the actual issue that is taking place...

 

I use @Paul Slocum's test cart to verify full paddle functions on both controller ports for all 4 players. I do this using a single set of paddles that I know work and swapping them between the ports during the testing. On this console, paddle 3 will stay all the way to the left, and then as a point is crossed, it is instantly on the right hand side. There is no jittery motion or any motion at all. It is like a digital reading vs analog? 

 

In looking at the schematics, there isn't much of anything between the controller port and the paddle read pins on the TIA. Just a single capacitor. I've verified continuity between pin 5 of the second controller port to C220 and from c220 back to pin 38 of the TIA. No breaks in that connection and resistance is never above 1Ω on average and I get the same for the other read lines, so the traces are good.

 

Using the 2.6 diagnostics program set in controller matrix mode with the controller plugs in place, I see the pulse on pins 40,39, and 37. But pin 38 shows low at all times. I've pulled the TIA itself as it is a Signetics made one and it passes in my chip tester pro. So the TIA appears to be good. But in posting about it in the Backbit.io forums, @evietron did confirm that she is testing all of the control lines off the TIA but only digitally so it is possible that the TIA is bad in regards to the analog reads from pin 38 and would still pass the test given that it does show a change in the paddle but not until a certain point is reached as I stated above.

 

So any others with EEs have any other ideas for me to try before I resort to trying a different TIA? I don't have ready access to them as I would like but I can remove them from working units in my collection to at least test or possibly from donors I have. Although there is a reason those consoles are donors and not working units being used by folks.

 

And while I didn't test it outright, I did go ahead and pull the cap off c220 and put another in its place from a donor board. No change.

 

[+-^CrossBow^-]

Well, that was quick. I found an old 4 switcher in my garage I picked up years ago as a parts unit. Still had all the chips in it and it actually does work but all the switches are shot. Anyway, more important is that it had a fully working TIA. Popped it into this other 4 switcher and I've got proper response on all paddle lines again. So yeah.. apparently this other TIA has something messed up internally that can't read the full analog values of paddle 3. It can only see a 0 or 1 or equivalent that reads all the way left or all the way right. Nothing in between.

 

 

[zylon]

I find TIA paddle issue failures, to run much higher on 4sw and later models, since they removed the hex buffer from the lines. I was going to suggest using a test-TIA was best option, but you're already past that. :) I stick to L6ers for myself, but still tinker/repair the others. 4sw socketed board does make a great IC testor unit though.🤔