if you want to be amused, read on. :) Sharing my experience so maybe someone else can learn from it although I doubt someone here is equally retarded as I am.
A week ago I changed wiring in my CT2 Falcon because, well, it looked horrible, wires were dirty, old and most importantly - loose at some places what definitely didn't help with CT2 stability in general. Since I'm not so bad with soldering anymore I was like OK, it's 14 wires, most of them soldered to ICs and/or IC sockets.
So I was really careful, the end result was very nice (I'm going to post something about that later on), proudly switched the Falcon ON... and nothing. No video output whatsoever, both VGA and RGB.
Double checked, triple checked, ... all the connections, pins, contacts, CT2 pin numbers, motherboard solder points... nope.
So OK, my last possible hope - Jookie and his oscilloscope. For reasons I'm not going deeper into (you wouldn't believe anyway ;)) we were forced to meet at 4.30am (yes, that's in the morning). He checked my wiring, PCB, ... nothing suspicious.
Let's fire up the oscilloscope then... thanks to the F030 Service Guide we knew what to look for first - HALT and BERR pins. This finally brought something interesting, Jookie managed to change "no video" to "white screen" on the first reset. Sometimes. What's more interesting - you would think the white screen mean good, right?
Wrong. During the white screen the SDMA received 8/12.5 MHz instead of regular 12/25 MHz (/ = CT2 clock). During the no video state the SDMA clock was as it should. However both XHALT and EBERR on the expansion slot were in "1", as they should in both scenarios.
Actually, it wasn't that clear from the beginning because Jookie were speculating whether HALT (CPU) --> XHALT (expansion slot) doesn't mean inverted HALT however after checking in a working Falcon this theory was busted. Another funny mistake was measuring wrong pins because the Service Guide describes pins from the connector point of view, i.e. on the motherboard they are in reverse order (be careful about that!)
So that leaves us... nowhere because obviously the basic stuff works. Even Ax/Dx lines worked. So Jookie tried some 'magic touches', blows (?), even some CPU resoldering (on the CT2, with heatsink on!) ... nope. And then I would look at the PCB and ...
... put the memory card back. ;)
Yes. That was is. Naturally, the question "from what moment exactly this machine has started working?" was in the air and nobody could really tell. I'm not even sure whether I had *ever* put that card back before rechecking the wiring etc.
But anyway, it seems there must have been something wrong at least on some level because the white screen indicated something really messed up.
What does it all mean? Well... don't diagnose your Falcon at 5am in the morning. ;)