Falcon CA401332-002-J

[whomper]

@whomper Falcon

Y4 = Factory
37 = July 1993
0 = Pilot production
2 = Pal
008557 = number 8557

It seems it is a revision J with the following text on Version History web sitenot sure what it means:
"Strangely, there are records of Falcons from 07/1993 with this revision. Maybe done by an Atari dealer in post production to D/C PCBs?"

It is the one with clock patch v1.2 in which it works ok with Cubase audio but once DFB1X is installed, Cubase Audio experiences problems-> SDMA issues. Other than that, the computer works great accelerated.

IMG_9924.jpeg

IMG_9930.jpeg

IMG_9927.jpeg

IMG_9928.jpeg

[exxos]

:thanksyellow:

So is it a early version or a later one... @mikro?

Looking the the combel compares to mine the date on @whomper boards is 1993, mine is 1992..So I'd assume he has a later board..

@stephen_usher what revision was your board?

[dml]

The more usual fix (and the one recommended by R Czuba with the CT60 boards) is to add a termination resistor between the SDMA clock and GND.

The clock trace is long so terminating the line helps with clock edges & I guess reflections from the SDMA.

Not sure of the correct value but around 100R seems typical for this kind of thing, not less than 50R.

Maybe someone knows more about that specific fix? I'm going from memory here but pretty sure its terminating to GND and not a pullup (which would be more like 1k - 10k).


(enjoy my quality artwork below!)

term.jpg

[exxos]

@dml it was 47R. It stopped my falcon from working totally as the sdma voltage ended up at about 0.5v. It's why I started doing patches myself. But my V4 patch has series termination and while I had parallel termination, some people said it made things worse. The signal never seems bad to me either way. So it seemed more like a timing issue that noise.

Its ultimately why I found low logic level voltage on the sdma. Some had 2v DC offset. So that 47r termination actually crowbared the DC level so the logic low level voltage was more what it was supposed to be so I believe that was the actual fix not the actual termination of a signal. That is what was done on my version 3 clock patch, but also it was rather unknown about delays on the clock as well, so it had adjustable delays but that was more for when people were trying to use it with the Phantom accelerator and I just removed that option on the V4 as did not see any reason to keep it at the time.

It was all documentation here anyway https://www.exxosforum.co.uk/atari/last ... m#CP2022RS

[dml]

Interesting. I think one or maybe two of my boards has this mod. I should measure and see whats going on there and what value was used (I didn't make that change myself it was there already).