Icky - Progress

[Icky]

Did you do the image test ?

FC2 is on RP8.

Ah yes. Missed that on this small screen.

Yes I did the image test. Sometimes it looked almost perfect (not very often). Sometimes there were parts not quite right.

I’ll post photos tonight to show.

The screen is definitely better since the GLUE Dataline swap but still not there yet.

EDIT: would be good to get a pic of the image from an original ST.

[exxos]

I haven’t had time to run all the tests and decipher the messages from the corrupted display and over written terminal (term does not clear between each page so keeps overwriting itself - too little time last night to sort that one out).

Works fine with putty, what are you using ?

However I did see the message bus error in the some of the ram tests. The messages werent very clear so will have to try tonight to sort the term refresh clearing of the screen.

odd.. didnt the atari cart pass ram test ?

[exxos]

Is the MMU getting VSYNC ?

Diagrom might not work with 4MB yet.. I think it only finds 2MB ?, unless that was fixed...

You could lift MAD9 (R909) and wire the RAM card side to 0v, to enable 1MB. Then try diagrom again..

[Icky]

Redo the DiagROM test on a known working ST. I remember from my tests with DiagROM that the IRQ tests failed even on a good machine. I think this is simply one of the areas where DiagROM hasn't still been ported from the Amiga. So the failures you get are most likely not indicative of any problems with your board.

BTW: "No (I6) error" from the Atari diagnostic cart is not an error.

Thanks - No error for I6

Will when I get time compare original and STF on DiagROM

[Icky]

I haven’t had time to run all the tests and decipher the messages from the corrupted display and over written terminal (term does not clear between each page so keeps overwriting itself - too little time last night to sort that one out).

Works fine with putty, what are you using ?

Mac - will try on Windoze with PuTTY

However I did see the message bus error in the some of the ram tests. The messages werent very clear so will have to try tonight to sort the term refresh clearing of the screen.

odd.. didnt the atari cart pass ram test ?

Yes - it did. I think in the later thread you hit on it only looks at 2M.

[Icky]

Is the MMU getting VSYNC ?

Diagrom might not work with 4MB yet.. I think it only finds 2MB ?, unless that was fixed...

You could lift MAD9 (R909) and wire the RAM card side to 0v, to enable 1MB. Then try diagrom again..

I now have a list of things to do tonight. Work is getting in the way of Atari fun.

[PhilC]

I can confirm that vsync is wired to the Mmu as it's one of the things I checked the other day.

[exxos]

If it display the "girl" picture on diagrom initially, then what changes to start getting corruption.... its like something isn't connected and something ends up random corrupting the screen data..

if the atari cart passes ram tests, then its not that... must be something wrong with the shifter reading / being loaded from ram or something. I don't know what though.

[exxos]

I keep checking the schematics over and over and I am either missing something stupid every time or there is just nothing else wrong anywhere.

I think it really needs a fresh pair of eyes to look the whole thing over make sure I am not missing something somewhere.

The only thing I can do is check the lib pinouts do actually match the IC pinouts, they look correct in the schematic, but its the only thing I can really think of checking now.

Only thing else, is to scope out every pin on all the main chips, make sure they all switching, and make sure there is no really bad ringing on any of the signals anywhere.. also note that if there is a problem on some pin somewhere, the scope loading may improve issues or make things worse... its really looking for "leads" to start investigating... Currently I am at a loss where to even start looking :shrug:

[troed]

FWIW the screen glitches look like timing's off between MMU and Shifter, or just between Shifter and RAM. Not sure that affects anything else.