Question on modded 64k Atari 600XL and SIO2SD reader

[repetto74]

Hi,

 

I started having a lot of issues when using an SIO2SD reader with a 600XL with internal 64k mod.

The machine is unstable and will at any random moment freeze and display garbled colored text and characters in a rolling pattern. A power cycle will set all back to normal for a short time before starting to freeze again.

Now I assume the mod is correct as the memory test is passing green with no fault detected. I cannot recall which tutorial I followed for this mod but there are maybe some more. Anyway it was posted here on Atari Age.

Is there any incompatibility I may not have been aware of concerning a 64k 600xL and any sd reader ?

[Rybags]

Try the memory test in a continuous loop and just leave it for an hour to see if it hangs.

Chances are you've got a bad connection somewhere in the memory upgrade. It's a simple upgrade so shouldn't be too much work to track down the problem.

 

I doubt the SIO2SD would be causing problems, though if your power supply isn't real good the extra memory + SIO2SD could draw enough extra power to make a marginal PS go bad.

Not sure what the power usage of SIO2SD is, I imagine the backlight and the LCD use most of the power.

 

For your testing, it's probably not a good idea to pull the SIO2SD off the computer while it's powered up, plenty of people have had the device become bricked which requires reprogramming of the PIC to fix it.

[repetto74]

Hi Rybags,

 

 

Thanks for your suggestions. I may start thinking also that the OS rom is somehow corrupted and unstable. The PSU may be also the problem but the thing is that even when disconnecting the SIO2SD and power recycling I will still have the problem popping out. It is like an issue popping out after a short period of time from a cold power on and once it is there I need to cool down for a while before turning the computer on again.

[repetto74]

Ok now the fault has moved to a worst scenario. I have a death RED screen on power up. Any idea where to check?

[Rybags]

Is this in all cases? If so then maybe your faulty OS Rom theory is right. A reddish brown screen that doesn't go to black after about 3/4 of a second means the OS startup isn't executing which can be one of several problems - CPU, MMU, OS being the most common.

[repetto74]

I am dumb sorry I did not swapped back the OS hip I tested on another working board (which is giving same crashes) so I was turning on without it. Anyway I had this red screen in the past sometimes together with those garbled video crashes.

[dakidski]

I had issues with the sio2sd from Lotharek and my 600Xl. When I got the one from Manchester_zp everything worked great.

[repetto74]

This is the kind of crash I have after a short period of time the machine is running :

 

 

[repetto74]

Hi Dakidski,

 

Problem is that I have this also with the SIO2SD disconnected...

[greblus]

Some time ago I bought 600xl and in my case the video glitches and crashes were caused by faulty CPU.

Stryker adviced me to run acid800 and at the begining all tests passed ok, but after a few tries it started to show errors during basic and illegal instructions tests.

 

http://www.virtualdub.org/downloads/Acid800-1.1.7z

 

After these problems were gone as an extra "bonus" I also found that the a PIA chip is broken . It was quite funny, because all games and programs I tested were working, doesn't matter what memory settings were selected in U1MB settings. But in 1088k mode when I was copying something through SIO in SDX, it hanged, while in 576k Compy Shop mode it worked fine.

 

Luckily my 600xl is fully socketed so it was easy to replace the chips.

[repetto74]

Well I have a spare of all main chips with me but I have tried to swap them all and had the same problem. Now I may try again one by one and see the result.

[_The Doctor__]

what are the two support chips that fail in the 800xl and what are the similar support chips in the 600XL? The failure resembles same issue covered multiple times with 800XL's so by extension could be the logical for that to start showing up in 600XL's. That's my hint...

 

so memory, support chips, mmu... that's the path to check and follow if you are sure it's not antic

 

C061618 in U2 is the mmu, you already know the ram chips, I will try to remember the support chips

Edited April 12, 2017 by _The Doctor__

[Rybags]

You've probably got a bad solder connection for the Ram upgrade. It's not like there's many of them so it's not a big effort to reflow the whole lot. Pry up slightly and reseat the CPU and MMU while you're at it if they're socketed.

[1050]

The 600XL has it's memory multiplexing hook up a tad bit different

than the 800XL, which in my opinion is just not right at all.

 

I mentioned this aspect in the 64k 600XL thread and declared

if I were to EVER attempt that mod, I would first go back and

rewire the 158s to 800XL standard so that the full 64k could be

refreshed properly in the first place.

 

I won't say this is the issue, I'm only offering a guarantee

of preventing this issue from being that issue. Please consult

schematics of 600XL and 800XL for further information.

[_The Doctor__]

1050 can you find your post and link it here... also please go ahead with your mod idea and post / construct it. I just bought the ram to upgrade a couple 600XL's and would like to play with the differences

Edited April 13, 2017 by _The Doctor__

[1050]

There is zero point to linking other thread, all information I have is in #14 above.

I don't recall if I even own a 600XL much less where it would be and wasn't going

to mod it at some later date either. You are pretty much on your own from here on

out, but I did look at the jerzy made 600XL schematic to see that address lines A14

and A15 are run thru an extra 375 that the 800XL doesn't use and the result of this

logic among other inputs is the 600XL /CAS signal. Just how the jumpers for 64k

mod done to the 600XL affects this circuit isn't clear to me because I did not look

into that. So if I were to 'fix' a 600XL to 800XL standards this extra 375 would have

to go too.

 

Was not going there, still aren't, sorry if you misunderstood that part.

The posts then are just a heads up to a potential issue caused by the difference

in address multiplexing done by the 158s on the 600XL.

 

http://www.dereatari.republika.pl/schematy.htm

 

As usual, caution is advised in using jerzy schematics as sometimes errors creep

in. One I just noticed is in the 800XL U27 notation for address lines used. It is a

simple repeat of the same lines used for U26 which is correct for U26 but not so

for U27. Proper address lines used for inputs on U27 are A4-A12, A5-A13, A6-A14,

and A7-A15 for the paired inputs on 1, 2, 3, and 4 inputs for the U27 158. Notice that

the address lines are offset all the way through by the number 8 on both U26 and U27

on the 800XL - this is proper and full address multiplexing for 64k ram. There are

other schemes, but this one is proven.

 

 

I have attached a somewhat corrected version of jerzy's 800XL schematic of interest,

not all connections are represented, notably pin one of ram RFSH is MIA. It is

tied by common rail to a pull up resistor R32 of 3k to +5 volts on the 800XL blueprint

schematic which in fact may be the missing schematic from FSM? As an addendum,

it is not attached directly to FSM since it's 18 inches by 24 inches in size unfolded.

And folded it is still oversize of standard 8 1/2 x 11 paper so a glued on pouch for it

on the FSM inside covers could not be used to contain it either. It could have been

folded six ways (9?) instead of just 2 x 2 - why not, I can not answer that one.

 

I will attempt someday to get Kinkos/FedEx to make a PDF of this document, both full

size and a shrink to four page version so the common man can just print it off to cut and tape

it together. Or use it as an exercise in the four page shuffle game.

 

Verifying the 600XL schematic in this particular area would be step one. I would need

to find a 600XL to even begin doing that and not very likely as I was way behind in

other projects going back years. I used MS Paint to alter the original gif for those

wondering how that was done. I make no claims to corrected data, just for the record.

Thanks jerzy - what an awesome total amount of work done there. Who can say they

have done more for the love of the machine. And not very many is the answer.

[repetto74]

Hi guys

 

Thanks to all of your for your intervention :-). I have figured out what was wrong. The IC delay module in U13 was the faulty one. I swapped it with the one coming from another board and all was back to business !

Where can I find this chip ? It looks this is a custom made one by Atari. Ref C060472-D. :-)

[Rybags]

Supposedly the delay line itself is a generic IC but Atari used one with the unused delay tap legs cut off at the factory.

The delay line is a vital part used for the generation of Ras/Cas for Ram accesses and switching signal for the multiplexed row/column addresses to the Ram chips.

 

There seems to be not a lot of info around about it or suggestions to what modern day part could substitute. Chances are the usual Atari suppliers probably carry it but AFAIK Best Electronics has the $20 minimum order.

[+ZuluGula]

GIS for this particular part number turned out this result.

http://www.atari.org.pl/forum/viewtopic.php?id=13365

[russg]

Hi guys

 

Thanks to all of your for your intervention :-). I have figured out what was wrong. The IC delay module in U13 was the faulty one. I swapped it with the one coming from another board and all was back to business !

Where can I find this chip ? It looks this is a custom made one by Atari. Ref C060472-D. :-)

Best Electronics has them.

 

http:/www.best-electronics-ca.com

 

408-278-1070

1:30 - 5:30 PM Pacific time

Edited April 19, 2017 by russg

[repetto74]

Hi guys. Thanks a lot for pointing to Best Electronic but sorry I did not précisé I am located in Old Europe :-) and shipping cost from California are just too high.

I will try to dig it out here in the continent