Another 1040 STFM and Exxos memory upgrade (1/4mb)

[rubber_jonnie]

Just with my moderator head on, could you please edit your posts so that the images are placed inline and not attachments. Thanks.

With regards to the problem, , I've been looking at C070789-001 REV.D mainboard pictures, and that large blue capacitor is not on any of the images I have or can find for that board. What's going on there?

I'd also agree with @Steve regarding the traces. I had a couple of these boards fail when I upgraded to 1MB in the original slots because of bad traces, and I had to fix them with bodge wire.

They looked just fine until I buzzed them out.

[derkom]

Just with my moderator head on, could you please edit your posts so that the images are placed inline and not attachments. Thanks.

I have taken care of this. But I will take the opportunity to remind people to please use the "attachments" tab and "place inline" to stick your images into your posts instead of leave them as dangling attachments. The forum is a lot tidier with the images placed into the posts.

[rubber_jonnie]

Just with my moderator head on, could you please edit your posts so that the images are placed inline and not attachments. Thanks.

I have taken care of this. But I will take the opportunity to remind people to please use the "attachments" tab and "place inline" to stick your images into your posts instead of leave them as dangling attachments. The forum is a lot tidier with the images placed into the posts.

Thanks, I was a little busy to do it earlier.

[ChaoticBiker]

Thank you for fixing the images. Will ensure everything is in line from there on

With regards to the problem, , I've been looking at C070789-001 REV.D mainboard pictures, and that large blue capacitor is not on any of the images I have or can find for that board. What's going on there?

That is the C49 which normally would be to the right of the new memory board. Not sure why but the instructions said to move it here. I've tried running it in both places, no change. You can see this being moved in similar manner in other mods on this forum - I can understand why it required to be moved with the older board as there was space constraints for the SIMM connector - not sure why instructions for the 2019 model still require it to be moved as space is not an issue anymore. Do not believe it is an issue here.
I have also moved the MMU pin wire to the throughhole the instructions recommend.

I'd also agree with @Steve regarding the traces. I had a couple of these boards fail when I upgraded to 1MB in the original slots because of bad traces, and I had to fix them with bodge wire.

They looked just fine until I buzzed them out.

Thank you for all your help so far.

Broken traces is what I'm currently betting at. It is very time consuming to buzz everything out. I already found 2 broken traces earlier, both placed higher up where the old chips used to be, I fixed them, and now tested almost everything that's available around the memory board.
I am getting very close to a point where I'll need to unsolder the board and check underneath it... or is there any way or schematic I could use to avoid this?
If not, well I'll try to desolder it tomorrow and we'll see from there.

[PhilC]

You may not need to desolder the board if you can think in reverse. You can test the board from the bottom with a multimeter.

[ChaoticBiker]

Just thought I'll update how are things.
Still no luck.

What I've done so far:

Found a 3rd broken trace via buzzing literally everything and fixed it. Each and every pin appears to buzz out using multimeter for continuity test. I also buzzed them against any point I could find elsewhere on the board - no issues now.

There are no shorts anywhere (checked all where I soldered and where the removed chips were)

Board has 5v power and both ram chips on the upgrade are getting 5.02v measured directly on their pins too.

I replaced that C49 16v 1000uF capacitor with a new one ( a 25v 1000uF), soldered back where it originally was meant to be.

Note I am still using 33ohm resistors instead of 68ohm ones in R71 R72 and R73 but even if this was the issue from what I read atari would start but would throw errors if anything, and with these boards it's the other waya around - people revert to 47ohm or 33 ohm as it can throw errors with 68ohm.

I still have some ideas to try (including further testing on the traces), but am open to any and all suggestions where to go next.

[rubber_jonnie]

Just thought I'll update how are things.
Still no luck.

What I've done so far:

Found a 3rd broken trace via buzzing literally everything and fixed it. Each and every pin appears to buzz out using multimeter for continuity test. I also buzzed them against any point I could find elsewhere on the board - no issues now.

There are no shorts anywhere (checked all where I soldered and where the removed chips were)

Board has 5v power and both ram chips on the upgrade are getting 5.02v measured directly on their pins too.

I replaced that C49 100uF capacitor with a new one, soldered back where it originally was meant to be.

Note I am still using 33ohm resistors instead of 68ohm ones in R71 R72 and R73 but even if this was the issue from what I read atari would start but would throw errors if anything, and with these boards it's the other waya around - people revert to 47ohm or 33 ohm as it can throw errors with 68ohm.

I still have some ideas to try (including further testing on the traces), but am open to any and all suggestions where to go next.

I can't recall if this was asked, but is this for both 1MB and 4MB settings?

Do you have a diag cart? If so worth trying that with a serial cable off to a terminal. I'm pretty sure it should spit out some results there even with no RAM. Worth a try regardless.

[ChaoticBiker]

I can't recall if this was asked, but is this for both 1MB and 4MB settings?

Do you have a diag cart? If so worth trying that with a serial cable off to a terminal. I'm pretty sure it should spit out some results there even with no RAM. Worth a try regardless.

yes this is both in 1mb and 4mb settings, behaviour is exactly the same.
Don't own a diagnostic cart - but willing to invest in one. Any advice, what exactly should I buy?

[rubber_jonnie]

I can't recall if this was asked, but is this for both 1MB and 4MB settings?

Do you have a diag cart? If so worth trying that with a serial cable off to a terminal. I'm pretty sure it should spit out some results there even with no RAM. Worth a try regardless.

yes this is both in 1mb and 4mb settings, behaviour is exactly the same.
Don't own a diagnostic cart - but willing to invest in one. Any advice, what exactly should I buy?

I'd suggest this: https://www.exxosforum.co.uk/atari/store2/#0165

You'll also need a null-modem serial cable and another machine with a serial port running some kind of terminal software. I can't guarantee it will show you anything, but it may, so up to you to decide if the investment is worthwhile.

BTW I know you checked the address lines, but have to checked cas/ras etc? Also, have you reseated the MMU? Handling these boards can lead to bad connections in the sockets.

Have a look here too: https://www.exxosforum.co.uk/forum/viewt ... =91&t=3143

[stephen_usher]

Handling these boards can lead to bad connections in the sockets.

Agreed, and other breakages. I always handle old PCBs with kid gloves, trying not to bump them or bend them and always lifting with two hands. They are just too fragile.

Also, reseating the socketed chips is always the first port of call if there's a failure. It's the simplest and safest thing to do.

Remember, with PLCC chips you REQUIRE a proper PLCC chip extraction tool (they're cheap). Also, try not to pull the chips out, the extractors are meant to jack them out as you push the two sides of the handle together. You *may* need to rock gently to ease the chip out but you have to do this ever so carefully.