Need advice: Upgraded 1200XL from 24 pin rev 10 ROM to 28 pin 800XL OS. Now won't recognize cartridges.

[ACML]

I upgraded a 1200XL from the stock 24 pin ROM (rev 10) to utilize 28 pin EPROMs with both the 800XL and Omniview OS on it.  I've done this about 20 times before and have had no issue.   As usual, the new OS(s) seems to work fine (repurposed the RF channel 3/4 switch as OS selector) without a cartridge, but when one is inserted, it does not recognize it and won't boot correctly.  I did verify the machine would boot the same cartridges prior to the OS mod (see photos).

 

1)  BASIC cart makes the farting boot sound, but no READY prompt (locked up). 

2)  Defender cart just goes straight to solid black screen.

3)  Ultimate cart hangs and internal light just flicker.

 

I've never seen this behavior before. 

Troubleshooting so far:

1)  Swapped Mexican Rockwell SALLY with Synertek and no change.

2)  Swapped 74LS08 with 74F08 and no change.

 

 

 

[Peri Noid]

Are you sure, you did not break any connection while desoldering old sockets?

[Edit]

I can see you just added some parts of sockets. The Wx jupers soldering looks a bit weird. Did you make sure, there were no shorts anywhere there?

And did you follow this description? 

 

Edited July 17, 2023 by Peri Noid

[ACML]

11 minutes ago, Peri Noid said:

Are you sure, you did not break any connection while desoldering old sockets?

[Edit]

I can see you just added some parts of sockets. The Wx jupers soldering looks a bit weird. Did you make sure, there were no shorts anywhere there?

And did you follow this description? 

 

I just added four SIP 2x1s to extend the 24 to 28 pin.  Original sockets still there.

 

I removed W6-9 and jumped W11-13 to go from 24 pin to 28 pin.

 

I retouched my solder pads to make sure they were good and did not bleed over.  Still same issue.

 

I can't tell you how many times I've done this mod.

 

 

[ClausB]

Could an EPROM have been damaged from soldering?

[ACML]

6 minutes ago, ClausB said:

Could an EPROM have been damaged from soldering?

I don't know.  Both OS(s) 800XL (0000-1FFF) and Omniview (2000-3FFF) appear to work fine.'

 

Clarification:  I'm using two 27128 EPROMs for U12 and U13 so the lower half (0000-1FFF) of each is 800XL and the upper half (2000-3FFF) is Omniview.

 

Edited July 17, 2023 by ACML

[_The Doctor__]

I would double check socket cleanliness and leg cleanliness, sometimes the old socket go bad in these as well, so you might have been better off replacing the sockets. The usual inspection for contaminants.

Also consider another set of eproms, they've been finicky at times, I've had some systems come to life by using a different brand of proms before. Wouldn't be the first time eproms were out of spec.

[ACML]

5 hours ago, _The Doctor__ said:

I would double check socket cleanliness and leg cleanliness, sometimes the old socket go bad in these as well, so you might have been better off replacing the sockets. The usual inspection for contaminants.

Also consider another set of eproms, they've been finicky at times, I've had some systems come to life by using a different brand of proms before. Wouldn't be the first time eproms were out of spec.

I don't think the sockets are the issue. I think it has more to do with the W6-13 jumpers and the traces.  I learned a valuable lesson today.  I said that I have accomplished this mod successfully probably 15-20 times, but always using my trusty cheap 1980's Radio Shack 30-40 watt solder wands.

 

Today was different.  Today, I decided to try my new solder station YIAUH 939D for the first time.

 

I was using 60/40 tin/lead solder with a melting temp of ~370 deg F.  I set the iron to 440 deg F and it melted the solder off the roll, but when I toughed it to the board to heat up the joint to remove the W7 jumper, it did not melt the existing solder.  So I increased the temp to 500 and not enough, then 600, not enough, then 700 and it melted the existing solder.  On the back burner of my mind, I thought that 700 deg F was a good temp for modern lead free solder, but too high for 60/40 resin tin/lead solder.  I think the critical mistake was the tip was to small of mass (very narrow and pointed).  That made me need to put more pressure and wait longer than the old Radio Shack 30 watt wand.  Basically, I think it destroyed the trace around the through holes.  Not good!  That will teach me to try something new and untested on a valuable retro main board.

 

I decided to return the W6-9 jumpers and return it to Rev 10 24 pin ROMs.  Now, I just get a solid dark green screen. So I bricked the board.  I have attached two photos.

 

What it should look like.

 

 

 

What it looks like now.

 

Look at the left side of W8 and W9.  I think I burnt the trace(s).

 

[+DrVenkman]

It’s still repairable - you know the pads you worked on, so you know the traces involved. Use some fine kynar wire to repair the traces and turn this into a “well-loved” project board.

[reifsnyderb]

They make pad repair kits for circuit boards.  A quick search on amazon shows they sell them.  It's possible other places have them available as well.

 

Also, sometimes it's hard to melt old solder joints on these boards.  If I have this problem, I add new solder to the joints to get them to melt.

Edited July 18, 2023 by reifsnyderb

[flashjazzcat]

Made a video about this a few weeks back (a 1200XL which had the exact same work done on it and no longer recognised carts). There was a broken trace on a data or address line under the front OS ROM socket (signal going from the CPU to the cart port via both OS ROM sockets).

 

Edited July 18, 2023 by flashjazzcat

[ACML]

7 minutes ago, flashjazzcat said:

Made a video about this a few weeks back (a 1200XL which had the exact same work done on it and no longer recognised carts). There was a broken trace on a data or address line under the front OS ROM socket (signal going from the CPU to the cart port via both OS ROM sockets).

Thanks for the video.  I added two (2x1) SIP extenders to make the 24 pin a pseudo 28 pin socket.  Your video gave me the idea to remove those 2x1 SIPs and see if the stock Rev 10 will boot.  If that does not work, I need to examine the W8 and W9 traces as DrVenkman points out.

[+DrVenkman]

23 minutes ago, ACML said:

Your video gave me the idea to remove those 2x1 SIPs and see if the stock Rev 10 will boot.

You shouldn’t need to remove the “extenders.” If the jumper configuration is correct, 24-pin mask ROMS work fine in 28-pin socket configurations. 

[ACML]

1 hour ago, DrVenkman said:

You shouldn’t need to remove the “extenders.” If the jumper configuration is correct, 24-pin mask ROMS work fine in 28-pin socket configurations. 

You are correct.  I just wanted to ensure that I didn't have some solder blob underneath that might have shorted an adjacent pin.  No joy, machine still immediately goes to solid green screen.  I've swapped ROMs and the SALLY, but no change.  I checked the traces going out from W9, which was the most suspect and they appear to be intact (photo is showing the flux around the contact).  I could get continuity at outer points following the traces.  

 

The keyboard fix went through without issue, the board tested great and the OS mod should have been a breeze.  What's that line from the original TOP GUN?  "That was some of the best flying I've ever seen, right up to the point where you got killed"

 

Edited July 18, 2023 by ACML

[Peri Noid]

19 minutes ago, ACML said:

You are correct.  I just wanted to ensure that I didn't have some solder blob underneath that might have shorted an adjacent pin.

It can be simply measured.

[+remowilliams]

3 hours ago, DrVenkman said:

It’s still repairable - you know the pads you worked on, so you know the traces involved. Use some fine kynar wire to repair the traces and turn this into a “well-loved” project board.

Kynar is definitely your friend, and trace repair/bypass should do you well if done carefully.  e.g. the 1200 I had to restore traces on that were cut for a previous upgrade...

 

 

 

[ACML]

21 minutes ago, remowilliams said:

Kynar is definitely your friend, and trace repair/bypass should do you well if done carefully.  e.g. the 1200 I had to restore traces on that were cut for a previous upgrade...

Is Kynar essentially 30 gauge wire?

[+remowilliams]

5 minutes ago, ACML said:

Is Kynar essentially 30 gauge wire?

 

Yes, 30 AWG solid wire wrap.

[+DrVenkman]

When I removed the two old 2K DRAM chip in my Sinclair ZX-81 two years ago, despite years of experience in removing ICs, I still lifted two traces off the board. Notice there’s not even solder mask there to help. Sir Clive’s best work, that, lol. Anyway, two short strands of Kynar bridging the pads which fit right under the new socket for the 32K SRAM upgrade and voila. 

 

 

Note: here I used flexible stranded wire for the actual jumpers to the SRAM for this upgrade since it requires bending a few legs out of the socket and soldering wires to them in lieu of cutting traces on the board intentionally. The repaired traces under the socket and chip are essentially invisible. 

[ACML]

UPDATE:  Life diverted me to other things, but I had a chance to revisit this board.  I looked under magnification and can't see any smoking gun.  I had de-populated the board of all ICs (VLSI, memory, 74 series TTL logic, 4050, MMU).  Basically anything socketed.  I re-populated the board with ICs new to this board and got the very same outcome, just immediate solid green screen.  I read up on AA threads related to the "solid green screen" phantom and a couple alluded to a possible issue with the delay line which is soldered in.  I chewed on that idea for a bit, then discounted it since the board, originally unmodified, had no issues and booted fine.   Then I removed W6-9 and added jumpers W11-13 to make it 28 pin.  It would boot and run fine, but testing farther I found it did not like it when a cartridge was present.  It would not boot a cartridge, but still no solid green screen behavior.  I removed W11-13, reinstalled W6-9 and put the 24 pins ROMs back in.  It's been green screen ever since.  I replaced the ROMs multiple times with known good ROMs (CO60617A for U12 and CO60616A for U13).  I believe what happened is that when using the new solder station, I left the iron on the board too long and the temp was too high which has damaged it in a way that I can't decipher. So I doubt the delay line just went bad when it worked fine and none of the work was near it.

 

I'm going to leave this one sit a bit and slowly gather more info when able, but it might take a while before I spend anymore time on this one.

 

 

 

[_The Doctor__]

betcha you have some bad sockets. clean the chips and use a straightener with new sockets for the 28 pin conversion area etc.