exxos's Atari, Amiga & retro forum

One "common point" the the 5 I highlighted is only under the PLCC GLUE.


It goes a bit "random" when mixing in all the others you highlight though.


Might be worth checking from GND to DTACK for example, on all the chips DTACK goes, using the GND closest to DTACK on the same chip that is.. Then see if there is any resistance changes. If you can detect a lower resistance on one area, it gives a clue. Though normal meters might not have the resolution for it..

Such a mystery - I thought I would just blast around the GLUE and CPU with my desolder gun. No change when finished with the GLUE however when running around the CPU socket the resistance between the DTACK and GND went from 1K4 to 600R. I think there may be a solder bridge under the CPU socket or something. Will have to carefully remove the CPU socket now.

Icky wrote: 01 Jan 2023 17:36 In fact IPL0, IPL1, IPL2, BERR, VPA, VMA, RESET, HALT, BGACK, BR have 1K4 to GND. Which is odd as for example the IPL0 line doesn't go near any chip on the board just like the DTACK but only has a pull-up resistor.

Just a crazy idea. Did you try removing the pull-ups? Just in case, for some reason, the pull-ups turned to pull-down? Or that is not possible?

ijor wrote: 02 Jan 2023 15:37

Icky wrote: 01 Jan 2023 17:36 In fact IPL0, IPL1, IPL2, BERR, VPA, VMA, RESET, HALT, BGACK, BR have 1K4 to GND. Which is odd as for example the IPL0 line doesn't go near any chip on the board just like the DTACK but only has a pull-up resistor.

Just a crazy idea. Did you try removing the pull-ups? Just in case, for some reason, the pull-ups turned to pull-down? Or that is not possible?

Nice thought but all the pull-ups are SMD resistor networks and all on the VCC bus mostly next to each other in a run. There are no pull-downs. That's why it's a mystery as there should be an open circuit with the chips removed from VCC to GND but this board has resistance. E.g. in the board layout below the RN15 it has the DTACK and RESET line on the 1K and RN12 which is RW is open between RW and GND.


I did think may be a power rail cap was accidentally in error replaced with a resistor. No such luck although in looking I did find a power rail cap that was half on a pad and just slightly touching the other. Fixed that and still the same issue.

At the moment it's a case by case basis here. i.e. am removing the CPU socket to see if that is the issue then on to the next if not until I end up with an empty board :( - not that it will come to that.

Icky wrote: 02 Jan 2023 16:14 At the moment it's a case by case basis here. i.e. am removing the CPU socket to see if that is the issue then on to the next if not until I end up with an empty board :( - not that it will come to that.

Stop messing about and get those 2 sockets off :chairsmack:

exxos wrote: 02 Jan 2023 16:16

Icky wrote: 02 Jan 2023 16:14 At the moment it's a case by case basis here. i.e. am removing the CPU socket to see if that is the issue then on to the next if not until I end up with an empty board :( - not that it will come to that.

Stop messing about and get those 2 sockets off :chairsmack:

I would but have just been crawling through old loft spaces this afternoon to remove an old 40-50's electrical cable that turned out to still be live and shouldn't.

Well there wasn't anything under the CPU socket. Still have got the issue. GLUE socket is next.

Icky wrote: 04 Jan 2023 22:42 Well there wasn't anything under the CPU socket. Still have got the issue. GLUE socket is next.

:( What the "gunk" on the SMT GLUE ?

exxos wrote: 04 Jan 2023 22:49

Icky wrote: 04 Jan 2023 22:42 Well there wasn't anything under the CPU socket. Still have got the issue. GLUE socket is next.

:( What the "gunk" on the SMT GLUE ?

That be some flux. I removed some of the solder on the SMD footprint as it looked like there might be some slight solder bridging. Haven't cleaned it up yet.

Sent some more tests and theories to you @Icky . I have a WIP new theory.

Problem with my current thought of it being solder.. I don't see where the solder could come from unless it was a pasting issue which was never cleaned off. Hard to tell until the GLUE is removed. But got a new new idea now :lol: