exxos's Atari, Amiga & retro forum

Sounds a lot like a bad socket.

russellnash wrote: 15 Mar 2026 08:18 I have looked on the original photo before I made it 800 pixels wide for the forum and on very high zoom I can see that all three pads on JP2 were linked, that means that must have been how it was as it was working, so I have restored it to this state and checked continuity between all pads.

This is an invalid configuration and won't work.

Only 2 pads at a time should be connected, either top and middle or bottom and middle, if all 3 are connected you've shorted A19 and A23 which will cause a white screen amongst other things.

The point of 3 pads is that only 2 are connected at any one time.

I'd suggest you go back and remove the solder bridge and check for continuity between pads, by default the bottom and middle pads should be connected.

If they aren't then it's likely it was cut previously so by a process of elimination you should have the top and middle pads joined.

Looking in closely at your image I believe that the middle and bottom pads were joined.

I will reverse it. The link between the bottom two is the original PCB link fixed to the board, it was never cut.
Here is what it looked like before I did anything.

russellnash wrote: 16 Mar 2026 06:22 I will reverse it. The link between the bottom two is the original PCB link fixed to the board, it was never cut.
Here is what it looked like before I did anything.

a500jp2.jpeg

Cool.

The question is, where did the solder blob come from. Might be worth looking at other jumper pads, see if any have been cut and if so then they may need rejoining.

stephen_usher wrote: 15 Mar 2026 09:24 Sounds a lot like a bad socket.

Is the Agnes socket just a standard 84 pin PLCC socket, or something special, Stephen? Don't want to get the wrong one.

rubber_jonnie wrote: 16 Mar 2026 09:45

russellnash wrote: 16 Mar 2026 06:22 I will reverse it. The link between the bottom two is the original PCB link fixed to the board, it was never cut.
Here is what it looked like before I did anything.

a500jp2.jpeg

Cool.

The question is, where did the solder blob come from. Might be worth looking at other jumper pads, see if any have been cut and if so then they may need rejoining.

I am beginning to think it was just some sort of strange coincidence and red herring.

russellnash wrote: 16 Mar 2026 14:28

stephen_usher wrote: 15 Mar 2026 09:24 Sounds a lot like a bad socket.

Is the Agnes socket just a standard 84 pin PLCC socket, or something special, Stephen? Don't want to get the wrong one.

rubber_jonnie wrote: 16 Mar 2026 09:45

Cool.

The question is, where did the solder blob come from. Might be worth looking at other jumper pads, see if any have been cut and if so then they may need rejoining.

I am beginning to think it was just some sort of strange coincidence and red herring.

IIRC and I fitted a new one to my A500, the Agnus socket is standard, but if you search eBay for Agnus or Fat Agnus socket you'll get some decent results.

Red herrings happen, but better to get rid of a rogue solder blob than have it sliding around inside.

It sounds a bit that someone wanted to do a 1MB chipmem mod, but that will not work with a 8371 Agnus.
It would also give the issues you described.

Just remove the solderblob and leave the standard connection intact.

Yes, someone put an upgrade in there decades ago according to how it was stuck in place with all the green corrosion on it. Perhaps it had a different Agnus chip in it at one time but they didn't cut the track, anyway. I reversed it, it has changed nothing. I am going to take a break for a while, I have run out of steam for now, but I learned something about the Commodore Amiga A500. As always, you only find out stuff when it breaks :)

I replaced the logic chips U12 and U13 with a quick bodge up solder job. At first there was no change. I decided to tidy it up with sockets and still nothing happened until I changed U13 again, then pins 2 and 14 of the top row of RAM chips started pulsing. I did the same with U10 and U11, then the bottom row of RAM started pulsing on pins 2 to 14 also. I thought to myself, that must be it, but no, it still has a green screen.

On the Amiga A500 troubleshooter chart it shows the Fat Agnus and Gary chips as potential problems by green screen. I decided to swap these to see if it makes any difference and surprise surprise, it doesn't. The CAS, pin 15 on the RAM, stays stuck high. I wanted to try that before attempting to replace the Fat Agnus socket. The PCB is extremely delicate. After desoldering I normally clean out holes that are full of solder with a sewing needle in a drill. It has no cutting edge but the heat melts the solder away without cutting into the hole. This works generally very well, but the pads on this board are very prone to lifting when I do this. The trace remains intact and I just push it back down and solder it, but it is certainly something I have never seen before. The search goes on.