exxos's Atari, Amiga & retro forum

Don't know what it can be!

Is there a way of loading workbench from cd drive? Or how way to test it on built in 020 cpu?

arkadiusz.makarenko wrote: 22 Feb 2021 20:06 Don't know what it can be!

Is there a way of loading workbench from cd drive? Or how way to test it on built in 020 cpu?

All you have to do is hold both mouse buttons to access the early startup menu, and then you can move the mouse around and try. Otherwise you can boot from CU Amiga Coverdisks, if you have one, but it takes forever. EDIT: Also tried it on a second console with Akiko rev0, no difference. I need to find a way to upload a video here to show you the effect.

Ok, check this mouse video.

This is with 030 at 40 mhz, mouse simply plugged in, not moving, and the latest jed. It was crazier with the original jed. With 50 mhz 030 the effect is a lot more subdued but similar.

alenppc wrote: 22 Feb 2021 20:38 Ok, check this mouse video.

This is with 030 at 40 mhz, mouse simply plugged in, not moving, and the latest jed. It was crazier with the original jed. With 50 mhz 030 the effect is a lot more subdued but similar.

Three times the charm. It works ok on my just on 020 (boot menu).
Lets see if it is ok on yours.

EY, but this might yet be different issue on your video. It looks like address was not decoded correctly ...

Looks like the mouse is perfect on 020 now. Still glitches on 030 (like on the video). Hmm...

alenppc wrote: 22 Feb 2021 21:16 Looks like the mouse is perfect on 020 now. Still glitches on 030 (like on the video). Hmm...

Movement like this I had when A5 or A3 was missing, but it could be any address line either on cpld or arm...
But why it is on your 030 when it is perfect on my 030? Tomorrow will dig out different 030 cpu.

arkadiusz.makarenko wrote: 22 Feb 2021 21:40

alenppc wrote: 22 Feb 2021 21:16 Looks like the mouse is perfect on 020 now. Still glitches on 030 (like on the video). Hmm...

Movement like this I had when A5 or A3 was missing, but it could be any address line either on cpld or arm...
But why it is on your 030 when it is perfect on my 030? Tomorrow will dig out different 030 cpu.

If it was that, it would show up on 020 or 060 also, but it doesn't. So weird. I tried 4 different TF330s though, and they all suffer from this. Slowing down the CPU clock speed is the easiest way to reproduce this, as it is a lot less noticeable at 50 mhz.

alenppc wrote: 22 Feb 2021 22:05

arkadiusz.makarenko wrote: 22 Feb 2021 21:40

Movement like this I had when A5 or A3 was missing, but it could be any address line either on cpld or arm...
But why it is on your 030 when it is perfect on my 030? Tomorrow will dig out different 030 cpu.

If it was that, it would show up on 020 or 060 also, but it doesn't. So weird. I tried 4 different TF330s though, and they all suffer from this. Slowing down the CPU clock speed is the easiest way to reproduce this, as it is a lot less noticeable at 50 mhz.

Issue is that few people use it on cd32+tf330 setup and no one complaint before about similar issues. There was report about 020 issue which I solved today.
@kulamario built 5 sets recently and all pass... so the only variable is CD32 itself?
Can be there anything which could cause this issue?

What I can think of is that I am perhaps using a different version of the ARM chip? Maybe I have a wrong resistor somewhere?

I tried this on two different consoles, I have a third one so I can try there too.