TF CD32 Riser Revision 2 Design Complete

[PhilC]

Been a crazy week but we've proved that this riser will pass through FMV video.

@arkadiusz.makarenko Sorry i've not been keeping up with your progress this week.

Well done guys. Makes me want to get a CD32

[arkadiusz.makarenko]

OK.

After some struggle with implementation of microsecond blocking delay I managed to get following.
I hope it is good enough for Amiga. At this point it is from USB to Amiga. If this will work as expected I will have a look at sending reports back to USB keyboard (led light controls).

[terriblefire]

OK.

After some struggle with implementation of microsecond blocking delay I managed to get following.
I hope it is good enough for Amiga. At this point it is from USB to Amiga. If this will work as expected I will have a look at sending reports back to USB keyboard (led light controls).

Yeah dont think the delays need to be perfect. anywhere between 1/2 a microsecond and 2 microseconds will do i recon. Its a clocked signal after all.

EDIT: If you upload this to github i'll try and see if it works later

[arkadiusz.makarenko]

OK.

After some struggle with implementation of microsecond blocking delay I managed to get following.
I hope it is good enough for Amiga. At this point it is from USB to Amiga. If this will work as expected I will have a look at sending reports back to USB keyboard (led light controls).

Yeah dont think the delays need to be perfect. anywhere between 1/2 a microsecond and 2 microseconds will do i recon. Its a clocked signal after all.

EDIT: If you upload this to github i'll try and see if it works later

I have pushed changes.

[terriblefire]

I have pushed changes.

Thanks ... have you assembled up the riser for the CD32 yet or just testing on the bench?

[arkadiusz.makarenko]

I have pushed changes.

Thanks ... have you assembled up the riser for the CD32 yet or just testing on the bench?

It was on bench. Now I am trying to make it work on CD32, but it is not too successful. I managed to make it work at some point, but keys were very sticky. But most of the time it seems that stm32 is hanging or crashing.

[arkadiusz.makarenko]

Now I can't connect to the chip. It says Disable read out protection and try again...

[arkadiusz.makarenko]

OK All back to normal. I just reprogrammed chip. I assume there are anti-tampering features which may be turn on by default? Don't know.

Anyway. Keyboard seems very sticky, but I know why it is. At this moment data is sent only when value on keyboard is changed, which should work fine for mouse and gamepads, but not for this keyboard implementation. So for example if you hold button it will be recorded only once. I will need to move some processing from main to different place.

And I have updated initial config, as it was holding Amiga in reset until you connected keyboard and pressed button. Now it should be ok.
Not too bad for first try.

[terriblefire]

OK All back to normal. I just reprogrammed chip. I assume there are anti-tampering features which may be turn on by default? Don't know.

Anyway. Keyboard seems very sticky, but I know why it is. At this moment data is sent only when value on keyboard is changed, which should work fine for mouse and gamepads, but not for this keyboard implementation. So for example if you hold button it will be recorded only once. I will need to move some processing from main to different place.

And I have updated initial config, as it was holding Amiga in reset until you connected keyboard and pressed button. Now it should be ok.
Not too bad for first try.

For the CD32 its v important just to do the keyboard reset once even if you change keyboards... keyboard reset is system reset on CD32.

[arkadiusz.makarenko]

I have played a little with it, and I changed us delay implementation. Now it is marginally better than it was, but still far from perfect.

Edit:
New udelay is bad. DWT_Delay is actually much better (Measured on my cheap logic analyser).

But no luck, I start suspecting my CD32...