Page 10 of 60

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 29 Apr 2020 19:59
by arkadiusz.makarenko
@terriblefire

I am not 100% sure yet, as I have not finished reading datasheets on RTC, but we might be missing LSE oscillator, 32,768kHz for Vbat mode?

"The RTC is functional in V BAT mode and in all low-power modes when it is clocked by the
LSE. When clocked by the LSI, the RTC is not functional in V BAT mode, but is functional in
all low-power modes."

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 29 Apr 2020 22:58
by terriblefire
oh ok.. we can look at adding that for the next version.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 30 Apr 2020 08:53
by wairnair
Just checking in to let you know that you have audience :) I enjoy reading your updates and I'm waiting for the final board patiently.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 04 May 2020 22:14
by arkadiusz.makarenko
Ok. I started porting code for keyboard.
I may rewrite it in the future, but for now it is good enough.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 04 May 2020 22:22
by terriblefire
arkadiusz.makarenko wrote: 04 May 2020 22:14 Ok. I started porting code for keyboard.
I may rewrite it in the future, but for now it is good enough.
Awesome. Should be dead easy to port over the keyboard code. I'm going to write a module to allow the arm to respond to specific address requests.

EDIT: I think even though VBAT isnt working we should be able to get the rest of this working by feeding a fake value to the Amiga at boot. We can sort this out on the next revision.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 05 May 2020 00:15
by arkadiusz.makarenko
terriblefire wrote: 04 May 2020 22:22
arkadiusz.makarenko wrote: 04 May 2020 22:14 Ok. I started porting code for keyboard.
I may rewrite it in the future, but for now it is good enough.
Awesome. Should be dead easy to port over the keyboard code. I'm going to write a module to allow the arm to respond to specific address requests.

EDIT: I think even though VBAT isnt working we should be able to get the rest of this working by feeding a fake value to the Amiga at boot. We can sort this out on the next revision.
Keyboard support seems to be blocking code. Do you think this can have impact on communiation arm-amiga?

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 05 May 2020 08:53
by terriblefire
arkadiusz.makarenko wrote: 05 May 2020 00:15 Keyboard support seems to be blocking code. Do you think this can have impact on communiation arm-amiga?
Its interruptable code. arm-amiga will generate a fast interrupt. Do stuff. Return. Keyboard wont notice that.

Pattern should be.

Interrupt generated.

Read code from CPLD + Data direction (RW).
if write then grab 8 bits from PB0-7 and pulse an ack to the cpld (it completes the cycle)
if read then put 8 bits on the data bus, pulse ack via cpld and then tristate again.

Should be possible to do this in under 500ns.

The CPLD holds the Amiga until the arm responds.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 05 May 2020 20:16
by arkadiusz.makarenko
terriblefire wrote: 05 May 2020 08:53
arkadiusz.makarenko wrote: 05 May 2020 00:15 Keyboard support seems to be blocking code. Do you think this can have impact on communiation arm-amiga?
Its interruptable code. arm-amiga will generate a fast interrupt. Do stuff. Return. Keyboard wont notice that.

Pattern should be.

Interrupt generated.

Read code from CPLD + Data direction (RW).
if write then grab 8 bits from PB0-7 and pulse an ack to the cpld (it completes the cycle)
if read then put 8 bits on the data bus, pulse ack via cpld and then tristate again.

Should be possible to do this in under 500ns.

The CPLD holds the Amiga until the arm responds.
Now it is clear.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 05 May 2020 20:18
by terriblefire
arkadiusz.makarenko wrote: 05 May 2020 20:16 Now it is clear.
All set up on the bench.

Re: TF CD32 Riser Revision 2 Design Complete

Posted: 09 May 2020 10:22
by terriblefire
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.