Re: ECB PropTerm

[Sergey <skis...@gmail.com>]

Yoda,

I was checking the possibility of re-using PropIO design before I
designed the ParPortProp board. (I believe a schematic of that attempt
is still somewhere on the Wiki), including the modification that
you're suggesting here - using an SR flip-flop to put the CPU in the /
WAIT state until propeller had processed the command and ready. And it
has some drawbacks (unknown FF state upon reset or power on, it is
impossible to reset propeller from the software, it adds wait
states...).

8255 on other hand allows a very simple decoupling of the Propeller
from the ECB bus. PPI in this case uses mode 2 which is intended
exactly for this purpose (and there is no bit banging necessary). It
also can support interrupt to indicate ready to send / receive data.

This approach has been tested with Zeta SBC and runs reliably even on
20 MHz (of course on the ECB systems the bus will limit the
frequency).

I don't mind designing the PCB myself. The PropIO board needs to be
updated anyway to fix this frequency limitation issue, and to make it
ECB/Kontron compatible.

Thanks,
Sergey

On Feb 12, 7:49 am, yoda <y....@r2d2.org> wrote:
> Why do we need another propIO which this is with a parallel port stuck in
> front of it?  PropIO works reliably under 6MHz and can be made reliable
> above that by just putting in a set/reset flipflop of the CS / enable
> connect to the wait.  I believe there are spare NAND gates on the board to
> do this but I have not had time to do that - seems it would be more
> productive to implement the solution on the current board and get a re spin
> then to make a whole new board.
>
>
>
>
>
>
>
> On Tuesday, February 12, 2013 2:37:16 AM UTC-6, Sergey wrote:
>
> > Hi,
>
> > I uploaded the initial schematics for ECB PropTerm:
>
> >http://n8vem-sbc.pbworks.com/w/file/63550295/ECB%20PropTerm%20-%20Sch...
>
> > Basically it is a ParPortProp together with an 8255 PPI and ECB
> > interface attached.
>
> > It provides simple and reliable terminal (and also SD card storage)
> > capabilities to ECB systems.
>
> > Perhaps schematic can be simplified a bit (I think 74LS241 buffers on
> > EBC are redundant).
>
> > I'd like to get any comments and suggestions...
>
> > Thanks,
> > Sergey