[DSP-10] ver380

[Luis Cupido]

Bob,

Thanks for your reply.
Ok, I understand clearly the implications, and yes
if it needs a deep structural change it surely became
another radio... DSP11 ?! ;-) I don't want that for
the moment.

Fine, I'm convinced, I'll try to see if I can understand
what comes out of the DSP and play a bit along that line.
I'll keep you posted.

Luis Cupido
ct1dmk.


Bob Larkin wrote:
> Hi Luis,
> 
> The 9600 serial baud rate is locked in tightly to the architecture.  The 
> AD21xx handles a single interrupt very well.  But, it is awkward with 
> multiple interrupts, in part because there are only two sets of registers 
> that can be switched simply.  So to keep things manageable, the 9600 is 
> derived by dividing the 48K rate for the CODEC by 5.  Then the same clock 
> runs the UART and everything else.
> 
> I believe the higher baud rates would require running the ADI style of UART 
> which runs off a timer. This then needs two interrupts.  All possible, but 
> not simple.  The DSPx would probably have the added power to support a 
> faster rate, but the EZKit is running out of "real time."
> 
> Another, again not trivial approach, would be to connect up a UART as a 
> piece of hardware.  That might make sense.
> 
> I have certainly had times when the 9600 seem restrictive.  On the other 
> hand, it is also easy to grow in all directions, like bandwidth and dynamic 
> range, in which case it becomes a different radio ;-)
> 
> Good things to think about, though.
> 
> 73, Bob  W7PUA
> 
> 
> At 03:00 AM 1/12/2009 +0000, Luis Cupido wrote:
>> Bob,
>>
>> How difficult it is to put the DSP working at
>> 38400b/s or 96800b/s on the serial port ?
>>
>> Would you think the mumber of packets could
>> be increased to 5 to 10 spectral updates per second ?
>> Or is there another impediment?
>>
>> Just dreaming out loud ;-)
>>
>> Luis Cupido.
>> ct1dmk.
>>
>> _______________________________________________
>> DSP-10 mailing list
>> DSP-10 at mailman.qth.net
>> http://mailman.qth.net/mailman/listinfo/dsp-10
> 
> 
> _______________________________________________
> DSP-10 mailing list
> DSP-10 at mailman.qth.net
> http://mailman.qth.net/mailman/listinfo/dsp-10
>