[hpsdr] Software development and DttSP

[David McQuate]

Hi Alberto,
  The source code for the DOS programs used by the script file, 
initozy11.bat
would contain the info needed to load the Ozy FPGA code (and thereby 
"awaken" Ozy).
One of these days I'll try to extract the necessary bits.

  Mercury receives "command and control" bytes as part of the regular 
transmission of
samples to it.  The details are contained in a document I mentioned my 
recent posting to this list.
Included in this info are the NCO frequency, the sampling rate, the 
preamp state, etc.

(I believe that the anti-alias filters frequencies scale directly with
the clock frequency.)

A side note - DttSP source can be found at
https://www.cgran.org/browser/projects/dttsp

look in branches / ab2kt / sdr-core / base

In the sub-directory doc, there's an interesting paper "DTTSP.doc" 
describing how DttSP works.

73,
Dave / wa8ywq

--Alberto wrote--
In this vein, I think that it could foster some software development for 
HPSDR the availability of a detailed list, step by step, of what is 
needed to get data out of Mercury,
> to send commands to Mercury and audio data to Penelope, etc. etc., to 
> not mention Janus.
>
> What I mean is something rather detailed, without assuming any 
> knowledge from the would-be implementer.
>
> - Is libusb the way to go ?
> - Does it exist an API list for libusb ? Where ?
> - Which commands need to be sent initially to Ozy to awake it ? 
> Preferably without resorting to a script file, but directly from the code.
> - When Ozy is no more sleeping, how can I convince it to talk with 
> Mercury ?
> - Which commands Mercury knows and acknowledges ?
> - How can I change the NCO frequency ? And the sampling rate ?
> - Are there commands to set the anti-alias filters when changing 
> sampling rate or Mercury is smart enough to do it by itself ?
>
> That is a non exhaustive list, with the only purpose to make clear 
> what I have in mind when asking for detailed information.
> I don't know is something similar already exists. I know, I could read 
> the PowerSDR source code, but, even if I haven't tried
> that yet, I think that would be a very time consuming way. An 
> implementer's guide would be much better.
>
> Thanks for listening
>
> 73  Alberto  I2PHD


 1235095299.0