[hpsdr] Looking for guidance

[Jeremy McDermond]

On Nov 10, 2010, at 2:35 PM, Berndt Josef Wulf wrote:

> ***** High Performance Software Defined Radio Discussion List *****
> 
> G'day All,
> 
> Many thanks to all of you who replied to my email. I now have a better understanding about the project and 
> I'm currently favouring an Atlas based system. With this in mind, allow me to ask a few more questions:
> 
> * Are circuit diagrams for Mercury, Penelope, Metis and Atlas available?

Yes.  If you look on the Wiki, there should be links in there to the schematics for all of the boards.  Part of the OpenHPSDR project is having open source hardware as well as software.

> * Is the source code of firmware, including Verilog code for the FPGA's, available?

Yes.  It's in the SVN repository.  See http://openhpsdr.org/resources.html for how to get access to the code.

> * Can Mercury and Penelope operate independently, e.g. full duplex?

Yes.  In fact that's how I implement "monitor" functionality on MacHPSDR/Heterodyne.  I just don't turn off the receiver chain, and you hear whatever's being actually transmitted.

> * What kind of data is streamed between HW and PC?

The protocol document for the USB packets is in the SVN repository in /trunk/Documentation.  The Metis wrapper around those packets is documented in the same directory.

In short, Metis/Magister is sending you packets with 63 3-byte I & Q samples from the ADC, and 63 2-byte mono samples from the Mic connector on Penelope (there's a header too, see the docs for more details).  The computer sends to Metis/Magister 63 sets of samples composed 2-bytes for each of I and Q, and a 16-byt each right and left sample to go to the headphones connector on Mercury.  Again, the document is very complete, and it should answer most of your questions.  I used it to implement MacHPSDR.

> * Will Metis allow for full remote operation of RX and TX?

It's as remote as gigabit ethernet gets you.  Right now Metis only works on a local segment, and does not route anything through a router.  This doesn't necessarily have to be true, but consider that the data rate from Metis can be immense.  Technically speaking, though, since it's standard gigabit ethernet, you could use a fiber optic media converter and run Metis as far away as the fiber could reach.  If you wanted to pay the thousands of dollars for ZX optics, you could place Metis at the end of a 70km link.

> Not being a black box operator, I like to customize hardware and software to suit my needs and interest, 
> which includes ELF, VLF, LF and radio astronomy.

There's plenty of opportunity for that in the project.  People try to ensure that everything is as open as possible for playing.  That being said, it's also not a project that's necessarily easy to set up and get into, as you've already found.  Some things are rough around the edges because it truly is an experimenters project, and not a "product" per se.

> 73, Berndt
> VK5ABN

--
Jeremy McDermond (NH6Z)
Xenotropic Systems
mcdermj at xenotropic.com




 1289430860.0