[hpsdr] Introducing Angelia

[Joe Martin K5SO]

Hi Phil, 

Passing the phase word instead of the actual frequency seems to improve things when using multiple Mercury receiver boards on Atlas but it's not good enough, I'm afraid.  The phase relationships between boards seem to be stable under frequency/band changes alright but the values change after a power cycle of the Atlas bus.  

It appears to me that the approach used in Angelia is the better method; i.e.,  multiple ADCs on a single board.   Unless, of course, I'm not implementing the approach you suggested in the precise way that you did.  Actually, we can't do it precisely the same way, as you know, because you are working with a single board, Hermes, not a Penny/Merc combination for your VNA.  

In my test I simply calculate the phase word in Metis as it normally would be done in Mercury then pass the 32-bit phase word to the Merc boards as one of the frequencis in the Atlas broadcast to the Mercs which then use it as a direct input to the receiver module(s) in Mercury instead of using the original C122_sync_phase_word variable that is calculated in Mercury.  

Thanks for the suggestion though; it does seem to help, but not enough for repeatable nulling/steering/DF applications with multiple Mercs, I think.  To solve this issue properly we'll need to put all the ADCs on a single board, in my opinion.  

Abhi and I are presently discussing how we might do that very thing using 4 ADCs on a single board on the Atlas bus.

73,  Joe K5SO


On Aug 5, 2012, at 6:13 AM, Joe Martin K5SO wrote:

> ***** High Performance Software Defined Radio Discussion List *****
> 
> Hi Phil, 
> 
> That is wonderful news!!  In fact, if the "phase word" approach works for multiple Mercury boards too then this may be the breakthrough approach I was seeking and could make using perhaps up to five Mercury boards for DF and beam steering feasible with the Atlas bus.  
> 
> I'll definitely check it out!  Great observation!   Thanks!!
> 
> 73,  Joe K5SO
> 
> On Aug 5, 2012, at 1:08 AM, Phil Harman wrote:
> 
>> Hi Joe,
>>  
>> There may be a solution to this issue.  I’ve been writing the code to turn Hermes into a Vector Network Analyser. IN which case I need to guarantee the phase relationship between the Tx and Rx is totally consistent and repeatable across numerous power cycles.  Since the Tx and Rx FPGA code  run off the same clock on the Hermes board then that is half the problem solved.
>>  
>> What I found is if I feed the same frequency to the individual frequency-to-phase converter code for the  Tx and Rx CORDICs then every time I change frequency the phase relationship varies in a totally uncontrolled manner.  However, if I calculate the phase word from the frequency and feed the phase word to *both*  CORDICs then the phase relationship is rock steady under all conditions that I’ve tested so far.
>>  
>> In which case if we run all the Mercury boards off the same 122.88 MHZ clock and feed the phase word, calculated in say Metis,  over the Atlas bus rather than the frequency then we may find an improvement.
>>  
>> 73 Phil....VK6APH
>>  
>>  
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at openhpsdr.org
> Subscription help: http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
> HPSDR web page: http://openhpsdr.org
> Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20120807/f113f3b8/attachment-0003.htm>