[hpsdr] Multiple RX issues

Helmut dc6ny at gmx.de
Tue Oct 14 00:14:04 PDT 2014 

Hi Bob,

Let me add to Joe’s fine clarifications: There exist some considerations to
eliminate the bottleneck Atlas. Please see Phil’s VK6PH presentation
regarding to Gen2 Hardware at
http://video.openhpsdr.org/HRF2014/News+Future%201.2.mp4
BTW amp linearization by pre-distortion is also possible with standard
Atlas-based Mercury and Pennylane using the previous filters on Mercury, but
not with PowerSDR yet. Alfred, HB9EPU, implemented a very effective
algorithm in his own specific software environment which improves at least a
300W LDMOS amp up to more than -50dBc IMD.

73, Helmut, DC6NY

-----Ursprüngliche Nachricht-----
Von: Hpsdr [mailto:hpsdr-bounces at lists.openhpsdr.org] Im Auftrag von Joe
Martin K5SO
Gesendet: Montag, 13. Oktober 2014 23:22
An: Bob Cowdery
Cc: hpsdr at lists.openhpsdr.org
Betreff: Re: [hpsdr] Multiple RX issues

***** High Performance Software Defined Radio Discussion List *****

Hi Bob, 

It’s been a while since I’ve looked at the issue with regard to the
Atlas-based system, but I seem to recall that the reduction of the number of
receivers on Mercury was mostly due to the inability of the Atlas bus to
adequatley handle the data rates at the higher sampling rates and multiple
receiver configurations everyone was seeking.  The reduction from four
receivers in the Mercury firmware occurred at the same time that the new
polyphase-filter receiver modules were implemented and the higher sampling
rate options were implemented.  We did have a four receiver version of
Mercury firmware, v3.2 I believe, but it was limited to 192ksps for sampling
rate and did not have the numerous C&C selectable options that the current
Mercury firmware provides.  

To utilize the full potential of cuSDR (or dual Rx PowerSDR with PureSignal
for that matter) it is necessary to use platforms other than the Atlas-based
HPSDR radio, at least at the present time.  

There was some effort a while back, I believe, to implement PureSignal for
the Atlas-based rigs but it has not been successfully achieved to my
knowledge.  The simple truth is that the high data rates that are required
for multiple receivers (more than 3) are accommodated better with
single-board platforms (Hermes, Angelia, Orion) because they do not have the
speed issues that are associated with the Atlas bus.  cuSDR should work fine
for you as long as you limit yourself to 3 receivers when running an
Atlas-based system.  PowerSDR internally limits things to 2 receivers when
the “HPSDR” hardware (Atlas-based system) is selected, I think; other
hardware selections (non-Atlas-based selections) within PowerSDR permit
automatic use of greater numbers of receivers, as for example, to support
PureSignal, etc.  

73, Joe K5SO

On Oct 13, 2014, at 2:49 PM, Bob Cowdery wrote:

> Hi Joe,
> 
> Well, that certainly explains it and why Rx4 is all zeros. I really 
> hadn't appreciated that as I took the spec at face value and seeing 
> CuSDR allowing 7 I thought it must be implemented. Is that a hardware 
> limitation with Atlas do you know?
> 
> 73, Bob G3UKB
> 
> On 13/10/2014 21:34, Joe Martin K5SO wrote:
>> Hi Bob,
>> 
>> Mercury v3.4 implements a maximum of 3 receviers in the firmware.  
>> 
>> Metis_v2.6, which I think you said you are using, allows for 4 receivers,
in principle, but that does you no good on the Atlas-based system as Mercury
firmware limits the max number of receivers to 3.  
>> 
>> 73, Joe K5SO
>> 
>> On Oct 13, 2014, at 2:18 PM, Bob Cowdery wrote:
>> 
>>> ***** High Performance Software Defined Radio Discussion List *****
>>> 
>>> Just confirmed that the I and Q data for RX4 is all zeros so that 
>>> has to be coming out the hardware. Any ideas anyone?
>>> 
>>> 73, Bob G3UKB
>>> 
>>> On 13/10/2014 20:43, Bob Cowdery wrote:
>>>> ***** High Performance Software Defined Radio Discussion List *****
>>>> 
>>>> Hi Warren,
>>>> 
>>>> Thanks for the input. I'm only running 48K and I set CuSDR to 48K 
>>>> as well. I don't think its running out of steam. CPU is about 10% 
>>>> with 4RX and network about 25%, but only goes up to 27% with 5 RX. 
>>>> Still higher than I expected though.
>>>> 
>>>> The strange thing is that with 4 RX, the first 3 still work 
>>>> correctly and there are no dropped packets or sync errors. I've not 
>>>> checked the actual data of the 4th RX yet but I'm assuming its all 
>>>> very low numbers close to zero as there is no audio and panadapter 
>>>> flat lines at the bottom on CuSDR.
>>>> 
>>>> Just to see what happened I tried 6 and 7 receivers, still no 
>>>> dropped packets, rubbish out the audio and usage 15% CPU 31% 
>>>> network. I still think something is going funny in the firmware.
>>>> 
>>>> 73, Bob G3UKB
>>>> 
>>>> 
>>>> On 13/10/2014 19:57, Warren C. Pratt wrote:
>>>>> ***** High Performance Software Defined Radio Discussion List 
>>>>> *****
>>>>> 
>>>>> Hi Bob,
>>>>> 
>>>>> Perhaps your problem with sync errors when the 5th receiver is 
>>>>> enabled is that the 100Mbit interface will not support five 
>>>>> receivers at 384K.  IF you are trying that at 384K, you'll need to
back down to 192K.
>>>>> 
>>>>> I'm not sure I understand exactly what you mean about "no data and 
>>>>> thus no audio."  If you are running PowerSDR, it only generates 
>>>>> audio for RX1 (and its sub-receiver) and for RX2.  RX3, RX4, and 
>>>>> RX5 are used for things such as the stitched panadapter, 
>>>>> PureSignal, and diversity reception.  I'm not familiar enough with 
>>>>> cuSDR to comment there.
>>>>> 
>>>>> Hope that helps.
>>>>> 
>>>>> 73,
>>>>> Warren  NR0V
>>>>> 
>>>>> On 10/13/2014 11:52 AM, Bob Cowdery wrote:
>>>>>> ***** High Performance Software Defined Radio Discussion List 
>>>>>> *****
>>>>>> 
>>>>>> Hi,
>>>>>> 
>>>>>> I've been playing around with multiple receivers on some software 
>>>>>> I'm writing and something strange is going on.
>>>>>> 
>>>>>> I can direct audio from each receiver to different local output 
>>>>>> devices left/right/both so I can hear what's going on on each RX. 
>>>>>> Everything is fine up to 3 receivers, the 4th receiver gives no 
>>>>>> data and thus no audio and the 5th starts giving sync errors i.e. 
>>>>>> the 7f7f7f is not where it should be.
>>>>>> 
>>>>>> I installed and ran up CuSDR32 and I get exactly the same result. 
>>>>>> I can't hear what's going on but RX4 has no panadapter output and 
>>>>>> RX5 continually flashes the sync indicator.
>>>>>> 
>>>>>> I've not updated the hardware for a while, using Metis 2.6 and 
>>>>>> Mercury 3.4. Anyone know if there are any known issues with these
versions?
>>>>>> Failing that is anybody getting receivers above 3 working with 
>>>>>> those versions.
>>>>>> 
>>>>>> Thanks for any help
>>>>>> 
>>>>>> 73 Bob (G3UKB)
>>>>>> _______________________________________________
>>>>>> 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/
>>>>> _______________________________________________
>>>>> 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/
>>>> _______________________________________________
>>>> 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/
>>> _______________________________________________
>>> 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/
> 

_______________________________________________
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/


 1413270844.0

More information about the Hpsdr mailing list