[Skimmertalk] [RBN-OPS] Red Pitaya 122-16 overload levels in CW Skimmer duty - DATA

[Jeff Blaine]

Got the RP122-16 checked out last night.

Test method:

Elecraft KG3 as CW test source
choke
HP 355/D/C step attn
9 pole LPF
7 pole LPF
MC 3:1 splitter

With the KG3 set at -33 dBm, the net output after losses of the filters 
is -40 dBm.  The filtering is needed to be sure the KG3's square wave 
and thus inclusive of 10m content (which runs about -35 to -40 dBc) is 
reduced to undetectable levels.

The test procedure would be to setup the KG3 to call CQ, and look at the 
SNR reported by skim server.  Increase input amplitude, repeat.

With this configuration, the results were that the onset of 10m spot 
generation occurred at the point that skim server reported 60 dB SNR 
levels.  The point was pretty crisp, in that I could get 58-59 dB SNR 
consistently, but once I stepped up to the 60 dB SNR, I would start to 
see 10m spot levels in the low-single-digit SNR, along with 20m spots.  
The 20m spots would sometimes show a correct frequency and 60-61 dB SNR, 
but more often than not would show significant errors in both frequency 
and amplitude reported. In about 1/3-1/2 of the cases no spot was reported.

Change in BW (48/96/192) or threads allocated (1-32) did not affect the 
results.  Other bands were not checked (I don't have a another set of 
hammers-of-hell LPF to kill the 2nd harmonic of the KG3) and it may be 
easier to figure out how to gate one of the HP sig gens to allow the 
sending of code than to address the filtering - a project for another day.

If this test is an accurate reflection of actual use conditions, it 
suggests that in the 20m case, the skimmer system's limits are these:

Overload point    -40 + 28 = -69 dBm    (~S9+8)
MDS / 2dB SNR:   - 126 dBm

Skim server is pretty responsive with spots down to the low single 
digits but once you get to about 4dB SNR, they start coming slower.  I 
saw a few 2dB spots, one 1dB spot and zero 0 dB SNR reports.  Hence the 
MDS indicated at the 2 dB SNR level.

Points measured have roughly a +/-1 dB initiated tolerance - plus 
instrumentation error.  Which means that these absolute levels may be 
off by a couple of dB, but the relative DR of the system will be closer.

With my normal distribution feed, the peak 20m signal levels normally 
encountered by the skimmer antenna results in a roughly -45 dBm peak 
level.  This is about 20+ dBm too hot, in my case.

The normal caveats apply here.  This is a one-off test, of a novel test 
method, which was completed at 3AM.  So I would love to see if another 
guy who may have a suitable set setup could run a similar test to see if 
the results could be duplicated.

73/jeff/ac0c
alpha-charlie-zero-charlie
www.ac0c.com


On 6/15/22 9:39 PM, Jeff AC0C wrote:
> 73/jeff/ac0c
> alpha-charlie-zero-charlie
> www.ac0c.com
>
>
> For those who have the capability to measure, I'm curious as to what 
> signal levels with the RP resulted in false spots with CW skimmer server.
>
> I've run the QS1R for years and found the overload level to be around 
> the +6 dBm level.  As long as peak signal levels were below that, I 
> had virtually zero false spots (here, I mean as an example, the low 
> SNR 10m spot at the 2x harmonic frequency of 20m actual frequencies).
>
> The same antenna configuration feeding a pair of RP 122-16 look like 
> they are exhibiting overload conditions well under -40 dBm; perhaps as 
> low as -50 dBm.
>
> Initial testing of the RP vs. QS1R showed the RP to be a few dB more 
> sensitive than the QS1R.  However, given similar hardware typologies, 
> I'm stunned at the apparently low level signals required to cause the 
> harmonic phantom spots.
>
> Wanted to see if some of you other RP ops had conducted any 
> measurements on your unit, and what your experience has been.
>
> 73/jeff/ac0c
> www.ac0c.com
>
>
> 73/jeff/ac0c
> alpha-charlie-zero-charlie
> www.ac0c.com
>
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Groups.io Links: You receive all messages sent to this group.
> View/Reply Online (#6178): https://groups.io/g/RBN-OPS/message/6178
> Mute This Topic: https://groups.io/mt/91791597/383159
> Group Owner: RBN-OPS+owner at groups.io
> Unsubscribe: https://groups.io/g/RBN-OPS/unsub [keepwalking188 at ac0c.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>
>