[aprssig] Interesting Preliminary Findings on Soundcard "Soft TNC" Shootout

[Stephen H. Smith]

On 12/26/2011 11:42 AM, Lynn W. Deffenbaugh (Mr) wrote:
> On 12/26/2011 11:32 AM, Stephen H. Smith wrote:
>>
>> I.e. the "cliff" between "always" and "mostly failing" is surprisingly 
>> abrupt, not at all like the perceived "gradually getting noisier" one 
>> experiences on voice as the signal weakens.
>
> Not surprising to me if you think about it.   Drop one bit in an AX.25 
> packet, and the entire packet is not decoded.  There's no such thing as 
> "almost" with AX.25 decoding.  And that's not a fault of AX.25 exclusively.  
> Any digital mode that uses checksums or CRCs or other error detection 
> mechanisms without any form of error correction will suffer the same fate.

I realize that, but before I did the test,  I had been expecting that one would 
gradually get an increasing failure rate as the RF level dropped; something like:

Level0               100%
Level0 - 1dB      80%
Level0 - 2 dB     30%
Level0 - 3 dB     10%


The critical "all or nothing"  threshold appears to be when the soft hiss 
partial-quieting noise that is many (audio) dB below the recovered  modulating 
audio tones starts giving way to the "popcorn" noise.

Now that I think about it, this makes sense. The popcorn represents spikes that 
can be at the same level (or higher) than the modulating audio tones.       
Since even just ONE of these pops at the right instant can add a zero crossing 
that turns a 0 into a 1 (trashing the whole packet), the rather abrupt "falling 
off a cliff" effect occurs when even just a few pops start happening.

(By the way, I did the tests with the signal generator modulated to exactly 3.0 
KHz, as measured by the IFR's own deviation meter, on both tones.)

The IF bandwidth (i.e. selectivity) of the receiver has a significant effect on 
this.    Wider-bandwidth receivers tend to more gradually loose quieting, 
getting much  "hissier" as RF level drops, before the popping starts, than 
narrowband ones.     I would expect to see a more gradual slope of increasing 
failure rates as the RF level drops in wideband receivers.

I guess this adds another dimension to the more comprehensive tests to come -- 
I will characterize the bandwidth of each receiver before running tests on it.

-------------------------------------------------------------------------------

--

Stephen H. Smith    wa8lmf (at) aol.com
Skype:        WA8LMF
Home Page:          http://wa8lmf.net

*****  NEW   Precision Mapping 9 For UIview Released! *****
    http://wa8lmf.net/aprs/PMap9_Notes.htm


  Vista & Win7 Install Issues for UI-View and Precision Mapping
     http://wa8lmf.net/aprs/UIview_Notes.htm#VistaWin7

30-meter HF APRS over PSK63
    http://wa8lmf.net/APRS_PSK63/index.htm

"APRS 101"  Explanation of APRS Path Selection & Digipeating
   http://wa8lmf.net/DigiPaths