Re: LF: Re: LF EbNaut tonite - interference rejection in SpecLab

[Markus Vester <markusvester@aol.com>]

Assuming a linear receiver with no AGC action, the major mechanism for desense is usually the interaction of a strong inband signal with a (software) noiseblanker. On one hand it can raise the NB trigger threshold such that the more spherics come through. More importantly, the NB chops up the strong signal itself, so it encompasses itself with noise-like AM sidebands which brighten up the spectrogram around it.

Fortunately SpecLab has some clever FFT-filter options to deal with a strong narrowband interferer. For example, you can select "custom filter" and graphically draw a notch around a known interferer frequency. Then there's a versatile autonotch function, albeit when setting this up you will have to beware not to "throw out the baby with the bathwater". Furthermore there's a very useful frequency selective limiter, which automatically lowers the gain only for bins which contain a signal above a certain threshold (reminiscent of "LEILA" on Oscar satellite treansponders). Although this is not complertely artifact-free, the remaining contamination is usually limited to very few Hz around the offender, and the rest of the spectrum becomes interference-free. This "Leila" is also inserted at DL0AO before the narrowband instances and decoders, and the effect is sometimes visible when one of us is transmitting. 

The FFT filter options are well documented in SpecLab's help files. Note that for the desired effect, the noiseblanker must be inserted only after the FFT filter, using the second "blackbox" (L4-L5). 

Another source of interference are keying sidebands from the transmitter itself, which unfortunately cannot be mitigated on the receive side. In that respect, phase-continuous modes like WSPR or MSK are relatively benign. On the other hand, hard-keyed modes like EbNaut, Opera (and particularly endless aural-speed Morse beacons) can be quite obnoxious in the neighbourhood of a strong transmitter.

Best regards,
Markus (DF6NM)

-----Ursprüngliche Mitteilung-----
Von: N1BUG <paul@n1bug.com>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Mo, 12. Nov. 2018 22:04
Betreff: Re: LF: Re: LF EbNaut tonite

Hi Stefan,

OK but keep it in mind for the future. I understand we have only 2.1
kHz to play in and we have to share the resource. The DL0AO decodes
were 0544 and 0550z so I think you were not transmitting at the time.

73,
Paul

On 11/12/18 3:36 PM, DK7FC wrote:
> Hi Paul,
>
> No problem, don't worry about the QRM. It is normal that we interfere
> each other sometimes. I'm sure i also reduced your WSPR SNR at DL0AO if
> you were so close to me ;-)
> Tonite i'll stay at 137.620 kHz.
>
> 73, Stefan
>
> Am 12.11.2018 21:23, schrieb N1BUG:
>> Stefan, Markus,
>>
>> If you want to stay on 137.465 let me know... I can shut my
>> transmitter off for the night. Sorry I didn't think to do that last
>> night. I thought I was far enough away (in kilometers) from anyone
>> with interest in EbNaut...
>>
>> Paul N1BUG