QSB on the beacon signal?

  • Hello. I am a total newcomer to microwaves and QO100. I started experimenting with the RX part of the system before I try to transmit. At the moment I use a small dish (60 cm) behind a window. The beacon is about 12 dB over noise, so I am quite happy. I am waiting for a 110 com dish, and things will be much better.


    The big question I have is that I notice quite some QSB on the beacon signal, with as much as 6dB range and a pattern not unlike what you could see on HF. I have clear line of sight to the satellite, the dish is indoor, and I really do not understand how that could happen.


    Any wisdom from the experts?


    Thanks in advance and 73.


    Pete MM0TWX

  • On 10GHz heavy rain clouds could cause this. I can see it on my equipment as well. I live in a valley and the mountain in the south is 1245m over NN. When there is rain on the summit it also could happen that my Astra TV has no signal.

    • Official Post

    The beacon is about 12 dB over noise, so I am quite happy

    That's quite weak, even for a 60cm dish.. perhaps you are on a side-lobe or the window is metalized and has attenuation. Make sure you have AZ/EL optimized in several iterations, also don't forget the tilt/skew of the LNB.

    6dB variation seems a lot.. Unless there is heavy rain or big clouds, I see almost have no variation.. the CW beacon is 38dB over the noise (in SDR console) here on a 80cm dish..

  • Thank you all for your inputs. Concerning the low signal, I believe it's partly due to double glazed widow, and possibly interference with the lobes from the window's metal frame. For me it was a success to hear anything at all!


    Concerning fading, clouds and rain could be an explanation, for we have a lot of that in Glasgow...

  • That's quite weak, even for a 60cm dish.. perhaps you are on a side-lobe or the window is metalized and has attenuation. Make sure you have AZ/EL optimized in several iterations, also don't forget the tilt/skew of the LNB.

    6dB variation seems a lot.. Unless there is heavy rain or big clouds, I see almost have no variation.. the CW beacon is 38dB over the noise (in SDR console) here on a 80cm dish..

    How are you measuring that 38dB Peter? I don't see that on my 90cm but I do see the transponder noise floor so I must be doing something wrong with SDR Console.


    73 Chris

  • Hello Peter,


    I was running a much older version of SDR Console (3.0.9) and just comparing the power in the beacon and with the same bandwidth of noise floor next to it for a rough ratio. I've updated to 3.0.18 and using the SNR function get the same 38dB as you.


    73s Chris

  • Hi, i never cared too much for the qsb on my rx-system, but now that it is a topic here, i have some questions.

    My setup is a 1.5m pf dish , poty with lnb . In sdr-console, i see the noisefloor at about -110dBm, the cw-beacon reaches -74dBm (with some qsb) on the scale when it transmits a steady carrier. That makes it 36dB SNR, right? The sdr-console meter shows 40.5dB SNR . What causes this discrepancy?

    Also, i see fast qsb on the PSK-beacon. The level changes about 4-6dB within less than 2 seconds. Is there an explanation for that?.

    Third, the noise jumps up 1 or 2 dB every now and then, but i think this has been discussed somewhere else in this forum and is something everyone sees, not just me.

    Maybe the qsb is the reason why decoding FreeDV fails (not completely, but is no fun)?

    73 Martin

  • dm4im:


    be carefull with the SNR display of the SDR Console, it depends on how you have set the visible bandwidth.

    With a smaller bandwidth the SNR value increases and with larger bandwidth it decreases.

    I have set the VB that the CW beacon is on the left end and the PSK beacon is on the right of the spectrum/waterfall. (ca 260kHz BW)

    That gives a reading of 39-41dB SNR (depending on the weather) of the CW beacon on my 120cm dish.

    The transponder floor already shows a SNR of 6-8dB without any signal.

    So the real SNR is around 32-33dB. This was already discussed some time ago in this Forum and should be the correct value.

    I have no QSB on the beacon and those noise jumps are likely from a "friendly radar" (as peter called it someday).

    FreeDV runs here 100% with Mode2020, just checked it yesterday in a hour-long QSO with DJ4RAM.


    73s DB8TF

  • Well, not here. I decreased bandwith from 2048 to 300. I see the noisefloor roll off abt. 3 dB starting 15khz above cw bcn and 15khz below upper bcn when i set the beacons to the left and the right margin. SNR on the cw bcn drops to 37.something .Thus SNR of the cw beacon is 3dB lower compared to 2048 bandwith.


    But this is only a function of levels. What bugs more is the qsb. So after almost a year i realize i still have work to do to improve rx.

  • To clarify what i mean with the changing SNR value with different bandwidths i took some screenshots.


    The SDR-BW is always 2,4MHz! Nothing changed except for the Spectrum/Waterfall zoom!


    Shown SNR of the CW beacon:


    BW 400kHz > 39dB

    BW 100kHz > 41dB

    BW 20kHz > 45dB

    BW 5kHz > 51dB


    73s DB8TF

  • Ok, to sort that out: The Receive Bandwith is always 2,4Mhz (or the max. your dongle provides, in my case 2048khz) , your Filterbandwith is always 2.8khz. You only change the Zoom level.

    So what's happening here? sdr-console does not measure the signal alone, rather it does an interpolation of the levels inside the passband?

  • If you reduce the bandwidth the noise-floor rises too. If I reduce the bandwidth to 10KHz as you did in your screenshot I get an SNR reading of ca. 45, but the noise floor (without any signal present) has increased to ca. 15 which gives me a "real" SNR of ca. 30 which is about the same as with a bandwidth setting of 260 KHz, which is my standard setting.

  • Just for information, I tried the 60 cm dish outside, on a clear day. SDR Console's snr meter now says 24 dB on the CW beacon. However, copy on SSB signals is possible but not very comfortable. At my latitude, 60 cm is evidently not enough.