Strange waterfall on VHF satellites; UHF is fine

Our station has two M2 yagi antennas, one for 2 meters and one for 70 cm. The 70 cm reception is great - the waterfalls are strong and the there are many decodes. However, on the VHF frequencies we almost always have a blank waterfall, even on direct overhead passes of the NOAA satellites which should be solid copy. However, one recent pass (SatNOGS Network - Observation 7908692) seems to show a solid signal at the right side of the waterfall with nothing in the middle. There doesn’t seem to be any RF-related reason that this would occur, which suggests possibly a program issue - except that it doesn’t occur on 70 cm passes. Any thoughts on this issue?


Jon WB2MNF sysop for W2MMD

Hi Jon,
I’m running double antennas at station 1864 and it works pretty well.
The only odd thing I see in the configuration is SATNOGS_RX_BANDWIDTH is set to 2MHz, not sure what this does on a rtlsdr, probably nothing but still should probably not be there.

What does the rest of the setup look like ? preamps, cabling, filters, diplexer ?

The rtldsr is a 8bit device, it is pretty easy to overwhelm and you basically get no signal out of it. It can sometimes be seen in the DC portion as a huge spike that just dominates. In this case it is out of view due to the lo-offset.

What I’m suggesting is if the VHF signals (out of band) are very strong and the gain on the rtlsdr is high (which 32.8 is), you can get flat waterfalls.
Even getting in strong broadcast FM can deteriorate reception significantly.
Checking the spectrum with gqrx, changing the gain, using a fm-notch…

Thanks - I think part of the issue might be extremely strong signals on UHF which comes in on the same feedline. We have a 42 element M2 yagi with a preamp up there (also used for Greencube communications) so it might be clobbering the VHF signals. I’ll try shutting off the preamp to see if it makes a difference. I also brought the gain down to 16.5 which didn’t help. Any suggestions on setting RX bandwidth?


Not on rtlsdr, I leave it unset. iirc it’s at least 3MHz wide and fixed, as it’s built for TV reception. The bandscan often show imaging at the edges of a 2MHz wide run. If you have a websdr/gqrx and tune to a strong station at the edge you can see it appear on the opposite side as well. fwiw I think that setting is just ignored.

Deactivate UHF preamp for VHF passes, use FM notch. In my experience, the rtl has a harder time when strong signals are closer to the tuned frequency than if it is waay off. that is, 108MHz is a lot closer to 137/145 than 435 is.

Thanks - my immediate option was to deactivate the 2 meter reception and stick with 70 cm which appears to be where most of the action is. Thinking about replacing the RTL-SDR with an Airspy but installing the software on a Pi doesn’t appear straightforward - I tried a few of the steps and got errors so apparently there’s required knowledge that I’m missing.

when building soapy stuff you should make sure the appropriate -dev packages are installed on the system. starting to build everything from source and installing on the host just makes a mess of everything.
I don’t deal with airspy, but we do have some packages that we maintain over at obs. I know @PE0SAT has more experience with airspy.