Dear Friends, We are building a system with our high school team.
We used SDR# with an RTL-SDR dongle.
We used two antennae:
1 - The V dipole that came with the dongle kit.
2 - We built this QFH antenna. (ranch.com/how-to-build-a-qfh-quadrifilar-helix-antenna-to-download-images-from-weather-satellites/)
3 - We also used wideband LNAs for both.
For the most part, we followed advice that several of you generously gave to my July post.
We have not yet been able to convert data to image. We had a satellite pass directly overhead last Friday. The QFH had better signal than the V-dipole. We processed the data to the correct sampling rate in Audacity. But we think the signal strength might be too weak for WXtoIMG to convert to image.
What would the minimum signal strength in dB be to successfully convert an image?
Are there ways to increase signal strength?
Or perhaps we have a different problem somewhere in the chain…?
I think you are in the good conditions to succeed (signal strength should be enough).
The main trick with WXtoIMG (at least with my version under Linux) is to convert the wav audio file sampling at 11025Hz. I have never succeded to make WXtoIMG work at any other frequency however the command line suggests it could. Other arguments to WXtoIMG are, I believe, more intuitive.
You can also search the Libre Space site with the WXtoIMG keyword as there is other references to it.
Yes, @blueskies ; that’s the one.
We got a 36’ coaxial to the antenna mount on the roof. Sealed it in white flex seal so it wouldn’t get stolen for the copper! An LNA in a little outlet box worked really well.
You can do adjustments during the reception if you use SDR#. However, with this hardware, you will probably end near the maximum gain.
QFH is a good choice for a static antenna.
This Thursday I am going to have the students do the observations “mobile” – outdoors with laptops, one team out in an open area with the dipole antenna, one of the roof with the QFH. We’ll see how it goes! If that doesn’t work, I would be fairly certain we have a software problem.
@Frederick If you’re using SDR#, you should be able to see the waterfall and the spectrum in real time (live), so you should get an idea of whether your system is picking up any signals from the satellite(s).
You can check if the system is deaf by tuning to e.g. FM stations and see if you’re picking any signals up (i.e. debug if the signal travels through the antenna, LNA and SDR as it should). To check if the LNA works, you can plug it in/out of the antenna (an LNA should generally be placed immediately after the antenna). If the overall noise increases, then the LNA is working, as it is indeed amplifying input signals and internal noise that it unintentionally produces.
In SDR#, I would make sure to click the Gear button at the top left corner and ensure the options “Offset Tuning”, “RTL AGC” and “Tuner AGC” are all unchecked, and “Frequency correction (ppm)” is set to 0, like so:
Your RF gain should not be set to 0 dB (otherwise the RTL would be pretty deaf), but it should not be maximized either (e.g. to avoid over-saturation/clipping by the ADC (if your environment is RFI-rich)). Ideally, you should start from RF Gain = 0, and slowly increase the gain until you see that the signal-to-noise ratio (S/N) in the live spectrum maximizes. At some point, increasing the RF Gain will no longer improve the S/N. Usually for an RTL, the nominal gain should be around 8 dB, but I think it’s safer for you to set it to 12 dB to ensure the RTL will pick the signal up.
Here’s how it went. One group of students went out in a field with a laptop and the dipole antenna, the other group went on the roof and hooked directly into the QFH helical antenna (bypassing the LNA). Our results:
The dipole (left) received a partial image: The ocean and clouds are visible. (We live in Hawaii.) The helical antenna received nothing but fuzz.
So we probably could tweak the software (SDR#) for the dipole. I think it should have done better. But per the helical, probably it’s a hardware problem: it’s very likely we made some mistake when building it. The helical should have received a better image.
