search

Xerbo / LeanHRPT-Demod

An easy to use HRPT demodulator
GNU General Public License v3.0
58 stars 7 forks source link

MetOp Demodulation creates a circle in constellation diagram #10

Closed Crosswalkersam closed 2 years ago

Crosswalkersam commented 2 years ago

Hello, when I try to decode MetOp (live and offline), I get a circle in the constellation diagram and not the 4 dots typical for QPSK. NOAA and METEOR BPSK demodulates fine with 3MSPS, QPSK does not (neither 3MSPS nor 6MSPS). It sporadically goes to 4 dots for 1 or 2 seconds, but then quickly returns to the circle.

Bildschirmfoto 2022-02-21 um 08 47 46

I recorded MetOp @ 6MSPS with an AirSpy mini using Windows 10. The SNR is 9-14dB. In case its useful, I uploaded a .wav baseband here. I used the newest release of LeanHRPT-Demod. Any Ideas on what could be wrong? Thanks!

Xerbo commented 2 years ago

The only thing I can think of is sample drops, I have tried that baseband in other software and got the exact same result.

zeke800 commented 2 years ago

Hmm... whenever I decode MetOp there are four dots. I don't know too much about AirSpy but as Xerbo suggested give it a go at 3 MSPS, or switch to a different USB port. I found that on my computer there are some ports that produce a massive amount of drops.

zeke800 commented 2 years ago

I know this is off-topic but I also never seem to get a signal from FengYun 3B or 3C... any ideas?

Crosswalkersam commented 2 years ago

The only thing I can think of is sample drops, I have tried that baseband in other software and got the exact same result.

Is there any reliable way to check if I am loosing samples? I would use rtl_test, buts since its an Airspy device, this does not work.

I know this is off-topic but I also never seem to get a signal from FengYun 3B or 3C... any ideas?

Yes. They are all offline right now. 3A is gone, 3B is in maintenance until atleast June and 3C is only active over China I believe.

zeke800 commented 2 years ago

Is there any reliable way to check if I am loosing samples? I would use rtl_test, buts since its an Airspy device, this does not work.

Try demodulating Inmarsat or something relatively easy to receive at the same sample rate and see if the constellation locks. That is the only way I can think up off the top of my head.

Xerbo commented 2 years ago

Looking at your NOAA baseband I noticed some weird stuff happening around the carrier, potentially overloading? Try turning down the gain image

Xerbo commented 2 years ago

Closing due to inactivity

zeke800 commented 2 years ago

Actually I had this happen once. Unplugged my SDR and restarted. Gone.