Testing per-sample IQ Doppler correction in SpectraFlux. Satellite tracking software sends frequency updates via rigctld — instead of retuning the SDR (which causes PLL transients and audio glitches), the correction is applied as a phase-continuous NCO rotation on every IQ sample. SDR parks at a clean center frequency, Doppler offset handled entirely in the digital domain. Zero retune latency, no phase discontinuities. Here's a Python script driving a ±50 kHz sinusoidal S-curve through rigctld while the spectrum tracks it smoothly.
SpectraFluxのIQドップラー補正をテスト中。 衛星追尾ソフトからrigctldで周波数更新を受信し、SDRの再チューニングではなく、IQサンプルごとにNCO位相回転で補正を適用。PLLの過渡応答やオーディオの途切れがゼロ。Pythonスクリプトで±50kHzの正弦波ドップラーカーブをシミュレーションしています。
@BoscoMac Thanks, only HP/Agilent/Keysight gear in the lab! 🤓🤓Yep adjustable height, it are OK to test other equipment but I dont like for small DUTs with a lot of cables, is better a large table
I set up the first devices in the new laboratory. The electronics laboratory is now fully air-conditioned. Stable temperatures for calibration. Soon the work tables will be set up. #hp#agilent#keysight#tektronix#rohdeubdschwarz
@BoscoMac Great, send some a picture pelase when you finish your setuo,always is nice get inspiration for others. I’musing small tables with wheels to move in front of every rack, but this tables are too small for some DUT’s. But for the new lab I’m thinging in more bigger fixed tables.
SpectraFlux can now see and decode airplanes. PlutoSDR at 1090 MHz, ADS-B decoder written from scratch in Swift. 70 aircraft from a tiny antenna on my desk. It helps to be on the corridor to KATL.
Got pySpectraVue running way faster. Now it processes the full sample buffer and batches waterfall lines in a single repaint. Huge difference.
Python/PyQt6/PyTorch/NumPy. Torch handles GPU FFTs on Apple MPS and CUDA. Talks to USRP B210 and RFSPACE NetSDR.
🚀 Major progress on PySpectraVue, a Python-based spectrum analyzer optimized for precision! Image shows testing with 512k FFT and various colormap files. Features pixel-perfect rendering (no smearing in time or frequency), exact frequency displays, high-res zooming, and hardware-accelerated FFT via PyTorch (MPS/CUDA). Inspired by the big guys, PySpectraVue builds on 25+ years of SpectraVue SDR legacy. #SDR#SignalProcessing#Python