jvierine / chirpsounder2 Goto Github PK

In some configurations of the code, the data-write to disk can be 360GB/hour (e.g. when streaming DC-25MHz raw IQ). Given typical SSD terabyte write (TBW) limits of around 100 - 3000 TBW, we might expect disk failure within a few weeks to a year.

I think the short term fix is a health warning and the long-term fix is an implementation that performs all processing on IQ stored in memory. This will likely require ~64 GB RAM or so (just for the IQ buffer), given the longest chirp is ~200 s and we probably need some buffer around it.

https://www.amazon.com/Inland-Premium-Internal-Solid-State/dp/B07X74M6PT?th=1

the final line of calc_ionograms.py uses variable "d" that is no longer set in the lines following if name == "main":

The make fails because stuff like UHD, Boost, digital_rf is not there unless you install it. Even with pip-installed digital_rf it fails. I suggest adding "sudo apt-get install libhdf5-dev" to the dependencies ahead of building digital_rf from source.

Note this appears to be fixed by installing digital_rf from source code. I noticed the digital_rf workarounds - maybe it can be removed from the makefile?

Hi,I'm using a receiver x300 and ubx with 12.5MHz center frequency, 25MHz sample rate and GPSDO.

thor.py -m 192.168.10.2 -d "A:0" -y "TX/RX" --clock_source 'gpsdo' --time_source 'gpsdo' -c ch1 -f 12.5e6 -r 25e6 -g 0 /dev/shm

I read the h5 data using matlab and found that there is a frequency bias of around 3.053KHz in all the data, why is this? Is it because my parameter setting is wrong?

One more question, how can I know where the ionogram I receive is emitted from?

Thanks!