KD7LMO GPS sample and sample rate nomenclature


#1

Hi,

KD7LMO provides a binary GPS recording, presumably made with the USRP
(thank you!). It is advertised as “4msps / complex GPS Snapshot (500
mS)”. I use these ratios to determine what the file size should be:

500e-3 seconds
4e6 complex_samples / second
2 floats / complex_sample
32 bits / float
1/8 bytes / bit

Multiply them all together and I get 16e6 bytes. However, KD7LMO’s file
is 32e6 bytes large. Can someone explain the discrepancy or
misunderstanding?

Thank you!

Chris


#2

Either I recorded them as 64-bit floats (double) or I did the conversion
from time in seconds to samples incorrectly. The capture script I wrote
is located at:

http://www.kd7lmo.net/ground_gnuradio_software.html

The code calls the file_sink method. I’m not sure what units/types it
generates.

dst = gr.file_sink (gr.sizeof_gr_complex, filename)

head = gr.head(gr.sizeof_gr_complex, sample_count)

fg.connect(src, head, dst)

So the recording may be 1 seconds, instead of 500mS.


#3

removed_email_address@domain.invalid wrote:

Either I recorded them as 64-bit floats (double) or I did the conversion
from time in seconds to samples incorrectly. The capture script I wrote
is located at:

mgray,

Thanks for the response. Your script looks fine to me, but I’ll try
running it. Perhaps you typed one second at the command prompt.

I tried to read the data as 64 bit floats, but it looked strange
(typical values 1e13) and I quickly gave up on that.

It appears that the first 7000 samples are bad. I attached a plot of
the mag of the first 10000. The remainder look like the last chunk on
the plot. I still cannot acquire a signal, but I’m working on that
(could be my fault).

BTW, I am working with gps_4m_complex.21Apr2005.dat

Thanks,

Chris

PS: attachment rejected by gnu.org


#4

Chris S. wrote:

PS: attachment rejected by gnu.org

I put it here:

http://www.stankevitz.com/temp/samples.gif