...
 
Commits (2)
......@@ -43,6 +43,7 @@
## Build tool directories for auxiliary files
# latexrun
latex.out/
doc/_build
## Auxiliary and intermediate files from other packages:
# algorithms
......
......@@ -135,11 +135,10 @@ In the below figure you can see the morse signal detected in wu 102 on -4,7 kHz.
.. figure:: /graphics/ON0EME_Morse.png
Seti at Home analysis
---------------------
After splitting the 2.5MHz raw data into 256 workunits, these workunits are
processed by seti_boinc or the setiathome_v8 program provided by Seti at Home.
Also the seti grafics is used to visualize the processed workunits.
......@@ -160,6 +159,61 @@ same signal but attenuated by 20 dB.
.. figure:: /graphics/ON0EME_Signal.png
This figure depicts the ON0EME signal in frequency and time.
Checking frequency mirroring in workunit
----------------------------------------
During testing so doubts raised due to frequency mirroring in workunits.
Frequency mirroring can be caused by means of swapping IQ bits in the pipe line.
To test the frequency mirroring the following test has been carried out. Using
a frequency sweep generator the frequency was sweeped from the central frequency
to fc + 1MHz in 100 s. This equates to 10kHz per second, i.e. the signals sweeps
in 1 second from the low frequency to the high frequency in the targeted
workunits, which are 0 to 100.
SETI-playback is used to depict the recorded signal, see figure below:
.. figure:: /graphics/waterfall.png
Recorded Broadband signal using new DT backend
Next to the main signal this figure depicts cross product artifacts, which
complicates the investigation on frequency mirroring. This signal is splited
using splitter_pfb into 256 workunits. All 256 workunits were processed by
SaH software. From each result.sah file the spikes were extracted. In below
figure you can see the same artifats as in previous figure.
.. figure:: /graphics/spikes.png
Extracted spikes from SaH result file.
Workunit having more than 8 spikes are selected and plotted with gnuplot to
check the frequency as function of time. The overall idea is that the frequency
is increasing with time as expected, but due to the fast traversal of frequency
in 10kHz/s, it is advised to make an even slower frequency scan.
results
+++++++
.. figure:: /graphics/wu013_spikes.png
Workunit 13
.. figure:: /graphics/wu054_spikes.png
Workunit 54
.. figure:: /graphics/wu110_spikes.png
Workunit 110
.. figure:: /graphics/wu119_spikes.png
Workunit 119
.. figure:: /graphics/wu128_spikes.png
Workunit 128
.. note::
Frequency axis is not well define. It is number divided by 10 in kHz
Conlusion
---------
......
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.