Date
1 - 2 of 2
FST4W - Is spread really important ? #fst4
Erwin - PE3ES - F4VTQ
Due to a failing GPS subsystem of my KiwiSDR I was looking at some solutions. But also checking the reception and decoding of FST4W signals.
I found the following over the last 10 days. This includes the moment my GPS started to fail (roughly 7 days ago).
I have no GPS at the moment and no other reference signal to discipline my KiwiSDR.
Results are across 20-30-40-80 meter bands.
I receive my local own RFZero generated signal with a spread of 4-140 mHz
I receive RX3DHR with a spread of 12-489 mHz
I receive OZ0RF with a spread of 7-871 mHz
I do not see a statistically relevant change in spread between old decodes and new decodes : conclusion >> the KiwiSDr does not need a GPS reference to be able to decode FST4W
I do see the software being able to decode within a wide range of spread.
I do realise this might not be true for very critical situations and it would always be good practice to have a well disciplined KiwiSDR.
Your thoughts and opinions on this are appreciated !
Erwin
I found the following over the last 10 days. This includes the moment my GPS started to fail (roughly 7 days ago).
I have no GPS at the moment and no other reference signal to discipline my KiwiSDR.
Results are across 20-30-40-80 meter bands.
I receive my local own RFZero generated signal with a spread of 4-140 mHz
I receive RX3DHR with a spread of 12-489 mHz
I receive OZ0RF with a spread of 7-871 mHz
I do not see a statistically relevant change in spread between old decodes and new decodes : conclusion >> the KiwiSDr does not need a GPS reference to be able to decode FST4W
I do see the software being able to decode within a wide range of spread.
I do realise this might not be true for very critical situations and it would always be good practice to have a well disciplined KiwiSDR.
Your thoughts and opinions on this are appreciated !
Erwin
Erwin
My work with FST4W shows that spread is important, but it has to be looked at in terms of probability of decode, certainly for the -120 and -300 variants. As spread is even more important for -900 and -1800 with those variants the probability curves may look more like either decode or no-decode, 100% either way, depending on total spread. See http://wsprdaemon.org/ewExternalFiles/FST4W_on_HF_bands_V1-3.pdf for examples.
Unfortunately, as many other factors affect decode probability there is no quick and certain attribution to excess spreading, but reduced probability of decode due to excess spreading at your Kiwi without GPS may be why you decoded 143 out of 279 FST4W-300 from OZ0RF on 20 m that I decoded in the UK with a Kiwi and mini-Bodnar over the last seven days. It was 97 out of 245 the previous week.
Perhaps more importantly without GPS your reported frequencies are not right. The graph below shows the difference between the frequencies of spots on 20 m between you and G3ZIL with mini-Bodnar (the ones at +/-100 Hz from the central values are due to spots with spurs).
Gwyn G3ZIL
Available using usual login credentials at:
http://logs1.wsprdaemon.org:3000/d/RDnDSpZVk/wspr-frequency-comparisons?orgId=1&var-rx_id_A=G3ZIL&var-receiver_A=G3ZIL_4&var-rx_id_B=F4VTQ&var-receiver_B=KIWI_0&var-band_A=20&var-band_B=20&from=now-14d&to=now
My work with FST4W shows that spread is important, but it has to be looked at in terms of probability of decode, certainly for the -120 and -300 variants. As spread is even more important for -900 and -1800 with those variants the probability curves may look more like either decode or no-decode, 100% either way, depending on total spread. See http://wsprdaemon.org/ewExternalFiles/FST4W_on_HF_bands_V1-3.pdf for examples.
Unfortunately, as many other factors affect decode probability there is no quick and certain attribution to excess spreading, but reduced probability of decode due to excess spreading at your Kiwi without GPS may be why you decoded 143 out of 279 FST4W-300 from OZ0RF on 20 m that I decoded in the UK with a Kiwi and mini-Bodnar over the last seven days. It was 97 out of 245 the previous week.
Perhaps more importantly without GPS your reported frequencies are not right. The graph below shows the difference between the frequencies of spots on 20 m between you and G3ZIL with mini-Bodnar (the ones at +/-100 Hz from the central values are due to spots with spurs).
Gwyn G3ZIL
Available using usual login credentials at:
http://logs1.wsprdaemon.org:3000/d/RDnDSpZVk/wspr-frequency-comparisons?orgId=1&var-rx_id_A=G3ZIL&var-receiver_A=G3ZIL_4&var-rx_id_B=F4VTQ&var-receiver_B=KIWI_0&var-band_A=20&var-band_B=20&from=now-14d&to=now