Is DBPSK Differential?

GNU Radio
1

I am wondering if it is my understanding is wrong here.

In my ongoing saga (I have been trying to get a flowgraph that worked in
3.6 working in 3.7.2 and it just wont) examining the bits out of the
demodulator when fed with 01111110 AX25 flags, the DBPSK module in GR3.6
appears to be differentially coded and always outputs 1000001, but in
3.7 it isn’t consistent, sometimes it is 1000001 and sometimes 01111110?
My understanding of DBPSK is that to send a 1 the phase shift is 180
degrees and for a zero there is no phase shift. This means when a signal
is transmitted there is no need to sync to the phase ambiguity. That
being the case, one should never see what I see in 3.7.2 - it looks more
like the DBPSK demodulator is actually a BPSK demodulator and D means
something else?

I also notice a few more options in 3.7 - Core Affinity and Min Output
Buffer. I don’t think there are connected but would like to know what
they are.

Mike

2

Mike W. wrote in post #1124197:

I am wondering if it is my understanding is wrong here.

In my ongoing saga (I have been trying to get a flowgraph that worked in
3.6 working in 3.7.2 and it just wont) examining the bits out of the
demodulator when fed with 01111110 AX25 flags, the DBPSK module in GR3.6
appears to be differentially coded and always outputs 1000001, but in
3.7 it isn’t consistent, sometimes it is 1000001 and sometimes 01111110?
My understanding of DBPSK is that to send a 1 the phase shift is 180
degrees and for a zero there is no phase shift. This means when a signal
is transmitted there is no need to sync to the phase ambiguity. That
being the case, one should never see what I see in 3.7.2 - it looks more
like the DBPSK demodulator is actually a BPSK demodulator and D means
something else?

I also notice a few more options in 3.7 - Core Affinity and Min Output
Buffer. I don’t think there are connected but would like to know what
they are.

Mike

I was just me using DBPSK then? Can we have this block removed from the
tree - it seems redundant anyway as the PSK block can be configured to
do DBPSK and works as expected.

Mike