I am still trying to figure out why bit-perfect data has been
unattainable by me so far in RFNoC (with a custom module). I took a
step back and made a simple "splitter" module that takes in one input
and spits out two identical outputs (screenshot attached). I had the
two outputs go through RFNoC: FIFOs, and then through RFNoC:Keep 1 in N
(with an N of 1) and then to two file sinks. I used a source of a
UHD:USRP Source. When I let it run for a few seconds and then check the
output files, they are bit-perfect to each other (I get the same results
with the RFNoC:Radio as the source).
Next, with the exact same setup, I drop a file sink straight off of the
UHD:USRP Source to see what the incoming data looks like. I run it for
a few seconds and check my splitter outputs and they look the same as
each other. I then compare the splitter outputs with the incoming
output (the file sink connected directly to the source), and they don't
quite match up (this is the issue I've been debugging for weeks).
As an example, I see this on my inputted file sink:
48 21 47 3D 61 01 56 3D F6 20 95 3D 3F A1 41 3D FB 40 98 3D 1B C1 AB 3C
6F E1 5E 3D 72 41 E0 3C 4D 21 4A 3D 70 61 5F 3D FB 70 98 3D 12 81 26 3D
FC D0 98 3D 86 81 EC 3C 19 A1 AA 3D 3B 61 3F 3D
And this on the splitter's two outputs:
8E 21 47 3D AC 01 56 3D 2A 21 95 3D 83 A1 41 3D 30 41 98 3D 58 C1 AB 3C
BE E1 5E 3D C0 41 E0 3C 94 21 4A 3D BF 61 5F 3D 31 71 98 3D 4D 81 26 3D
32 D1 98 3D D9 81 EC 3C 55 A1 AA 3D 7F 61 3F 3D
As you can see, the two outputs are right sometimes, and then slightly
off other times. How am I losing bit-perfect sinks? I dug through the
log files of the build and all timings were met, there were no critical
errors beyond the handful that are there from the original design.
Nothing. I have been tracking down variations of this issue for about
1.5 months now, but I don't understand what I could be missing...
Has anyone tested for bit-perfect transmissions through RFNoC before (if
it was a real signal, it might be close enough to
decode/demodulate/hear/etc.)?
I am still trying to figure out why bit-perfect data has been
unattainable by me so far in RFNoC (with a custom module). I took a
step back and made a simple "splitter" module that takes in one input
and spits out two identical outputs (screenshot attached). I had the
two outputs go through RFNoC: FIFOs, and then through RFNoC:Keep 1 in N
(with an N of 1) and then to two file sinks. I used a source of a
UHD:USRP Source. When I let it run for a few seconds and then check the
output files, they are bit-perfect to each other (I get the same results
with the RFNoC:Radio as the source).
Next, with the exact same setup, I drop a file sink straight off of the
UHD:USRP Source to see what the incoming data looks like. I run it for
a few seconds and check my splitter outputs and they look the same as
each other. I then compare the splitter outputs with the incoming
output (the file sink connected directly to the source), and they don't
quite match up (this is the issue I've been debugging for weeks).
As an example, I see this on my inputted file sink:
48 21 47 3D 61 01 56 3D F6 20 95 3D 3F A1 41 3D FB 40 98 3D 1B C1 AB 3C
6F E1 5E 3D 72 41 E0 3C 4D 21 4A 3D 70 61 5F 3D FB 70 98 3D 12 81 26 3D
FC D0 98 3D 86 81 EC 3C 19 A1 AA 3D 3B 61 3F 3D
And this on the splitter's two outputs:
8E 21 47 3D AC 01 56 3D 2A 21 95 3D 83 A1 41 3D 30 41 98 3D 58 C1 AB 3C
BE E1 5E 3D C0 41 E0 3C 94 21 4A 3D BF 61 5F 3D 31 71 98 3D 4D 81 26 3D
32 D1 98 3D D9 81 EC 3C 55 A1 AA 3D 7F 61 3F 3D
As you can see, the two outputs are right sometimes, and then slightly
off other times. How am I losing bit-perfect sinks? I dug through the
log files of the build and all timings were met, there were no critical
errors beyond the handful that are there from the original design.
Nothing. I have been tracking down variations of this issue for about
1.5 months now, but I don't understand what I could be missing...
Has anyone tested for bit-perfect transmissions through RFNoC before (if
it was a real signal, it might be close enough to
decode/demodulate/hear/etc.)?
