First allow me to say that I love my F5K. I would not trade it for any other
radio. And I fully appreciate all the hard work Gerald and all the team have
put into making this technology available! So, please accept this as just my
observations after doing a little testing.
The CW transmission performance of my F5K has always disappointed me. So, my
Flex spends most of its time on the digital modes. And it does a great jobs
in those modes. After the latest PSDR updates I hoped the CW situation would
improve. My first impression was that, indeed, it did improve a lot. But I
still could not send good CW. I just could not get the timing down. At first
I thought it was just a matter of getting accustomed to a new rig. Try as I
might I just can't send CW with the F5K. When I try to send "test" it comes
out "tiht" more often than not. So, this weekend I tried a little
experiment.
My normal station setup is as follows:
F5K
PSDR V2.0.22
One of Neal's i7 computers running W7 32 bit (Runs at < 10% CPU usage w/HRD
- very low DPCs)
MFJ Keyer with Kent iambic paddle
Radio Interface - Audio->Primary Buffer Size:2048 Sample Rate:96000
I brought my Tektronix TDS2022 'scope home and measured the time delay
between key closure and the start of the sidetone. The test was conducted
using the MFJ keyer connected to the key jack and iambic set to false . The
keyer was set to produce a string of dots at 20 WPM (60 ms). One channel of
the scope was connected to the front panel headphone jack and the other
channel was connected to the keying line from the keyer. The scope was
synced on the keying line. Drive was set to zero and the radio was connected
to a dummy load.
Radio Interface - Size: 2048 Sample Rate: 96000 (my normal setup)
DSP Buffer Size CW RX:2048 CW TX:256 Delay
= 54 ms (Almost a full dit!)
DSP Buffer Size CW RX:2048 CW TX:512 Delay
= 54 ms
DSP Buffer Size CW RX:2048 CW TX:1024 Delay =
54 ms
DSP Buffer Size CW RX:2048 CW TX:2048 Delay = 54 ms
Not what I expected. It seems the DSP TX setting for CW makes no difference
in the delay. Neither does the DSP RX CW setting but this was expected. I
have been told that these settings have an effect on the delay. But in my
case they do not.
Here are some numbers for other buffer sizes and sample rates:
Radio Interface - Size: 2048 Sample Rate 19200 Delay =
35.6 ms
Radio Interface - Size: 1024 Sample Rate 19200 Delay =
30.8 ms
Radio Interface - Size: 512 Sample Rate 19200 Delay =
28.0 ms
Radio Interface - Size: 256 Sample Rate 19200 Delay =
28.0 ms
Radio Interface - Size: 1024 Sample Rate 9600 Delay =
36.0 ms
Radio Interface - Size: 512 Sample Rate 9600 Delay =
32.5 ms
Radio Interface - Size: 256 Sample Rate 9600 Delay =
28.0 ms
I continued the experiment with various combinations of buffer size and
sample rates. In every combination that I tried, the delay was still a
significant part of the length of a dot.
I went through these tests several times. These numbers varied a bit over
the three days that I conducted the tests. For instance, the 2048/9600
numbers varied from 48 ms to 54 ms.
Next I tried connecting the key directly to the Flex and setting iambic to
true. The shortest delay I measured was 28 ms.
Then I repeated the experiment with my old, hasn't been used since the Flex,
FT-1000. The delay was around 4 ms.
My conclusion is that I can't adjust my finger-ear coordination to handle
delays present in the Flex.
Now it did occur to me that the problem might be in my head so I tried one
more experiment. I asked a ham friend (CW op) to stop by. I sat him down at
the key, had him put the headphones on and send a little. He immediately
noticed the problem. He was having the same problem I was.
The only two ways I have found that allow me to send proper CW with the Flex
is to either use the keyboard or listen to the sidetone from the keyer.
I have been told by other Flex owners that they routinely operate CW at
speeds up to 30 WPM. I find it difficult to believe that any good CW op can
plug a paddle into the Flex and send good CW while listening to the
sidetone. Am I doing something wrong?
73,
Mark K5XH
_______________________________________________
Flexedge mailing list
[email protected]
http://mail.flex-radio.biz/mailman/listinfo/flexedge_flex-radio.biz
This is the FlexRadio Systems e-mail Reflector called FlexEdge. It is used for
posting topics related to SDR software development and experimentalist who are
using beta versions of the software.
