Key Clicks?


Lahra Flip Svare KT9X
 

In another forum, they were talking about changing the "rise time" to 6ms (from any faster setting) to prevent key clicks.  They also said many times the "key clicks" that receiving operators complain about are actually caused by their own misuse of the Noise Blanker.  So I've been checking into "rise time".  I don't think I fully understand it yet and how it relates to the possible perceived "key clicks". But the only setting I see on the FTDX101D that might relate is the "CW WAVE SHAPE".  Mine is currently set to 8ms, but it seems that 6ms is the default.  I believe one of our FTDX101 Group Zoom meetings mentioned to change that to 8ms, and that is probably why mine is not set to the default 6ms.  However I also see the option for 4ms.  Perhaps that faster setting might lead to key clicks??  I am thoroughly confused as to what may or may not cause the key clicks, but I'm leaving my "cw wave shape" set to 8ms, pending any replies from this prestigious group that my sway me toward another setting.  Or maybe the whole "rise time" issue isn't addressed by the wave shape at all, if that's the case, someone please set me straight.  I'm still learning this radio.

73, Flip
KT9X


Daniel W9DGI
 

The CW WAVE SHAPE setting changes the rise time of the keying waveform. Leave it at 6ms or 8ms; don't set it at 4ms -- a longer rise time is less likely to result in key clicks. The latest firmware release added the 8ms option to address any possible issues with key clicks.
73
Daniel W9DGI


Lou W7HV
 

Rob Sherwood has tested the bandwidth of the FTDX101MP at 8msec rise time and said it was the best he's seen.  That's for firmware 202104 and later.  In that and subsequent revs, the risetime was increased to more accurately reflect the setting number.  The actual risetime is still a bit faster than the number you set.  I've listened to it on another rig and can't tell the diff between 6 and 8 msec at 40+wpm so leave it at 8.  Here's a pic of the keying waveform at 60wpm and 8 msec.


Barry D. Jacobson
 

I am leaving mine on 4 ms for now. I like crisp, sharp Morse Code, not blurred out. Have heard all kinds of weird sounds over the years. But if I am told it has key clicks or splattering, would consider changing the rise time.

In my opinion, you'd be better off sharing a frequency domain plot of the waveform at different rise times so we can compare how wide the spectrum is, if you can do it. May need to send series of dots and trigger on repetition rate.

Wondering also what type of coupler or attenuator you use to get transceiver output into your scope.

Best,

Barry WA2VIU

--
Barry Jacobson
WA2VIU
bdj@...
@bdj_phd

On Wed, Jun 16, 2021, 4:05 PM Lou W7HV via groups.io <louandzip=yahoo.com@groups.io> wrote:
Rob Sherwood has tested the bandwidth of the FTDX101MP at 8msec rise time and said it was the best he's seen.  That's for firmware 202104 and later.  In that and subsequent revs, the risetime was increased to more accurately reflect the setting number.  The actual risetime is still a bit faster than the number you set.  I've listened to it on another rig and can't tell the diff between 6 and 8 msec at 40+wpm so leave it at 8.  Here's a pic of the keying waveform at 60wpm and 8 msec.


Lou W7HV
 

Yep.  A spectrum of the keying would be good.  Post one up if you have it.  I'll do that if/when I can get one.

In my experience, mushy code is due to changing the weighting so that the space is too short compare to the Morse element.  The "standard" is for a space to be equal to a dit and a dash to be equal to 3 dits.  Some ops will increase the basic weighting to make the space to short, and some like to extend the length of the dash to more than 3 dits.  All of this can be done depending on the keyer.  The FTDX101 CW WEIGHT setting only extends the length of the dah relative to the dit.  The default is the standard value of 3.

Listening again on another rig, at 40+wpm I find the sound at 8msec rise time to somewhat better than the faster rise times where dits produce more of a tap than a tone.  IMO, removing the higher frequency components produces a cleaner more readable and more listenable sound.  Of course that's me listening to my own rig.  Operating, I'm listening to whatever the other station is putting out.

I have a voltage divider at the input to a 50 ohm dummy consisting of 1k and 50 resistor to ground. If the 50 ohm dummy was perfect (it's actually more like 55 ohms!), the rig would see 47.6 ohms or roughly 1.05 SWR as a result of the added voltage divider. The 50 ohm coax from the divider sees the 50 ohm resistor of the voltage divider and is terminated into 50 ohms at the scope.


Lou W7HV
 

I've recently obtained a keying spectra for the FTDX101MP at 30 and 60wpm with the rise-time set to 8msec.  This is with the current 202105 firmware.  It's excellent, perhaps the best among current rigs.  A few posts down you can see the 101MP's keying envelop that I posted.  This is a huge improvement over Yaesu's previous clicky rigs and should be lauded.  Also attached is the spectrum of the FT1000MP MkV at 30wpm.  This is after long ago having done the W8JI clix mod on it.  It's twice as wide (note different frequency span).  


WF3T
 

Looks very good, and so does the keying envelope you posted earlier. Thanks Lou.


Barry D. Jacobson
 

I listened to my 4 ms risetime (haven't upgraded firmware) on a websdr and sounded clear as a bell.

--
Barry Jacobson
WA2VIU
bdj@...
@bdj_phd

On Sun, Jul 11, 2021, 4:53 PM WF3T via groups.io <wf3t=fastmail.com@groups.io> wrote:
Looks very good, and so does the keying envelope you posted earlier. Thanks Lou.


Lou W7HV
 

It'll sound good when you listen to it, but it will create more QRM for those listening close to your frequency.  To me, 4msec risetime sounds too clicky over 30wpm.  8msec sounds great up to 60, though I can't copy that fast.

Mushy sounding CW is due too too heavy or distorted dit vs dah weighting, not envelope rise time. 


John Bayne N4EEB
 
Edited

The option to change the rise time shouldn't even be in the menu. Set your rise time to 8 and forget about it


John Bayne N4EEB
 

4ms should not even be an option on the radio. You're foolish if you don't set it to 6 or 8.


Barry D. Jacobson
 

We should verify it on the air. Glad to make a sched.
--
Barry Jacobson
WA2VIU
bdj@...
@bdj_phd

On Sun, Jul 11, 2021 at 6:32 PM John Bayne N4EEB via groups.io
<john@...> wrote:

4ms should not even be an option on the radio. You're foolish if you don't set it to 6 or 8.


John Bayne N4EEB
 
Edited

There's nothing to verify on the air. Listen to Rob Sherwood's videos on rise times. You possess a radio that can create the cleanest cw signal in the world, and if you want to sabotage that, set yours at 4.  Good luck. 


Barry D. Jacobson
 

Hi John, not a matter of sabotaging it. In my long career in CW, I
have heard all kinds of pathological CW sounds. Mushy, scrapy, buzzy,
etc. What I heard from my signal on the websdr few nights ago left me
with a smile on my face at 4 ms. Was so crisp and sharp. But it's true
I adjusted the SDR receiver to listen at exactly my sending frequency.
I will repeat the test listening off-frequency and see what I hear. I
will also try in near future to simulate in Matlab a plot of the
spectrum of 4 ms vs. 8 ms rise-time. This should be a straight
signal-processing issue that does not even depend on the radio. I have
looked at my monitor scope and it is rising and falling linearly (not
some kind of curved sigmoidal shape), and I believe Lou, W7HV, sent a
picture of the time-domain response which was linear. The question is
what is the spectrum of a trapezoid of certain duration (depending on
CW speed) with two sides rising and falling at 8 vs 4 ms. This can be
computed by hand or simulated on a computer without knowing anything
about the radio. It is well-established Fourier analysis. I hope to
get around to it. But of course, listening is important, as well.
--
Barry Jacobson
WA2VIU
bdj@...
@bdj_phd


On Sun, Jul 11, 2021 at 6:49 PM John Bayne N4EEB via groups.io
<john@...> wrote:

[Edited Message Follows]

There's nothing to verify on the air. Listen to Rob Sherwood's videos on rise times. You possess a radio that can create the cleanest cw signal in the world, and if you want to sabotage that, set yours at 4. Good luck.


Lou W7HV
 

When referring to clix, or clicks, I am talking about spurious keying artifacts heard outside of a reasonable bandwidth from a CW signal.  It interferes with others that may be operating nearby, and certainly in crowded conditions and pileups.   It's unneighborly and inconsiderate to do so, but that is the characteristic of many radios, like my FT1000MP MkV, which I went through the effort to modify and improve.  It's roughly twice as good as it was before the mod, but still twice as wide as what the FTdx101 can do.  Why not transmit the cleanest signal that your radio is capable of if it's only a menu selection away?  There is no downside.


John Bayne N4EEB
 

With all that engineering and C W experience, you ought to try for the extra class ticket.


John Bayne N4EEB
 

Don't waste your breath, he doesn't get it.


Barry D. Jacobson
 

Hi Lou, I understand, but I need to convince myself there is no downside. Will model and listen and make up my mind. Will share any plots I make.

--
Barry Jacobson
WA2VIU
bdj@...
@bdj_phd

On Sun, Jul 11, 2021, 8:40 PM Lou W7HV via groups.io <louandzip=yahoo.com@groups.io> wrote:
When referring to clix, or clicks, I am talking about spurious keying artifacts heard outside of a reasonable bandwidth from a CW signal.  It interferes with others that may be operating nearby, and certainly in crowded conditions and pileups.   It's unneighborly and inconsiderate to do so, but that is the characteristic of many radios, like my FT1000MP MkV, which I went through the effort to modify and improve.  It's roughly twice as good as it was before the mod, but still twice as wide as what the FTdx101 can do.  Why not transmit the cleanest signal that your radio is capable of if it's only a menu selection away?  There is no downside.


Barry D. Jacobson
 

Was away from hobby for a number of years, except for annual Field Day at my club. FD was cancelled last year because of Covid, so I decided need to set up my own shack. Hasn't really had one since high school. Studying for Extra now. I took General at FCC. Real code exam.

--
Barry Jacobson
WA2VIU
bdj@...
@bdj_phd

On Sun, Jul 11, 2021, 8:40 PM John Bayne N4EEB via groups.io <john=n4eeb.com@groups.io> wrote:
With all that engineering and C W experience, you ought to try for the extra class ticket.


WF3T
 

A couple of observations. First of all, Barry, you will not hear your own key clicks so listening to yourself on an SDR or recording is meaningless in the context of this discussion.

Conventional wisdom says you should never set the rise time as low as 4ms. However, these transmitters may be so clean it's not an issue. Or, their 4ms setting may not be 4ms in actuality. Lou can answer that, he already has half the data, just needs to get shots keying at 4ms also. 

Second, 8ms is fine for slow speeds, but doesn't sound good at true QRQ speeds. 6ms works fine at any speed up to somewhere between 80 and 85, which is where the 101's absolutely excellent keying starts to breakdown. I'm guessing due to relay switching time. 

Steve, WF3T