[1000mp] Key Click Data on line

[Jim Smith]

These plots are very interesting.  

When looking at the plots I was struck by the pronounced keying 
sidebands for some of the radios and the lack of them in others.  e.g. 
for the unmodified MP in the region -1000 Hz to - 500 Hz wrt the carrier 
frequency there are about 10 identifiable keying sidebands.  (I picked 
this particular range because they were easy to count, not to support 
some personal bias against MPs.)  Other rigs don't show nearly as 
pronounced sidebands.  In particular, the K2 shows almost indiscernible 
keying sidebands in that particular frequency range.  i.e. they're 
really hard to count.

So, what am I looking at in the K2 plot?  Clearly, at least in that 
range, the keying sidebands provide an almost minuscule contribution to 
the power at any frequency in the selected range.  So where does that 
power come from?  Could it be that mysterious phase noise I hear about 
from time to time?  No BS here.  I really don't know what phase noise is 
but I AM interested in knowing what these plots represent.  It's obvious 
that the spectra show, for each Tx, that there is power being emitted on 
either side of the carrier frequency, some rigs more than others. 
 Superimposed on this power is the keying sidebands.  In the case of the 
unmodified MP the superimposed keying sidebands are clearly a more 
significant percent of the power generated in the frequency range I 
selected than many other rigs, the K2 being the most obvious.

What I am trying to communicate here is my belief that, if one were to 
plot the spectra for the same rigs under the same conditions except with 
a brick holding the key down (i.e. steady carrier), you would see 
exactly the same thing except for the wiggles due to keying sidebands. 
 It would be very interesting to me to see plots for the same rigs with 
the same test gear for the steady key down condition.

If I am correct in my belief, the numbers quoted for the K2 of -54 dB do 
show that it emits something 54 dB down 1kHz from the carrier but that 
key clicks are a very small contributor.

 From this perspective, the plots show that the unmodified MP clicks way 
more than the other rigs.  The other rigs do, however, emit something 
about as strong as the MP clicks but it would seem to be not as easily 
identified by the ear.

I guess I'm saying that the measurement of the power emitted 1 kHz away 
from the carrier frequency by a particular rig when keyed at 40 wpm 
doesn't say anything about key clicks as it looks to me as if, in all 
cases, the power in the keying sidebands is a relatively small - but 
very annoying if you're listening in key click alley during a major 
'test - addition to something else that's already there, perhaps the 
aforementioned phase noise.  In the case of the unmodified MP it looks 
like the keying sidebands' contribution is greater - the individual 
sidebands are not masked by the "other stuff" to as large an extent as 
evidenced by the fact that they are easier to count - and they are at 
about the same or higher level than most of the other rigs shown.

To see an example of the effect of masking, look at the lowest 500 Hz 
segment of the modified MP2 plot.  Here the keying sidebands are really 
obvious.  That's because their contribution to the total power in that 
frequency range is much greater than that of the "other stuff".

I find it interesting that the Inrad mod for the MP reduces the level of 
the MP's "other stuff" by about 20 dB 500 Hz off the carrier frequency. 
 While this "other stuff" is clearly not as obvious as clicks it must be 
contributing to the noise floor when you're listening near a strong 
signal from any of these rigs.  The K2 and Omni look like the worst in 
this regard as the "other stuff" is only 40 dB down at -500Hz.  (Of 
course, this does help to mask the clicks.)

Why don't I ever talk about anything on the high side of the carrier 
frequency?  Back in the old days one could get away with, "The high side 
cases are left as an exercise for the student."  The reality is that I 
have spent better than 2 hours composing this and building in the 
anti-flame shields and want to get on to getting my contest score in. 

Back down the dusty decades when I thought I knew everything I would 
have said that if you want to measure the power in the keying sidebands 
you would key the rig with a square wave generator to remove the 
spectral variations due to keying with morse characters with their 
varying duty cycle and use correlation detection to measure the power of 
the sidebands and ignore the rest.

Time to admit my personal bias.

K2 - I have actually seen 2 of them.  Total viewing time - about 10 minutes

MP - I own a MkV and love it.  Guess I should do the click mod.

I would be pleased to contribute whatever I can to any discussion this 
might provoke, including challenges to my stated beliefs.  I have a lot 
to learn in this area and there's nothing like a challenge honestly made 
and honestly met to refine one's thinking.

Hope this is of some interest.

73 de Jim Smith    VE7FO

      
George Cutsogeorge wrote:

>I have put keying spectral data up on the Inrad web site.  It includes MP
>before and after data as well as data on several other popular radios.
>http://www.qth.com/inrad/about-key-clicks.pdf
>
>George, W2VJN
>