[AMPS] 3/2 power law and Rich's very wide signal

[measures]

>
>Rich and others, collectively, wrote:
>
>>>>This is precisely whan happens.   The first cycle out of the exciter can
>>>>be anything up to the limit of the Final Unit.
>>>
>>>
>>>Surely not in practice.
>
>>This is what my oscilloscope indicates.
>
>That is the mark of a very poorly adjusted exciter, Rich.  

It is a misstatement.  My exciter is adjusted to deliver 100w, which is 
100v-pk..  When the VOX trips, the 100v level is reached in a bit under 
2mS.  This is not enough time to charge the screen bypass cap.  

>Just as the 3/2 
>power law is, in fact, a law that physics will enforce regardless of our 
>desires, so, too is the relationship between the time domain and the 
>frequency domain.  That relationship has been variously set forth by the 
>likes of Oliver Heaviside, Fourier, LaPlace, and many others who predate the 
>3/2 power law by hundreds of years.
>
>If your exciter really does produce a signal whose amplitude is full power 
>at the first RF cycle, you will be generating key clicks that are extremely 
>wide. 

True enough.  .  

> Far wider than the 100 or so kHz you are claiming was produced by a 
>DAF amplifier.
>
Only one of the DAFers was capable of producing such a signal, Fred.  The 
bandwidth of the other two was on the order of roughly 30KHz.  When CHR 
was really crankin', it was close to 150KHz.  

>A properly adjusted "cw" rig (I put "cw" in quotes because our on-off keying 
>method is not scientifically cw, which would be a continuous carrier, or A0 
>modulation), MUST shape the keying waveform to eliminate key clicks.  As 
>Steve alludes, the effective bandwidth of the signal is roughly equivalent 
>to 1/3*(rise time of the signal).  I emphasize the word "roughly", but it is 
>within an order of magnitude.  If properly shaped, a 'morse' code signal has 
>a bandwidth of about 1/5 the speed in words per minute.  If the rise time is 
>too fast, key clicks will result - the cw equivalent of "rotten splatter".  

indeed

>You can feed your oscilloscope's waveforms into any free version of SPICE, 
>or, for that matter, any Excel-based FFT algorithm to see for yourself.
> 

>Rise times on the order of 100 microseconds to 500 microseconds are what 
>I've measured in the past.
>
Agreed.   Do you feel that a typical screen bypass C can be fully charged 
in 100uS in a DAF amp.? This is the real issue here.  My guess is that 
Mr.  Thornley was not able - or was too lazy - to build a proper screen 
supply and  bias supply  .  

>I believe an RF source CAN produce a full power waveform in the first RF 
>cycle.  

This would require an antenna charging current of quite a few kA.  

>I just hope to not be on the band when it happens.  

W6CHR comes close.  

>I remember when 
>I first saw this in action.  I was listening about on 40 metres one day and 
>heard a lot of clatter.  I was tuning about the high end of SSB, perhaps 
>7.28 MHz, and wondered what the noise was.  On the fast AGC setting of my RX 
>switch, the peaks were about S9.  I tuned lower to find the signal.  It was 
>a local ham on about 7.040 calling CQ.  His main signal was only S9 + 15dB!  
>I called him on the phone to see what he was using for a rig.  He'd recently 
>installed "relay QSK".  He had a relay doing the TX/RX switching.  The 
>timing was such that the relay was making AND breaking with RF actuated.  So 
>it was a guarantee that his first RF cycle was full power and his waveform 
>also was cut off with the same rapidness.  Unfortunately, I could hope his 
>relay would wear out, but as he was running a mere 10 watts, I had no such 
>hope.  I was a lad of only 20, and he an old timer of 60, so there was also 
>no hope of convincing him otherwise.
>
Alas.  There is no fix for jello-brain.  If 17 year olds are too young to 
vote, 67 year olds should be too old to vote.    Older is definitely not 
always better.  

>Many of the reflectees have scopes.  Next time you think of it, see if you 
>can gather a sample of your cw waveform.  

I have.  Risetime is c. 2.5mS .  

> Don't even try it with an 
>oversampling digital scope.  You'll have to use a single-shot digitizing 
>scope or an analog scope to get the true picture.
>
>Cheerio,
>
thanks for straightening me out, Fred.  Your turn.  . 

-  Rich..., 805.386.3734, www.vcnet.com/measures.  
end


--
FAQ on WWW:               http://www.contesting.com/ampsfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm