[Amps] Non-linear subharmonics was: Nonsense, AL-1500 parasitic
jeff millar
wa1hco@adelphia.net
Sat, 18 May 2002 08:02:24 -0400
Here's two web pages that discusses using variable capacitance to produce
gain at f/2 as part of a larger discussion of varactor diodes and parametric
amplifiers.. These pages specifically discusses why f/2 works.
http://www.tpub.com/neets/book11/45k.htm
http://www.tpub.com/neets/book11/45l.htm
I know this doesn't directly address the question of oscillation or signals
at f/2, but apparently that technique lived and died before the internet so
only exists in libraries. Steve Thompson's email of 5/16 described the
mechanism well:
In linear circuits, I expect it's not possible, but few circuits are
without
non-linear aspects that become significant under some conditions. As
best I
remember the article I read, the argument goes that if anything (even
noise,
maybe) causes f/2 to appear in a circuit, then IMD between f/2 and f
produces
3f/2, then both f and 2f with 3f/2 produce f/2 and so on, so it can
become
stable and self sustaining if the circumstances are right. The same
analysis
can apply to f/3 etc. It's very common to find it when doing load pull
tests
on solid state PAs.
Steve
I'm still looking for the proper references. Does anyone have a textbook on
varactor multiplier theory and design?
In general parametric amplifier theory uses an RF "pump" to dump energy into
a non-linear device, which then can look like a negative resistance
terminating a transmissionn line. And negative resistance terminations have
the characteristic of reflecting more power than incident. Given gain and
a circuit, you can generate an oscillator...and that oscillator can
injection lock to the pump...thus transforming input energy to subharmonics.
jeff, wa1hco
----- Original Message -----
From: "Tom Rauch" <w8ji@contesting.com>
To: "'Amps'" <amps@contesting.com>; "David Robbins" <k1ttt@arrl.net>;
"Michael Tope" <W4EF@dellroy.com>
Sent: Saturday, May 18, 2002 6:45 AM
Subject: Re: [Amps] al-1500 parasitic?? or some other problem?
> > Tom, I thought Dave said that the 21 MHz spurious signal tracked his
> > fundamental and held a 3F/2 relationship with it all the way across
> > the 20 meter band. I don't see how a the 14 MHz fundamental beating
> > with a fixed frequency oscillation could do this (Dave was the spur
> > tracking really exactly 3*F/2 WRT to the 14 MHz signal??).
>
> A possible clue would be to change frequencies and watch the spurious
> frequency change.
>
> > Can a free running oscillator injection-lock to a pump signal near its
> > 2nd or 3rd harmonic? If so, perhaps that would explain it? Another
>
> Locks at half the frequency are nearly impossible, because the
> positive peak of the oscillating system would have to lock on both
> positive and negative excursions of the main signal. True sub-
> harmonics in non-linear devices are totally impossible.
>
> One thing that works is to understand the complexity of and the
> special devices with special characteristics required to **
> intentionally** make the same problem. Considering how we would do it
> intentionally, we can guess at how likely it is to occur in an
> accidental circuit.
>
> I remember in a local electronics class the teacher heard my 160-
> meter signal on a broadcast radio at about 900kHz when I was on about
> 1800kHz. he used it to demonstrate "sub-harmonics" to the class, and
> the question even appeared on an exam.
>
> When I caught wind of it, I had him do a simple test and beat an
> oscillator against my alleged "sub-harmonic" while tuning the radio
> dial. The pitch of the sub-harmonic changed when he tuned the radio,
> which it never would have done if it was a sub-harmonic.
>
> I explained to him it was a receiver deficiency, aggravated by the
> 455 IF and local oscillator choice. The local oscillator was at 1355
> when the receiver was on 900kHz. (455+900=1355) My signal was at
> 1810. 1810-1355=455
>
> What worked out to be a subharmonic really was a IF image.
>
> The point of all this is we sometimes miss simple in-your-face
> frequency relationships. With so many oscillators in the multiple
> conversion receivers and transmitters, and other signals in the
> outside world, a lot of stuff can be going on.
>
> A virtually impossible scenario is a f/2 unintentional divider
> remixing, because sub-harmonics are virtually impossible if not
> totally impossible to intentionally generate, let alone have by
> accident.
>
> On the other hand accidental mixers are simple and common, easy to
> build, and can occur many places in the outside world...but they
> require the correct accidental frequencies.
>
>
>
>
>
>
> > possibility would be some poorly shielded RF device with a frequency
> > divider that is capturing on Dave's nearby TX signal (or one of its
> > harmonics). The frequency divider could produce the F/2 component
> > which in turn could beat with the 14 MHz signal and to produce the
> > 3F/2 component. The fact that the spurious signal is bidirectional in
> > its response to Dave's antenna suggests to me that its due to a
> > harmonic of his signal that is created either in or before the 20
> > meter antenna (I would expect a Yagi to look bidirectional at its
> > harmonics, whereas it would be unidirectional at its operating
> > frequency).
> >
> > Interesting problem.
> >
> > 73 de Mike, W4EF...........................
> >
> > ----- Original Message -----
> > From: "Tom Rauch" <w8ji@contesting.com>
> > To: "'Amps'" <amps@contesting.com>; "David Robbins" <k1ttt@arrl.net>
> > Sent: Friday, May 17, 2002 6:45 PM Subject: RE: [Amps] al-1500
> > parasitic?? or some other problem?
> >
> >
> > >
> > >
> > > > Well, I can now say for sure that it is not the amp. I have found
> > > > that by feeding only the top 6 ele 20m beam and pointing it either
> > > > due east or west I can generate the f/2 and 3f/2 products with as
> > > > little as 45 watts. The beam heading is not what I was expecting
> > > > as it puts both the 40m and 15m towers off the ends of the 20m
> > > > elements. The low power needed does really narrow it down to
> > > > something related to that top antenna or something very close to
> > > > it... especially since I can not generate them with 1500w to a 4
> > > > ele 20m beam 45' below it.
> > > >
> > > > I can also say that its not the guy wires as the 17 year old rusty
> > > > ones were replaced yesterday.
> > >
> > > Hi Dave,
> > >
> > > I'm positive you are wasting time looking for non-linear joints
> > > producing a F/2 or 3F/2 product. Sub-harmonics are impossible to
> > > produce by virtue of a non-linear system. . Look for a mixing
> > > product with another local oscillator or signal of some sort, or a
> > > spurious oscillation in a device with gain.
> > > 73, Tom W8JI
> > > W8JI@contesting.com
> > >
> > > _______________________________________________
> > > Amps mailing list
> > > Amps@contesting.com
> > > http://lists.contesting.com/mailman/listinfo/amps
> > >
> >
> >
> >
> > _______________________________________________
> > Amps mailing list
> > Amps@contesting.com
> > http://lists.contesting.com/mailman/listinfo/amps
> >
>
> 73, Tom W8JI
> W8JI@contesting.com
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps