There seems to be some misunderstanding of how IMD happens. The voltage
output *divided-by* voltage input equation for any two port device is
always a Taylor series expansion, which includes linear, square, cubed,
etcetera terms in it. If one processes a single sinusoidal signal
through this expansion, the squared terms give rise to a second
harmonic, the cubed terms a third harmonic, and so forth. But then, how
does one get intermodulation, a mixing of two signals? If one delivers
two pure tones to an amplifier which exhibits such a Taylor series
expansion characteristic, one of the terms is:
( A*sin(2*pi*omega1*t) + B*sin(2*pi*omega2*t) ) ^2
where:
A,B are the peak voltages of the two separate signals,
omega1, omega2 are the angular frequencies of the two signals
pi and t are evident.
^2 is the best I could do at making a squared sign!
The expansion of this term gives us four separate frequencies: twice
each of the original frequencies, plus the sum and difference of the
two.
Now that we have the second harmonic of each, the very same second-order
term can cause a mixing of one of the second harmonics with one of the
fundamentals, producing a third order product. Similarly, a fifth order
product comes from the fourth power term plus the second,etc. An
interesting note here - a push-pull amplifier suppresses only the odd
order harmonics. Do you suppose it has improved IMD? As it turns out,
push pull amplifiers have strikingly high IMD for their relatively good
harmonic performance.
One fallout of this is that I cannot create a third order IMD product if
I do not create a second harmonic. This fact is used to advantage in
television transpose amplifiers. The designer aims the output network
such that it appears to the amplifying device as a short circuit at the
second harmonic, and they can achieve miraculous levels of third order
IMD - often 60dB down - in a class B amplifier. The drawback, of
course, is that tuning is very narrowband.
So if an amplifier has a stimulus of only 30 MHz, and we see some 350MHz
signals, then some of the 350 might be IMD, but since IMD cannot exist
without harmonics, then chasing IMD by itself would have little use, so
the focus should be harmonics.
Harmonics at the anode may actually be a good sign. If we want
efficiency, we can arrange to have the ouptut network appear very high
impedance at the odd harmonics, which tends to make the anode voltage
appear as a square wave, which means better efficiency. Classes D, E,
F, etc, are all based on controlling the waveshape at the output of the
device to be square-like. Therefore, I put little value in observations
of the anode voltage waveform or harmonic structure. It is not
representative, in most cases, of anything that would occur at the
antenna output.
Arlen
>From owner-amps@contesting.com Fri May 1 01:48:07 1998
>Received: from dayton.akorn.net (dayton.akorn.net [205.217.100.11])
> by dayton.akorn.net (8.8.5/8.8.5) with SMTP id EAA26816;
> Fri, 1 May 1998 04:47:50 -0400 (EDT)
>Received: by dayton.akorn.net (TLB v0.10a (1.23 tibbs 1997/01/09
00:29:32)); Fri, 01 May 1998 04:47:44 -0400 (EDT)
>Message-ID:
<F2B87FD6828AD11196F100A0C9410EC2617D9A@excalibur.swindon.msl.mitel.com>
>From: Peter Chadwick <Peter_Chadwick@mitel.com>
>To: "'amps'" <amps@contesting.com>
To: <amps@contesting.com>
>Date: Fri, 1 May 1998 09:47:00 +0100
>X-Mailer: Internet Mail Service (5.5.1960.3)
>Sender: owner-amps@contesting.com
>Precedence: bulk
>X-Sponsor: W4AN, KM3T, N5KO & AD1C
>
>>That's not an intermod problem, it's a harmonic problem.
>
>Unless you carefully define which frequencies are there, it could be
either,
>or both. However, if the analyser isn't being overloaded, and isn't
telling
>lies, then from what was said, it seems there's something not quite
right.
>It does depend a lot on the test set up and the coupling, though.
>
>73
>
>Peter G3RZP
>
>Out of circulation 2 thru 11 May.
>
>--
>FAQ on WWW: http://www.contesting.com/ampfaq.html
>Submissions: amps@contesting.com
>Administrative requests: amps-REQUEST@contesting.com
>Problems: owner-amps@contesting.com
>Search: http://www.contesting.com/km9p/search.htm
>
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|