Amps
[Top] [All Lists]

SV: [AMPS] G2DAF Circuit

To: <amps@contesting.com>
Subject: SV: [AMPS] G2DAF Circuit
From: W8JI@contesting.com (Tom Rauch)
Date: Sat, 28 Oct 2000 08:24:43 -0400
> I do get your point regarding the circuit but the point we are trying to
> make is that despite the fact that you and others try to tell us it
> doesn't work it does work. 

I keep saying, and I'll say it again, I've even seen class C PA's that 
"work" on SSB. 

I would not say a class C circuit is a "good idea"  based on the fact 
some can indeed produce, in some cases, acceptable IMD 
performance.  

WITH respectable IMD, the measurements that
> were carried out is proof of this. I think we have done far more extensive
> testing of these amplifiers than many others have done with their homebrew
> stuff. Far more..

Maybe the tests were "extensive" compared to others, but I'm sure 
we all agree the test was not conclusive about the system in other 
applications...which is the point I keep trying to make. 

> Mr DAF might very well have the wrong numbers in his article, I will not
> and have not contested you on that as you are much more capable to check
> them out than me. But that is another story.

When the voltage across a resistor does not equal the PEP power 
applied, someone missed something pretty basic. My only point 
with that is we can't rely on the numbers for understanding what is 
going on when the numbers are so obviously wrong.

When the loading on the exciter varies by around a 2:1 ratio over 
the RF cycle and envelope cycle, that usually isn't good. It doesn't 
mean the circuit will not "work" in some cases, it just means the 
design isn't very good. However, I can take that varying load and 
find some exciter that is happier with it than a constant load.

That's why ANY IMD test requires an almost perfect source. For 
broadband tests, I use an exciter that is more than 60 dB down in 
out-of-passband crud. For two or three tone tests, I mix single tone 
transmitters in an almost perfectly IMD-free system.

Such measurements are repeatable, when the source is much less 
distorted than the DUT. I can send a PA out to the someone else 
not doing that, they can measure it five times, and get five different 
numbers.  

The same thing typically happens with antennas. Someone will 
install an antenna that really doesn't work like they claim, but in 
their environment they are quite happy with the antenna. That's 
because they stick the antenna in a cluttered environment, 
compare it to another antenna in a cluttered environment, and 
reach a conclusion that "it works this way". 

Coaxial dipoles are one example, quads are another, where the 
claims don't agree with the facts. Like with an antenna, there can 
be a lot going on in any IMD test. The user can be as happy as a 
pig in slop, but that doesn't mean the system works as claimed.

One fellow in Ohio was going to patent an amplifier that REDUCED 
exciter IMD based on his extensive IMD tests with a two-tone 
signal fed into the microphone input of a solid-state amateur 
transceiver. I understand quite well how in one particular test setup, 
it is possible to add a PA that actually improves IMD of an exciter 
with IMD.

I can add a FET amplifier behind my FT-1000D, and it actually 
cleans my FT-1000D up a noticeable amount. If I add that same 
PA behind a KWM-2, it gets notably WORSE. 

The problem is if that PA is moved to another exciter, it can make 
IMD significantly worse. After I suggested he move the PA to 
another few exciters and do the same test, he became totally 
silent about his future IMD reducing patent.

My only point is the G2DAF system is poorly engineered, and is 
more hopeful thinking than good circuitry. There certainly might be 
cases where it works, I'm not disagreeing with that.
73, Tom W8JI
w8ji@contesting.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com


<Prev in Thread] Current Thread [Next in Thread>