Just an observation from looking at
http://home3.swipnet.se/~w-30548/g2daf/pa5.gif
I believe the spectrum analyzer controls were set improperly. The
resolution bandwidth was set for 1 kHz. This is too wide for properly
analyzing an SSB signal produced by a male voice. The male voice
fundamentals are below 100Hz, and thus the spectrum is a sequence of
tones separated by about that amount (or double, depending on the
existence of odd and even order overtones). These tones are not visible
in the display.
Remember that a spectrum analyzer's RBW filter is not a receiver
filter. A spectrum analyzer's resolution bandwidth filter is Gaussian
in shape, so as to preserve pulse fidelity. As a result, it has very
gently sloping skirts. One kHz in this case represents the 3dB down
points. The apparently-triangular shape in the spectrum analyzer plots
is the shape of the spectrum analyzer filter. As you can see, the
filter is so wide that it covers many of the probable IMD products. My
suspicion is that an RBW of 100 Hz or less is needed. VBW can be set
quite wide (>10X RBW) because VBW filtering is not needed for this
peak-hold based measurement.
Also, the noise floor appears too high. The block diagram shows 98.5dB
of attenuation between the amplifier and the analyzer. And the analyzer
screen shows an internal 30dB attenuator is turned on, so you have a
total of 128.5dB of attenuation from the amplifier to the spectrum
analyzer's mixer in the front end. If the amplifier is delivering 1kW,
or +50dBm, then the analyzer's mixer is receiving only -68.5dBm peak
power. While this assures you won't blow the front end of the analyzer,
it also means you'll be noise-floor limited. Better to set the mixer
input level between -20 and -40dBm (at least for the 8590) to improve
the dynamic range of the analyzer. Also, an investigation should be
made of what the real attenuator values are. Based on the analyzer
display, the peak power delivered by this particular amplifer is +66dBm
or 4kW. I can't tell from the photographs what tubes are used in this
EBK-1, but could they possibly produce 4kW peak? So it would appear
something is amiss in the attenuator chain.
Remember, when you're doing IMD measurements, you have not achieved a
useful result until you can actually see the IMD. All this plot shows
is that the IMD is below the 8590 filter skirts. Whether that is
acceptable IMD or not is not certain. Let's say there was an IMD
product 20dB down from peak and in the next adjacent channel (3kHz
away). It could not be seen on this plot because the filter skirt is
covering it.
Unfortunately, all my ham gear is stowed, otherwise I'd pull out a rig
and pop it on an analyzer and see what sort or settings seem to be a
good compromise between speed and display of IMD.
I'm making no comments on the G2DAF circuit itself. I've been following
the thread with a bit of curiosity and admit some temptation to build
one just to see what it can do. But the reality of a job with lots of
travel and fixing up a sixty year old home in California tend to put a
serious clamp on hobby activities!
Regards,
Dave W8NF
--
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
|
