Topband: Key Clicks

Tom Rauch w8ji at contesting.com
Mon Jan 5 03:20:00 EST 2004


> clicks, hiss on CW-LSB, AGC, and increase of 70 Mhz IF gain are all
> available from Inrad for a total for all mods of $55 plus shipping.  The
> instructions are clear and while the key click mod is not the world's

Yaesu did a good job keeping gain ahead of the filters low, and they use
stout mixers. The primary source of hissing is lack of filters just in front
of the product detector and the very high IF gain just in front of that
point. I traced almost all the "hissing" to the last IF amplifier stage.
You might just try an audio filter instead of moving gain around.

If you are ambitious, I used a very low noise GASFET in the last IF
amplifier of my FT1000. The device I picked has lower gain than the stock
Yaesu part. The lower gain allowed me to move the gate up to a higher tap on
the IF transformer, instead of mismatching the input of the FET to control
gain and running the device gain really high. Doing that and using an audio
clean-up filter, I don't have to increase gain before the narrow filters
(adding 70MHz gain has virtually the same effect as adding an additional
preamplifier for signals within the roofing filter BW).

I'd like to make one additional comment. We really need to establish a
standard for measurement of transmitters on both SSB and CW, and it should
be the effective spurious power level on close adjacent frequencies. Looking
at one spot with slow sweep and peak storage on a spectrum analyzer really
is not a good test.

It's pretty easy to see why it is a bad idea to pick one frequency and use a
long average of maximum signal peaks with a narrow filter sweeping by, which
is what all spectrum analyzers do. For example, some radios like the Omni 6
have a sharp click on one edge, and nothing on the other. The peak energy is
very high, but the average is low. The Omni 6 can have higher peak level and
do less damage to adjacent channels than a rig that hammers on both make and
break, like the FT1000MK V does.

If I look on a peak sample and storage device like my spectrum analyzer, the
Omni 6 could look worse yet bother adjacent CW bandwidth channels LESS than
a lower level radio.

A narrow filter sweep with average storage of peaks also does not account
for the slope of the sidebands of the radio. I'd rather have a radio next to
me with a brick-wall drop of clicks at 500Hz than one with a gradual slope
and a little less level. A peak comparison at XX spacing does not account
for that.

This is actually the same problem we have with transmitter IM testing. The
FCC has changed some commercial tests to use spurious power level on
adjacent channels, rather than somewhat useless two-tone tests. A two-tone
test shows the very best a radio is likely to do, not the worse. The FCC
requires normal modulation and an adjacent channel power test now in some
cases.

CW and SSB testing should be the same way.

My last point is this. While I agree with W2VJN that many radios click (for
example the Elecraft isn't good), Yaesu deserves all the heat they get. All
they would have to do is change the VALUE of several components in
production and 99% of the problem would go away. Instead of responding to
feedback (which I am sure they had plenty of) they stonewalled or ignored
the problem and released a new series of FT1000 radios that are worse than
the original FT1000D (and probably the MP). I could understand if they
didn't know the poor selection of a couple capacitors and resistors was
causing grief, or if the change cost money or took time. But not doing
anything to at least make such a problem better in such popular radios is a
serious insult to their users. It shows CW operators are not important
enough to warrant them changing the VALUE of a few components, even though
they redesigned PA linearity to improve SSB performance.

Don't look for things to get better until we as consumers take some sort of
action. If anyone is interested in the rules that apply to modulation
bandwidth I put it up at:

http://www.w8ji.com/fcc_97_307.htm


73 Tom




More information about the Topband mailing list