> I noticed the same thing when installing the mod in my FT-1000D.
> Putting an NPN transistor in series with the keying line means
> that the switching diode is not reverse biased on receive. If I
> remember correctly, the signal out of the balanced modulator is
> on continously which could result in some leakage through the diode
> possibly causing other problems. In the end I decide not to install
> the mod as on the air checks indicated no problems with clicks.
The mod is a patch. Like all patches, it is less than perfect but it works
well and has never caused any type of problem. It isn't even necessary to
reverse bias the "switching" diode that isn't a switching diode.
Yaesu obviously doesn't worry much about the negative bias (which leaks in
from the -9 volt line through 100k resistor R2173), since they use
conventional polarized electrolytics on those lines. For example, look at
C3243.
Even if you don't turn off D3009, the signal goes absolutely nowhere. D2050
is off from positive voltage, so any worries about RF leakage around D3009
from the balanced modulator are unfounded. The signal has nowhere to go on
TX, and on receive everything is held off. I'm not even sure why 250 of the
diodes and at least 1000 other parts are in the radio. Perhaps they had a
surplus of parts they needed to get rid of.
> 100 us seems like an awfully fast rise time although that IS the
> time constant for the original components (R-3119 and C-3243). The
> rise time on my rig with those components seems to be about 2.5 ms.
Since the RF envelope does not follow the rise time linearily, the time
constants do not produce the same envelope time constant. While a GOOD
design would use a linear modulator, we are stuck with what Yaesu decided to
do. (Unless someone makes new RF, IF and AF boards for the radio.)
> This seems to be too fast for click free operation so I'm going to
> take some more measurements and revisit the problem. Perhaps changing
> some of the original component values will result in a softer waveform.
Been there, done that. Changing original values will work, and you can make
the waveform better, but few people would want to pull all the boards and
change dozens of parts. My various mods used PIN diodes in a linear
modulator driven by an op-amp multi-pole filter with 200Hz cutoff, multiple
tandem R/C filters, and using my 250 Hz transmitter filter. None of them
work noticeably better than the $1.00 mod that does not require extracting
components, even though they "look better" on a scope.
ALL FT1000's as shipped from the factory click excessively and needlessly.
You won't bother anyone unless you are loud, and close in frequency. The
bandwidth is not symmetrically distributed either, it actually has peaks at
certain spacing. Rise and fall is calculated at 10% and 90% of envelope,
not 0 and 100%. Unless you go 60WPM, no one will notice or complain about
soft keying.
73 Tom
|