Output Spike on Icom-706 (Summary)

[AA6KX at aol.com]

In response to the requst I put on the reflector the other day for
confirmation of the problems I'm seeing with the output waveform on my new
IC-706, I have received several interesting replies.  It certainly was not my
intent to start a religious discussion with this posting, but that seems to
have happened in some quarters:

>> ICOM s  are delicate radios.
>> 
> Hey...
>
>I thought we covered this LAST week!!!!!!

On a more serious note, one very helpful fellow did take a look at the
problem I'm seeing and send me the following:

"There is a "spike" that decays over several RF cycles. It changes from band
to band and with supply voltage. That indicates it is from ALC response time.
...
The IC-706 produced a maximum "spike" of 130 watts when first keyed when the
supply voltage was 16 volts. When the supply voltage was reduced the "spike"
(actually several decaying RF cycles) was reduced also.
...
The power could be set at any level and the "spike" (actually envelope
overshoot) isn't reduced. The worst case overshoot will always be with high
supply voltages and low rig output powers. For example, if the rig is set at
60 watts and the supply is 15 volts, the overshoot will be 120 watts or so
before settling down to 60 watts..."

At least now I know that the problem is generic and not peculiar just to my
rig.  I'm still trying to get a test bed set up to measure the duration of
the spike myself, but the way this thing zaps my 87A it sure looks like more
than a few RF cycles!  It's enough that it will fault the Alpha off-line if I
set the drive in the 706 for 70-80 watts, which is the drive needed to push
the 87A to 1500 watts.  Based on the above, the 706 is leading off with 130
watts instead of 70 watts until the ALC kicks in to clamp the output back.
 This is not good.  In fact, I consider it unusable this way because of the
risk of damage to my amp.

Several people wrote to mention that they were using the 706 with other Brand
X amplifiers and weren't having any problems.  I will counter that they can't
see the problems because most meters will not react quickly enough to show
it.  In the meantime, I suspect they are putting a stress on their amplifiers
that will reveal itself in time.  One person mentioned he had already been
down this painful and expensive road with a solid state amp.

Also, several people quoted the line in the March QST review that said the
706 cannot key an amp directly, and referred me to the optoisolator circuit
shown in the November QST (p 84).  The problem here is that the IC-706 T/R
switching cannot handle very high current or voltage (as is the case with
most modern solid-state rigs).  The Icom service people told me the limit was
about 20 ma.  In any case, the Alpha 87A only puts about 4.6V on the rig's
T/R switching and I measured only 1 ma of current drain.  This is well within
tolerance for the IC-706.  But just for the heck of it--and because I do
sometimes use older amps--I went ahead and built the electronic relay given
by K1KP in that tech note.  This was a little harder than I expected, because
the QST article gives the wrong pin-out for the 4N35 in the schematic!
 (After destroying one 4N35, I learned that the emitter to the triac is on
pin 4, not on pin 6 as is shown in the drawing.)  If typos are going to sneak
through the proofing process back in Newington, why why why do they have to
be in the pin-out on the critical part???

Bruce, AA6KX