Here's what I think is happening.

On 40 meters, the path from the tuner back along the outside of the coax
to the rig is about as attractive to RF as the dynaplate, so a good
fraction of the RF current being fed to the antenna goes that way. On 20
meters, the dynaplate is relatively more attractive due to a combination
of the dynaplate impedance, antenna input impedance, and coax length, so
more goes to the dynaplate and less along the coax.

The RF current goes along the outside of the feedline to the rig, over
the outside of the rig, and into your inverter via the power leads.
There are a couple of ways you can reduce the problem:

1. Put one or more good current baluns (common mode chokes) in the coax
feedline. If you just put one in, I'd put it at the rig, to force the
current there to be minimal.
2. Put a common mode choke in the power cable to the rig. Because the
wires are pretty large, it's probably not practical to wrap multiple
turns around a single ferrite core. So get some large clamp-on cores you
can clamp over both conductors at the same time. It's likely to take a
dozen or so, or more depending on the type of ferrite. I recommend type
43, or 70-series if you can find it in clamp-on cores.
3. Apply the same kind of choke to the inverter power leads from the
battery.

What you're doing here is to create a high impedance to common mode
current along the path from the tuner to the inverter.

If you'd like a quantitative measure of how bad a problem you've got and
how effective any solution is (or if this is really the problem), get a
clamp-on ferrite core you can clamp over the coax line or the power
leads to the rig. Wind about 10 turns to make a transformer secondary
winding -- the primary will be the wire passing through the clamped
core. Connect the winding through a series capacitor of around 0.001 to
0.01 uF to a shunt diode (any small signal type will do), and then to a
voltmeter:

----| |----.----
cap |
from winding _|_ to meter
diode /\
|
-----------.----

Sorry, I'm not very good with ASCII art. . .

Clamp this onto the coax or over both power leads. The meter reading
will be approximately proportional to the amount of RF current. The
object is to minimize it.

Roy Lewallen, W7EL

Newbie Ham wrote:
Hi Everybody

I hope this isn't the wrong place to ask. if it is, please let me know
where I should post.

I have a Kenwood TS50 installed on a boat with a SGC autotuner. The
tuner is grounded to a dynaplate.

I also have a 2000watt inverter/charger built into the boat for ac power
creation from 12v batteries.

Something strange is going on and I jusr don't know where to start
trouble shooting.

When I transmit on frequencies like 14300MHZ usb, no problem.

If I transmit on 7628 LSB 100watts the FET's in my inverter blow up! As
soon as I key the mike I hear a loud pop and there goes another fet.

I've changed them twice now and since the inverter is bolted into an
awkward spot and weighs 70 pounds (it's all transformer), removing the
inverter and changing them ain't fun.

Tonight I went so far as to disconnect (as in unplug) the AC supply to
the inverter and switch it completely off. As in no LEDS lit, everything
off.

Yet as soon as I keyed the mic, POOF. The fets just blew apart.

Any thoughts as to what might be happening?

FWIW, the coax from radio to tuner runs within about 2 feet of the
inverter, the radio and tuner are powered from the batteries which power
the inverter. So the inverter and radio share Pos. and Neg.

I can only think of a few ways this might be happening:

1) Radiated signal from the coax leaking into the inverter.
2) Radiated signal passing into the inverter via the shared positive or
negative feeds.
3) Some weird ground loop issue.

I have no ideas as to how to diagnose this and trouble shoot it without
having to replace fets everytime. And that's a big job.

Any thoughts would be greatly appreciated.