I think Jerry has a much simpler idea -- just
find a way to disconnect the gate bias during receive.
This would put the gates at 0v, and effectively shut off
the IRF510 drains. Probably a good idea.
It might work, but I wonder about the sudden increase and decrease
of voltage on the gates. it might precipitate oscillation or action in the
non-linear area of operation. The linear area we are using is quite
narrow and a bit unstable.
Probably it would work fine for most, but not so good for a few. It depends on the
actual device in use.