I received an overwhelming number of helpful suggestions both on and
off the lists. I also received a large number of requests to pass
along any solutions found. Thanks to all who responded. Since there
were so many requests for follow-up information I will do it through
the lists rather than individually.

I am *not* an expert by any means but will offer a few thoughts
based on experience and what I learned after asking this question. I
have extensive experience protecting remote (at the antenna) preamps
on VHF and UHF. What I wanted to do at LF and MF is a bit different
and I had little to no experience with it.

The ultimate receiver protection is to disconnect the antenna from
the receiver while you are transmitting. Very often a 50 ohm
resistor or in some cases a short will be switched across the
receiver input at the same time. You have to be sure the switching
happens before the start of RF from the transmitter.

There are various switching devices on the market under names such
as front end saver, front end protector, T/R switch, etc. Sometimes
such devices use 'RF VOX' where the presence of RF from the
transmitter triggers the device to switch. Although this can work,
it seems to me a bit like closing the barn door after the horse has
escaped.

Many of these devices instead require a logic input to initiate
switching. That seems a lot safer to me. If you only use WSJT,  I
was reminded that you can use its RTS/DTR control signal and set the
timing as required for your receiver protector. If you use modes
that don't involve the JT software, you need another approach.

If using a transceiver or exciter which provides external T/R logic
(often somewhat inappropriately called a PTT signal), and if that
logic is asserted some few milliseconds before the start of any RF
output, and if you use a relay with sufficient switching speed
(faster than the logic-to-RF delay) then implementation is  simple.
You connect the T/R logic to the relay or device which will switch
in the receiver protection. The key here is to watch the timing! I
have seen many devices come to my bench for repair because the user
did not understand or did not think about the timing issues
involved. Some older transceivers have no delay at all. W8JI
discusses timing issues and how to measure delays here:

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

If you don't have a T/R logic signal from the exciter/transmitter,
or if the delay is too short to be of use, then you need to create
the switching logic and the appropriate delay yourself. I will leave
it to you to figure out how to implement this with common LF/MF
setups. On VHF and UHF where remote preamps are very common, we use
devices called sequencers or T/R time delay generators to accomplish
this. For example on my 144 MHz setup I have a four output
sequencer. It takes a single low on transmit logic input from my
transceiver and provides four outputs which go low in sequence with
about 50 milliseconds delay between them (such that the last one
goes low 200 ms after the input logic event. I have the first output
switch my preamp to 'safe' mode (disconnect antenna and connect 50
ohm resistor to preamp input), the second to my 1500 watt amplifier
relays, the third to the 60 watt driver amp relays, and the fourth
to a transverter enable line. Thus the transverter cannot produce
any RF output until everything else is ready. The sequence is
reversed when going back to receive. This has been standard practice
on VHF and up for many years.

That is all pretty simple when you have one receiver, one
transmitter, etc.

In my case I have two receivers (one LF, one MF) which are
completely independent of the several different transmitters. This
is a bit more complicated, as every receiver must be protected and
this must be triggered by any exciter that might be used. Further
complicating things, I have one very indispensable exciter which I
don't entirely trust to get the timing correct every time. For these
reasons I prefer some sort of RF limiting on the receiver input such
that it can always be left connected to the receive antenna while
transmitting without risk of damage. I didn't (and still don't) know
how to design a limiter which would keep things to a safe level for
all receivers. I was hoping someone else knew. :)

Limiters can be as simple as the old school approach of two fast
diodes connected back to back across the receiver input. With 1N4148
diodes this should limit RF to about 10 dBm (10 milliwatts) but I
use a 20 dB gain 2N5109 preamp which may be very capable of
producing 30 dBm (one watt) output with 10 dBm input! Some receivers
might not like that! I was reminded that such simple circuits often
make use of a small incandescent lamp in series to provide some
additional soft limiting. The idea, I think, is that the filament
when cold (at normal receive signal levels) has low resistance, but
as it heats up (strong RF from a nearby transmitter), resistance
increases.

I was cautioned that if f you have strong RF on any frequency
reaching the diodes while receiving, they may create all sort of
harmonics and other trash which will impair your receive ability!
Also the harmonics could possibly be radiated by the antenna,
perhaps causing interference to someone else.

There are more complex (and better) techniques for the RF limiting
approach. As noted in my original post, ICE once produced a receiver
protector (ICE-196) which used a saturable core transformer along
with didoes and other components to provide limiting. I have one
permanently connected to the receive antenna input on my FT-2000.
Nice, but the device has too much insertion loss at LF and MF.
Reverse engineering it and designing one that works at our
frequencies is beyond my level of expertise, but would be a worthy
project for someone. DX Engineering currently markets a similar
looking device but it is only rated down to 500 kHz and will pass a
25 mW signal - well above the stated danger level for some SDRs.

As it turns out, luck was on my side this time. ;-) The diodes which
I already had should limit input to the 2N5109 preamp at around 10
dBm. A 2N5109 will laugh at that. I was worried about the LF and MF
receivers, but perhaps unnecessarily so. They are both modified
Softrock Lite II SDR's, the input stage being a FST3253 mux/demux. I
have been educated to understand these can take quite a beating
without damage. This probably explains why neither receiver blew up
last winter! I'm going to trust that will continue to be the case
and leave my system alone. I am not worried about strong RF from
other sources causing the diodes to conduct and make trash while I
am receiving. I have no local LF/MF operators (yet), and I use
bandpass filters before the preamps to knock down strong out of band
signals. This might not work so well on LF in Europe where you have
monsters like DCF39 and the LW BC stations to deal with! I am not
too worried about radiating harmonics generated by the diodes since
the didoes are inboard of the band pass filters which will attenuate
the second harmonic more than 60 dB before it can reach the receive
antenna. The receive antenna would be a most inefficient radiator.

73,
Paul N1BUG