switching RF between filters how to?

Hello, I am trying to find a way to switch a low level RX signal through multiple HF (1-30MHz) bandpass filters (one each time), to select the appropriate band. the filters are bidirectional, so bidirectional switching is needed.
I would like it to have the minimum interference to the RX signal.

Some ways I have found:

1. multi selector switch. Big bulky but no power required and no interference to the signal.
2. DIP switch. Small, no power required, no interference, but need to individually select more than one switches.
3. relay switching. Nice control, but relatively big and consumes lots of power unless (memory) toggle relays are used.

4. Diode switching. I am affraid about the effect of the diodes in the signal path.
I post a suitable diode circuit here. Two of them are needed on each filter, one at each end, so I assume a 3db loss. It is not clear to me how a diode can leave RF pass on the opposite direction, when biased?
I would also like to discuss with you the circuit in figure 3 on this link http://urgentcomm.com/networks-amp-s...antenna-switch

Any other way/examples you could suggest?

the diode switch will cause IPs, We used 5 X HG 5007 hot carrier diodes to get the IPs @1V Rf !, below -120 dB. Our one worked in a circuit impedance of 200 ohms. They also take a lot of current, unlike the circuit you showed.
Frank

how about this circuit? ../imgqa/eboard/Antenna/rf-s32yhzynpui.jpg
It uses the diode to short circuit the signal when needed, rather than passing signal through the diode.

Will it still generate harmonic distortion?

The circuit you show is OK for FM, but its IP performance will be poor AND it relies on the quarter wave line to get the RX diode to reflect back an open circuit to the TX PA.
Frank

Would you mind to give a specification (maximum power, impedance,...)?

Hello, I do not have impedance specs, I assume my switched filters are 50R. I attach you a picture. The signal levels are low, single transistor mixer signal levels. So you can get an idea.
The system is a simple qrp ssb transceiver.

- - - Updated - - -

A crude setup I was thinking, was that of a single shunt diode to short circuit the rf signal, when needed....?

For lowest power consumption, I would refer to analog switches. There's a number of low voltage switches, e.g. from Analog devices with low series resistance and cut-off frequency of at least several 100 MHz. Switch capacitance can be possibly compensated in filter design. GaAs RF switches can also an option, they have lowest loss and lower. capacitance than CMOS switches.

Since the filters have DC isolated inputs and putputs, why couldn't one use a single shunt diode at each end, to short circuit the rf signal, when needed?