Isolated RF switching

Jan 05, 2026 Last reply: 6 months ago 30 Replies

Mercury wetted reeds are good for 100 million operations, and the degradation is a gradual increase in contact resistance.That's 138 hours of continuous operation at 200Hz, and the reed capsule isn't all that expensive or hard to swap.

A variable delay in switching might upset the null, but the delay is essentially controlled by the mass of the reed, the distance it has to travel and the magnetic force which is getting it to move, all of which should be pretty consistent.

Each dipole will have to have a waterproof connecting box at the centre point, so a few RF electronic components could easily be contained inside it. The switching connections then can be made at 200 c/s by ordinary twin flex, with no need for co-ax and all the standing-wave and phasing problems it would cause because of the variable terminating impedance.

With faint signals that are almost in the noise level off the front of a directional array, swinging it to find a null just loses the signals completely.

The phase reference is in the receiver, the same 200c/s multivibrator operating the aerial switching and the synchronous detector simultaneously. The incoming signal is continuous and unaware of what the receiving station is doing.

This is the opposite way around from the Lorenz system, where it is the transmitter which is switched from side to side with unequal mark/space ratio and the pilot 'synchronises' his brain to hear signals as dots or dashes if he is off the centre of the beam.

I want one aerial that can be swung on a rotator and used for direction finding, there is no advantage to having two of them if a single piece of hardware can be made to do the job with the addition of just one extra reflector.

At 200 c/s switching frequency, 1N4148 sounds far more attractive - eapecially as I have a drawer full of them.

I don't need the good on/off properties of a reed relay in this appication and I certainly don't want to have to replace any of the components after 138 hours of operation.

The whole purpose of the system is to null the synchronous detector, so anything that upsets it or needs calibration to start with (or after a change of components) is unsuitable.

Reed relays are not the answer to this particular problem.

We SED had a similar discussion in October 2024 titled "IR detector system, biasing of photo diode". This was with laser IR beams.

Joe

Well, the ADRF5024 costs us $108 in quantity.

Take a look at the TS3USB30. It's a damned fine fast DPDT switch for

20 cents.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

The Ramsey DF1 Doppler direction finder works by switching an array of four dipole antennas into and out of the receiver circuit, using

1N4148 diodes as switches, at a rate of 1 kHz.

I haven't heard of any of these wearing out :-)

The DF1 works fine at 2-meter frequencies, and I can recall using in the 70 cm band as well.

I wondered if I might be re-inventing the wheel but couldn't find an exact equivalent already in existence. There are three questions with this system:

1) Is it likely to work? I think you have answered that with a "Yes" - as long as I can engineer it correctly. 2) If it works, what use is it. I envisige it being used for DF-ing faint signals. 3) Who would want to DF faint signals? I don't know the answer to that, so it looks as though I have invented a solution that is looking for a problem.

Perhaps it can alert a ham to hit the dirt when it detects an incoming FPV drone crossing a battle line?

Thank you guys for another interesting thread. The manual for the DF-1 mentioned by Dave is available at:

formatting link

Perhaps it can alert a ham to hit the dirt when it detects an incoming FPV drone crossing a battle line?

Thank you guys for another interesting thread. The manual for the DF-1 mentioned by Dave is available at:

formatting link

Addendum:

AI acts as an acceptable, although now and again fallible, search engine front end. A little prodding can increase its precision. For instance, although the above archive link contains the phrase "schematic diagram" the diagram itself is blank. With a little prodding, AI returns a better result:

formatting link

Not only does n0zyc provide the schematic missing from the earlier link, he embellishes it. He enumerates its errors, adds corrections, and presents an improved design.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required