The AM Forum

I have a 2-position MFJ remote coaxial switch, but I'm nervous about putting large amounts of power through it. I built this pair and it should be able to handle anything I'll use.

Jon

How much eff3ct does the relay have on SWR?

Nothing on 75 meters, but the SWR on 20 meters went from nil to 1.6:1

It is an increase, but I’d rather have my piece of crap inline rather than an MFJ piece of crap. The plan was to put two 4700pf doorkobs on each end, but I could only fit one. That probably contributes to the increased SWR.

Jon

I think that the increase of the SWR at 20 is due to the ground system you used Jon. not so much for the relay.
If you put first a piece of PCB material in the cover attached by the ground screws of the connectors, than you have a solid ground plane, and it is easy to work with, much easier than all that copper. Run the cables close to the ground plane. It is the cable inductance that brings the SWR up, not so much the capacitance of the relay.
I had bad experiences with car relays at RF, I burned the isolation material. Perhaps the modern ones have a kind of plastic that is better than the paper / fiber that had my relay

That’s why I added the flat copper piece to the outside unit. It probably isn’t sufficient. 75 meters is flat and the SWR goes up only a little at 20 meters, so I’m not too concerned. I don’t plan to operate on any higher bands at home anyway.

I’ll blast some high power into it in the next few days and see if anything blows up. I might even modulate the carrier this time... oh - my - god

Jon

Sorry, what is the purpose of the capacitors??

What sort of relay is in there?

Can't quite make heads nor tails of what is going on.

If I was interposing a relay between some SO-239 connectors, my thinking would be to run a
copper strip of about 1"or more width between them on opposite sides of the box, and then
run straight 14ga solid wire from each center to the relay contacts... in the case of two connectors
on one side, a similar arrangement. Or take one contact off to a SO-239 at a right angle...

Most direct path... shortest.

Btw, copper or PCB material makes a very good "box". Just solder up the seams...

       _-_-

The purpose of the doorknobs is to isolate the DC from the radio and antenna. I’m using a balun at the center of my dipole and it’s a short to DC.

My big concern at this point is how much power can a 40 amp car relay handle. I’m going to throw in 1kw peak right now to see how it does before I try any real power.

Jon

The RF rating of relays? good question! I think like you I try for overkill. Some relays have very flat contacts, some have obviously convex spherical contacts. Some have 'self-wiping" contacts. Lot of variables and little published information.

Looks perfect to me though the only way to find out is use it!

I use contactors and relays (but not a car relay yet..) and have not had 'serious' SWR issues on HF because of them but sometimes because of the wiring because of my mistakes -too long, sharp bends, etc. The relays themselves seem to have very little effect on the SWR and I have had no issues with arcing but most of those are insulated for 240VAC or 600V. Derate for RF to try to avoid arcing, right?

3   Watts into 50 Ohms is    17.3    peak volts, and    0.24   Amps
12   Watts into 50 Ohms is    34.6    peak volts, and    0.49   Amps
25   Watts into 50 Ohms is    50.0    peak volts, and    0.71   Amps
100   Watts into 50 Ohms is    100.0    peak volts, and    1.41   Amps
500   Watts into 50 Ohms is    223.6    peak volts, and    3.16   Amps
1000   Watts into 50 Ohms is    316.2    peak volts, and    4.47   Amps
1500   Watts into 50 Ohms is    387.3    peak volts, and    5.48   Amps
2000   Watts into 50 Ohms is    447.2    peak volts, and    6.32   Amps
3000   Watts into 50 Ohms is    547.7    peak volts, and    7.75   Amps
6000   Watts into 50 Ohms is    774.6    peak volts, and    10.95   Amps


I often wondered how one of those mercury contactors would work for KW level RF. Most older ones are just two pins that stick down into a glass tube, and the mercury rises to them when a plunger is pulled down by the solenoid's field.
Others have an electrode at each end, which might be less desirable because of the capacitance to the solenoid coil and the lower contact being immersed in mercury all the time.
Some are filled with an inert gas to try avoid arcing.
I have no experience with those for RF but they are darn rugged for AC power switching!
Has anyone ever tried one of those?

DC? what DC?
There's supposed to be a blocking cap at the output of ur rig??
Also RF choke to ground?

Maybe a static choke to ground a the bottom of the feed to the ant?

                          _-_-

True. My home-brew stuff has a cap in the output, but I'm leary of putting 12 volts on the antenna jack of a ricebox. 

Anyway, it's a moot point now since I removed it from my antenna line anyway.  I've switched to a tuner for everything (although resonant on 3870 KHz).

Jon

Ok, you lost me again!

What 12 volts on the antenna jack of a ricebox??

The control voltage only goes to the relay coil(s).
Are you referring to the control voltage?

I've been using an Ameritron RC4 for years. It seems to work fine. Ameritron is part of MFJ now, I believe. It's rated for 2500W but I run less than that. IIRC, it uses several paralleled 1kV disc caps for the RF path.

http://www.ameritron.com/Product.php?productid=RCS-4

I use the same with no issue. GORT doesn't bother it.

I got into an interesting conversation about this topic. A close friend says that he had a commercially built switch which ties all the ground connections of the 6 antennas connected, together. He claimed that all the wire antennas (grounded sections) distorted the antenna performance of his yagi. He rebuilt the switch to isolate both connections, so electricly they were not connected. I would think that these grounded sections migh act like a radial field. Thoughts on this?

GORT the robot?

Gort is my 813 AM rig named after the robot.

Jon,

I built a remote coaxial switch using an old MFJ-916B duplexer. I hollowed the guts out of it and squeezed a DPDT 12VDC relay inside with the poles paralleled. Fed the DC into it with a feed-through capacitor and a ferrite choke on the outside. 500 watts didn't kill it. I used it to switch a single antenna between the main operating position and the workshop, so it was never hot-switched.

The small housing size is a plus - if it can maintain a good SWR at UHF, then HF should be no problem. And it's completely shielded. You can buy a new MFJ-916B for about $30 on Ebay. I picked up a used one there for much less. Or you could buy three SO-239 connectors and a metal box for about that much, and still have some drilling and blasting to do. The way I looked at it, MFJ did the heavy-lifting. I just swapped the guts.

Don

Intriguing thought. The grounded side of a coax-fed antenna is usually connected to a whole bunch of stuff anyway, the possibly long feed line, the house ground system likely at the access panel but maybe way back at the shack. But it's not usually connected to say half of a dipole you aren't using, way up in the air.

I hope somebody who understands this better than I do can answer this.

Just use something like this. Much simpler. It'll  handle oodles of power.

Power good

Jon

...sorry, can't resist:

https://www.youtube.com/watch?v=FiEw1jcLztA (https://www.youtube.com/watch?v=FiEw1jcLztA)

That looks pretty complicated.