[TowerTalk] RF Relay QUestion

Jim Lux jimlux at earthlink.net
Sun Dec 16 14:28:05 EST 2007



-----Original Message-----
>From: Tim Duffy K3LR <k3lr at k3lr.com>
>Sent: Dec 16, 2007 9:53 AM
>To: 'Mike' <k4gmh at arrl.net>, 'David Gilbert' <xdavid at cis-broadband.com>, "'Dr. James C. Garland'" <4cx250b at muohio.edu>
>Cc: TowerTalk at contesting.com
>Subject: Re: [TowerTalk] RF Relay QUestion
>
>Great work Mike!
>
>I looked at the relay specs and they look like they would work great for
>1500 watts on all HF bands for an RF antenna switching application.

The other thing to look at it (depending on the application) is the isolation.  Some AC power relays don't have a lot of RF isolation when open, because the capacitance between the contacts is fairly high.

In numbers.. say you're looking for 40 dB isolation on 10m in a T/R application (so that your kilowatt is down to 100mW into your receiver).  That's a voltage ratio of 100:1.  You'd want the impedance on the "off" side to be at least 5000 ohms (i.e. 100 times the input impedance of your receiver).  

5000=1/(2E-12*pi*28E6*CpF)

C= 1/(5E3*2E-12*3.14*28E6) = 1/(31.4*28E-3) = 1.13 pF

That's not real huge (even if I screwed up by a factor of 10 somewhere)

Of course, in the receiver case, you could use a second relay to ground the receiver input.. the impedance bump in the mainline path will be negligible.

The point is that there's more to RF relay usage than whether the contacts will carry the current.

Also, remember that at RF, all the current is being carried in the surface of the contact (skin effect), so the 60 Hz current rating may or may not be applicable.  

Testing is the key.. Get one, hook it up to a dummy load and put the spurs to the amplifier and see if it melts.

Breakdown voltage is probably the least of your worries, unless there's some weird capacitive coupling to something with tiny clearances.  But even then, if the voltage is below about 300V, it can't spark over in air.  

Even in a well matched system, you could exceed that 300V at legal limit (1500W ->  274 Vrms in 50 ohms -> 383Vpk )

Jim, W6RMK



More information about the TowerTalk mailing list