[TowerTalk] Re: [FCG] Relays in RF usage

[Jim Lux]

At 03:13 PM 8/12/2003 -0500, Didier Juges wrote:
>At 11:29 PM 8/11/2003 +0000, you wrote:
> >I ask:
> >
> >1  Can RF and AC be equivalent?  I think not.
>
>RF is AC. However, at high frequencies (RF), skin effects tend to make life
>harder on contacts. At HF (<30MHz), skin effect is probably minimal and not
>a real concern.

Even at 100 kHz, skin effect is very significant! and at HF it's a real big 
deal: this is why copperweld wire works as well as solid copper 
(electrically).
Skin depth in copper at 100 MHz is  0.26 mils(.0066 mm)
at 30 MHz it's about 0.47 mil (0.012 mm)
at 10 MHz it is 0.82 mil (0.02 mm )
a 1 MHz it's about 2.6 mils (.066 mm)


skin depth scales as 1/sqrt(f)


> >I say, put in a BIG ONE and if it burns out.... it was too small.  AMATEUR
> >radio, after all.... 73,
>
>That will probably work well at HF, but at VHF and higher, the size of the
>relay will actually play against you, as the inductance of the contacts
>will cause VSWR. A smaller device will probably have lower inductance and
>be faster, which could be necessary to avoid transmitting into an open
>before the relay closes.

Parasitic C from the wiring would also be an issue..



>I am amazed at the relatively small size of the relays used in modern
>transceivers to switch the antenna and/or the tuner in and out of circuit.
>Those I have seen (in my TS-440SAT for instance) look like 2 or 3A AC
>service types if I recall. At 100W and with a 3:1 antenna VSWR, the current
>will still be well under 3A, so I guess it's OK, assuming no additional
>derating for RF currents versus 60 cycles.

Hopefully, too, the mfr has actually tested the particular relay at RF 
rather than blindly scaling from 60 Hz databook values.  If you're making a 
thousand radios, it's worth it to do so.

Jim, W6RMK