[sci.electronics] How to switch high-voltage RF?

jfw@ksr.com (John F. Woods) (05/20/91)

Having recently seen an advertisement for a "smart" antenna tuner, I
became curious about how they switch the inductances and capacitances
in and out of the circuit.  Since the smart tuner claims to be good
for 300 watts, it would seem that they need switches that can
withstand several hundred volts when open, but the case seemed a bit
small for 11 high-voltage relays ("64 input capacitances, 32 output
capacitances, and 256 values of inductance"; 11 binary switches).
So, what are they likely to be using?

mark@ve6mgs.uucp (Mark Salyzyn) (05/20/91)

In <3636@ksr.com> It was said:
>Having recently seen an advertisement for a "smart" antenna tuner, I
>became curious about how they switch the inductances and capacitances
>in and out of the circuit.  Since the smart tuner claims to be good
>for 300 watts, it would seem that they need switches that can
>withstand several hundred volts when open, but the case seemed a bit
>small for 11 high-voltage relays ("64 input capacitances, 32 output
>capacitances, and 256 values of inductance"; 11 binary switches).
>So, what are they likely to be using?
I have the `smart' tuner in question. Regular relays (well, higher quality
that Radio Shack) are used. No Gold or Silver contacts. No special
precautions on the input side of the Pi tuning network. The middle `L'
of the tuning circuit has the inductors shorted by the contacts, most are
toriod, but all, I am sure, have no mutual inductance. The output caps,
however, have TWO separate relays with their contacts in series. This is
where the voltage can get rather high.

Have Fun. 73 de VE6MGS/Mark -sk-

tomb@hplsla.HP.COM (Tom Bruhns) (05/23/91)

The question about what sort of switching was used in the "smart" tuner
reminded me of working with a '419T' antenna tuner from Collins; it was
a mate to a '618T' HF transceiver used on aircraft (both commercial and
military).  It has vacuum variable caps and 'slider' inductors (kinda
like roller inductors, but the slider was driven by a motor shaft that
stuck up in the middle of the inductor).  Relays were small vacuum guys.
This thing, as I recall, guaranteed to tune in five seconds, but we knew
that if it took over two seconds, something was likely wrong.  One
of the failure modes was operation without an antenna connected.  It
tried to tune its output terminal, and generally ended up creating a
very high voltage node somewhere, one that would arc over.  It could
usually tune a 5" piece of wire in the mid-HF range.  --  But maybe
there's a lesson here about high RF voltages in tuners:  they can cause
nasty breakdowns, so it's best to avoid trying to tune things that will
be outside the range the tuner can handle.

khoward@chook.ua.oz (Ken Howard) (05/23/91)

From article <5170137@hplsla.HP.COM>, by tomb@hplsla.HP.COM (Tom Bruhns):
Writes: (concerning automatic antenna tuners)

>........
> that if it took over two seconds, something was likely wrong.  One
> of the failure modes was operation without an antenna connected.  It
> tried to tune its output terminal, and generally ended up creating a
> very high voltage node somewhere, one that would arc over.  It could
> usually tune a 5" piece of wire in the mid-HF range.  --  But maybe
> there's a lesson here about high RF voltages in tuners:  they can cause
> nasty breakdowns, so it's best to avoid trying to tune things that will
> be outside the range the tuner can handle.

Probably a good place for another timely warning here. I spent a while
designing an automatic antenna tuner for use with HF linear amps up to
2KW for an Australian communications comapany called Codan. During the
development work we made the painful discovery that if you tune for best
SWR there can be two possible modes: a) maximum power into the antenna
and b) maximum power dissipated in the tuner ... fun at 2KW!!!
Solution: course tune for max antenna CURRENT, fine tune for best SWR!


			---Ken---

Ken Howard, University of Adelaide         |      Phone:  +61 8 228 5198
	    Department of Computer Science |      Telex:  UNIVAD AA 89141
	    G.P.O. Box 498, Adelaide 5001. |      Fax:    +61 8 223 1206
	    AUSTRALIA.                     |      ACSnet: khoward@cs.ua.oz