[Amps] Pin Diodes

[W0UN -- John Brosnahan]

At 07:38 AM 3/8/2006, you wrote:

>>I do quibble with your statement on speeds.   Vacuum relay speeds 
>>are measured in
>>milli-second switching times whereas PIN diodes can be switched in 
>>micro-seconds.
>
>John,
>
>I agree with your statement.  I don't mean that all other factors 
>being equal, a vacuum relay can approach the microsecond switching 
>speeds of the PIN diode.  But as a practical matter as used in HF 
>amps, I think one would be hard-pressed to find a condition where 
>the QSK performance of the relay is inferior to that of the PIN 
>diode.  I don't contest, so the multi-million operations that you 
>guys see could certainly be a long-term factor.  But even that 
>limitation is the result of the hardware mechanics and not the 
>result of premature failure from hot-switching.  A well-designed QSK 
>circuit like Rich Measures' can avoid hot-switching at pretty much 
>any CW switching speed.  Although, high-speed data mode switching 
>may be a different matter.
>
>Using my IC-7800, I can run CW as fast as my WinKey chip will go - 
>perhaps 75 or 80 WPM and the 77Dx and 70V completely keep up when 
>I'm watching the waveform on a scope.  Moreover, I can hear between 
>keyed elements up to about 35WPM on IC-7800, but the T/R turnaround 
>time is limited by the transceiver and not the amp.
>
>That said, I did try Tom Rauch's QSK-5 a few months back.  It was 
>going to be used as part a silent T/R system for a Drake C 
>line.  While testing the QSK-5, I was very impressed with its 
>operation and fast switching time.
>
>At the end of the day, I think one has to make a decision as to 
>which evil is the lesser one to live with: (i) the mechanical 
>failure of the vacuum relay; or (ii) electrical failure of the PIN 
>diode resulting from lightning, SWR greater than 2:1, etc.  And, 
>when either of these switches does fail, which is less costly and 
>easier to replace, especially in the middle of a contest weekend?
>
>If I was a contester with predictable antennas, I would probably 
>want a PIN diode switch.  Since I am not a contester and some of may 
>antennas (e.g., G5RV) may present a 3:1 SWR, which is easily 
>compensated for by the amp's Pi-L network, I prefer a vacuum relay.
>
>Paul, W9AC

Hi, Paul--

I love how my 781s and 87As play together.  On CW we can hear the 
back-scatter echoes
and turn the antennas as appropriate.  And even on SSB with PTT you 
can hear a slightly
different "kerchunk" at the end of a transmission that indicates 
back-scatter.  Something
that you will never hear with typical VOX delays.

I am building some monoband amps and may use a different style of TR switch for
these narrow-band amps. One with either 1/4 wave coax lines or lumped-element
equivalents.  Something I have been doing since 1978 in my HF/VHF radars.  The
1N4007 is a PIN structure, and I think there are others, since the A6A was also
mentioned as a candidate -- not sure of its package but there was also another
diode used for 2.66 MHz sounding with similar nomenclature that was 
in a white,
ceramic package that was about a 1/4 inch diameter sphere, and this one
seemed to work well.  Anyway, I want to keep it CHEAP so I hope the 4007s
will work -- just need to do some carrier lifetime measurements "one 
of these days".

Alpha limited their PIN diode protection circuitry to a reflected 
power of 300 watts
(rather than a VSWR) to make their warranty policy work out without 
undue expense.
But Dick Ehrhorn's 87As have been running with a limit of 400 watts 
for some time
(a decade maybe) with no ill effects.  So the design is a bit 
conservative as it should
be in any product.

The PIN diode vs. vacuum relay issue is a classic engineering tradeoff that can
go either way depending on what factors you choose to consider.  Much like
solid state amps vs. vacuum tube amps.  Depending on what questions you ask
either one can be the "right" answer.

73  John  W0UN