[Amps] Ten-Tec Centurion

[Ian White, G3SEK]

R. Measures wrote:
>> Assuming the spike is at the leading edge of the transmitted signal, 
>>then an
>> rf sensing circuit could delay the turn on of the rf amplifier until 
>>after
>> the passing of the spike.
>
>Colin -- The problem is that incoming RF does not go on a mini-vacation 
>immediately after the spike, it merely backs down to the ALC level that 
>is set.  Thus, there is RF on the closing NO contacts while they are 
>bouncing.  This results in hot-switching and current-transients.
>

Also there are two sets of relay contacts, at input and output. These 
contacts will bounce, and even if they are on the same DPCO relay, they 
will not bounce exactly together. This means the PA can have some 
exciting moments when the input relay is closed but the output contacts 
are not. The PA then has full drive but no load, which could lead to 
very high voltages across the tank components.

All of that can happen without any kind of oscillation, either at the 
signal frequency or parasitically at VHF.


>>  If that was the case, then the arcing from the
>> leading edge spike might not occur.  According to the facts given us 
>>by Tom,
>> he did not experience any arcs while the amplifier was rf sensing. 
>>That
>> would be one explanation.  It might be the most logical one, too.  It 
>>is a
>> simpler, less convoluted approach than parasitics.
>
>Measuring the actual resistance of the parasitic suppressor resistors 
>and eye-balling their appearance could eliminate parasitics as a 
>possible scenario in under 2-minutes - if the soldering-iron is hot. 
>Not measuring R-supp will not.

Fair point...

I certainly agree about the root cause of the problem, though: 
RF-activated switching will *always* be hot-switching. (And isn't it 
illegal under the anti-CB-amplifier rules?)


-- 
73 from Ian G3SEK         'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek