[Top] [All Lists]

Re: [Amps] Bandswicth failure

To: "Tom W8JI" <>
Subject: Re: [Amps] Bandswicth failure
From: R L Measures <>
Date: Thu, 29 Jun 2006 04:49:03 -0700
List-post: <>
On Jun 29, 2006, at 2:35 AM, Tom W8JI wrote:

> Joe wrote:
>>> However if the amplifier is underloaded, voltage in the
>>> output
>>> "tank" will increase.  The "limit" for underloading is
>>> operating
>>> into an open circuit (T/R relay not closed, etc.).  If
>>> the bandswitch
>>> is operating close to the limit already (as is the case
>>> in some
>>> amplifiers), overdriving an underloaded amplifier can
>>> damage the
>>> bandswitch.
> Rich replied:
>> Depends on whether the Tune-C or the bandswitch has the
>> lower BDV.
>> If the Tune-C arcs at 3600v and the bandswitch arcs at
>> 5000v
>> (SB-220), the Tune-C protects the bandswitch.
> Another technical misstatement.
> The SB-220 switch is from the same Centralab family used in
> the lower voltage Ameritron amps, Dentron amps, and many
> other places.


> Like any switch, it has a variety of breakdown
> voltages depending on what contact to what other contact and
> even how the soldering is done on the switch.

.  .  I'm semi-obviously  talking about the stationary open contacts  
that are the most prone to arc during the 110MHz oscillation -- i.  
e., , 10m, 15m. and, to a lesser extent, the 20m fixed  L contacts.   
At sea-level, when open, the measured breakdown of these contacts  is  
c. 5000Vdc.

> Like any
> system with high voltages, the shape of conductor surface
> where it extends and everything around it is nearly as
> important as the spacing for determining breakdown voltage.

  The shape of the contacts varied during SB-220.
> The safe reliable upper limit in that switch, even when the
> switch is optimally designed and connections have careful
> lead dress and soldering, is just over 3000 volts.

More smoke.  Exceeding the actual BDV of c. 5000 (sea level) is what  
eats the switch. Also, the arcing problem takes place at the springy  
end of the contact, not at the solder lug end of it, so solder can  
not be a factor unless the constructor does not know which end of the  
contact is which.

> I know that from months of actual testing in a lab

What's the difference between testing and actual testing?

> with new
> switches. It's actually a race to see if the tune cap or
> bandswitch fails first with excessive voltage.

Months?  I'm about as slow as humans go, and it probably took me  
under 15-minutes to measure the BDV of a SB-220's - and a TL-922's -  
requisite bandswitch contacts and Tune-C.
> ...
cheerz, Tom

R L MEASURES, AG6K. 805-386-3734

Amps mailing list

<Prev in Thread] Current Thread [Next in Thread>