[AMPS] Shorted Pi-Networks Turns revisited

[Tom Rauch]

> 
> Most home built amps use large switches that have a high enough voltage
> rating to prevent any problem - so any switch could be used.  The question
> may be more difficult to answer in the SB-220.  It used a receiving type
> band switch wafer.  So, by using a non-progressive shorting switch to
> replace the original progressive shorting switch, the question arose
> whether the voltage rating of the switch would be exceeded.  I did replace
> the original switch (which had failed) with a non progressive shorting
> switch and it has worked fine.  The answer may be that I was lucky.

The SB-220 used a series 300 switch, the same as most other 
commercial amplifiers with supply voltages less than 3000 volts.
The actual rating of that switch is 2.25 amperes@28VDC while 
switching under power.

The current carrying capacity is 12 amperes with the contacts 
used, and the rated dielectric breakdown is 1500VAC. That's hardly 
a "receiving switch".

The breakdown rating was "upped" by positioning the contacts with 
highest voltage away from grounded switch sections. A new switch 
stands off about 4 kV peak from the ten meter contact to all other 
contacts.      
 
> For commercial manufacturers, the question may be more important than for
> homebrewers.  For Heath to have used a larger bandswitch may have been
> tens of dollars increase in the cost of manufacture and could have
> represented the difference between making a profit or losing money.

The next switch up is a series 230 switch. The wholesale price was about 3 times as much.

> Otherwise, you are correct that the volts per turn in the Pi section would
> be less than in the Pi-L.

There are many cases where that statement would not be true, but 
that isn't where the arc occurs anyway. It is a moot point.
 
> For homebuilders, it is simple.  Use a big bandswitch with lots of voltage
> and current and it will not fail.  For commercial manufacturers, either
> you need to put a lot more engineering into the design or take the
> Microsoft approach and design an amplifier quickly and put it out into the
> field to see what fails.  After the failure points are determined, come
> out with an improved model for more money to solve the original weak
> points.  And good engineering will then build newer problems into the
> "improved" amp requiring an ever further improved model.

If you wire the contacts without some thought as to voltage 
differences, and especially if you don't short a portion of the unused 
coil, you can make a large switch fail in a very small PA.
 

73, Tom W8JI
w8ji@contesting.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com