km1h@juno.com wrote:
> >>Get the switch from Ameritron...it is a PnP replacement.
> >>
> >>Then quit underloading the amp; that is the major cause of zapped
> >>switches. Load it up to maximum obtainable output with 100W of drive.
> >>Then back the drive ONLY down until you have ~ 800W. Do not reduce
> >power
> >>with the amp controls.
This is good advice. Parasitics are absolutely NOT the cause of switch
arcing. This is quite easy to understand, if you look at the tank layout
and understand ohms law.
The anode of the tube is shunted by a few ohms of capoacitive reactance.
It would take many dozens of kilowatts of energy to develop enough
voltage across the low Xc to produce voltages causing an arc.
At HF, the situation is quite different. The tank capacitor and tank for
an energy storage system. The tube is a time varying resiustance that
gives the tank a "tug" at exactly the right instant, and if nothing
removes the energy form the tank it builds and builds. Eventually
something absorbs the energy or something arcs.
None of this has anything to do with it being a pi-network, or an L
network. As Carl pointed out it only has to do with the unloaded Q of
the network. Load the PA too lightly for the amount of drive applied to
the tube, and you can get in tank arcing trouble.
The SB-1000 band switch, when properly installed, will handle normal
operating voltages. Unfortunately several things went wrong in the
SB1000. Heath decided to NOT install an anti-corona washer. That reduced
the voltage breakdown. Heath also used a transformer that did not have
the buck-boost feature, and used a transformer with poorer HV regulation
than the AL-80A.
Finally, no one inspects the assemblers connections to be sure they are
smooth round connections with no sharp point, and the lead dress is
proper.
ALL of this contributes to switch failures in the SB-1000, earning it a
reputation for "eating switches" while the almost identical AL-80A has a
very good switch service history.
The actual test voltage specified for critical switch wafer contacts is
3750 volts peak. This voltage was selected to put the switch just beyond
the voltage rating of the tuning capacitor, since the capacitor can take
a momentary arc without damage. Normal properly loaded operating voltage
is 2500 volts peak. But if you unload the PA, and drive it with full
power, peak tank voltage is limited only by the voltage breakdown of the
tube or tank components. Voltage can easily reach 3500 volts or more.
None of this stuff is magic or rocket science. It's all common sense and
normal engineering, and can be verified by traditional math. I've
measured PA's producing over ten times the normal operating voltage in
the tank during load faults.
73 Tom
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|