SB220 Arcing

W8JITom at W8JITom at
Tue Apr 9 00:45:35 EDT 1996

Subj:	Re: 220 arcing
Date:	04/08/96

Hi Tom,
I hear this a lot, and I think it's of general interest. 
In a message dated 96-04-08 20:02:28 EDT, you write:

>The main problems are parasitic oscillation/arcing when I tune up, unless I
>remember to start with the loading cap half-meshed.  It is most pronounced
>on 80 and 15 meters...I was shocked the first time it 'spit' at me on 80 

Everyone now tends to connect arcing with parasitics. Although under rare
conditions that can occur, it's a very rare case. The reason you still have
this problem after installing the parasitic mod is the problem is almost
NEVER caused by parasitics.

When an amplifier operating in class AB through C is lightly loaded, the peak
anode and tank voltage can be * several times * the anode supply voltage.
This is caused by the loaded Q of the tank circuit increasing, and the tube
being driven into and out of saturation, by high drive power and light anode
loading. The effect is actually caused by the tube "ringing" the tank at the
operating frequency, not at VHF.

Let's look at it from the standpoint of an unwanted VHF signal. If the cause
was a "VHF parasitic", the tank capacitor would present a VERY low impedance
path. If the tank capacitor was adjusted to have 300 ohms of reactance at 14
MHz, it would have ~40 ohms of reactance at mid-VHF. You can calculate the
VHF current needed to drive the capacitor to ~4000 volts at that value of
reactance. It's getting close to a short circuit for VHF (after all, that's
why the amplifier tank has VHF harmonic supression).

You'll find the tubes totally incapable of supplying enough current into that
load impedance to cause an 4000 volt arc! In rare cases I've seen arcs from
HF, or near HF oscillations. But I've *never* seen tank arcing from VHF
oscillations in an HF amp. 

Here's another thing to consider. When a circuit oscillates, it's from the
overall circuit loss being less than the overall gain, and correct phase
shift. Circuits just don't "wildly oscillate" for a split second and then
quit. They either make good strong oscillators, or they don't

If circuits are at the edge of oscillating, they won't build up into an
uncontrolled high power oscillation. 

HF amps arc when mistuned because it's almost impossible, and certainly not
economical, to size tank components to handle extremely high voltages caused
by underloading the tank.

>I had sort-of gotten used to the 15m instability.  It will also randomly 
>start when 
>I first key-down -- is the circuit not  loaded heavily enough??

Absolutely. The loading is too light. Not only that, MANY exciters put out a
sharp rising edge transcient before the ALC reduces the drive. I've measured
Kenwoods that slap 150 or 200 watts into a PA for several RF cycles at the
moment of transmission. Imagine the peak voltages generated at the output
when this happens. Keep in mind the SB-220 was designed to be "tuned up" in
the CW position, and operated at 600 watts output on CW (PEP) and 1200 watts
on SSB (PEP). They weren't designed to be tuned in the SSB position.

>OK, so if I slap the grids directly to ground, what kind of parasitic choke 
>you recommend?  I had previously used a 3 turn copper strip (abt. 1/4 inch
>wide) with a pair of 100 ohm resistors (before I installed Rich Measure's 
>that includes the nichrome wire/resistors/coils in the anode).

The stock one is fine, actually it is a bit of "overkill" when the grids are
grounded. Just be sure the resistors are true carbon comps, and not
constructed with a spiral carbon or metal film inside. The reason you have
arcing, and still have it after all the mods, is the problem wasn't
parasitics in the first place. Tuning the amp incorrectly, especially tuning
it in the SSB position, and having an exciter that pulses the amp with high
drive levels, is what really causes tank arcing.

A secondary cause is relay sequencing, if the relay's output closes late
(after the input closes), the amp will operate at full drive with NO load for
a second.  

73 Tom

More information about the CQ-Contest mailing list