Here's an unusual problem I've never encountered before, which I offer in
case you ever run into it. I have an external power supply which provides HV
and 120VAC to an amplifier RF deck. The power supply runs off 240VAC and is
protected with a dual 25A circuit breaker. The 120VAC (which is one leg of
the 240VAC line) to the RF deck powers the blower, filaments and bias
supply and is fused in the RF deck. The fuse is 3 Amps, since the current
draw for these circuits is low.
The 120VAC line is switched by a SPDT relay (rated at 12 Amps), operated by
the on/off switch on the RF deck. (There is a separate switch that turns on
the HV.) The circuit is very simple and seemed to work properly, until I
powered OFF the amplifier. (The HV was already off when I did this.). When I
pressed the Power off button, there was a pop, the 25 Amp circuit breaker
tripped, and subsequently the 120VAC line to the RF deck was dead. I checked
the 3 Amp fuse in the 120VAC line and it was fine.
The SPDT relay is wired so that the load (blower, filaments, bias) is on the
Common pin, the 120VAC feed is on the NO pin, and the NC pin is grounded.
Thus, the amplifier 120VAC circuits are grounded through the relay's NC pin
when the amp is off, and connect to 120VAC when the amp is on. At first
blush, it seemed like a foolproof circuit and it was hard to imagine how
anything could be wrong.
The mystery seemed to be the tripped 25 Amp circuit breaker. I couldn't see
anything that could draw that amount of current from the 120VAC circuits,
especially since the 3 Amp fuse didn't blow. My first thought was that a
switching transient (e.g., an inductive back-EMF) from the fan or filament
transformer was tripping the circuit breaker, but I quickly ruled out that
explanation. True, there is an inductive spike, but it's not large and much
smaller than the spike from the HV transformer. Also, I'm using a magnetic
circuit breaker, which only responds to current and is relatively immune to
voltage spikes.
Eventually, I traced the explanation to the SPDT relay wiring. The relay is
a PCB mount power relay, which has a short travel for the Common contact as
it moves from the NC to the NO contacts. When I powered off the amp, thus
breaking the NO connection, there was in fact a slight spark between the
moving relay contact and the NO contact, resulting from the fan back EMF.
The spark was only a few milliamps of current, but it was enough to create a
momentary ionized path between the NO contact (which is connected to
120VAC ) and the moving common contact.(connected to the fan). The moving
contact carried the ionized path from the NO pin over to the grounded NC
pin, thus providing a conducting plasma which shorted the 120VAC line to
ground and blowing the circuit breaker.
The fix was very easy. I replaced the SPDT relay (which I cut open to verify
that the contacts were indeed melted and fused together) with a new one, but
clipped off the NC pin. Without an NC pin, there is no current path to
ground for the 120VAC line, and all is well.
What's interesting to me about this fault is that a modest spark of only a
few mA and a duration of a couple of milliseconds, could create a plasma
which initiates a massive sustained current flow in the relay. In effect,
the relay was acting like a mechanical SCR, whereas a small "control"
current (i.e., the spark) triggered a much larger current. The large
current, once initiatied, is self-sustaining, just as the current in an
SCR.
73,
Jim W8ZR
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|