[Home][2018 Index] [TCML] Ferroresonance, was My first ballast results [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[TCML] Ferroresonance, was My first ballast results

If you already have a working air cored ballast, you can freely experiment
with adding more reactance in series by any type of inductor. The current
will always be limited to the value set by the air core inductor, even if
the added inductor saturates because of an underdimensioned iron core. The
worst thing that could happen would be that it overheats. 

But my general impression is that many coilers use ballasts with marginal or
too small cores. If the core saturates, that is a very bad thing. You lose
control over the current, and bad resonances could appear with the tank cap,
if no gap fires so that the energy is removed. 

These types of resonances are well known and researched by the power
distribution engineers, google "ferroresonance" and a lot of articles comes
up. If a lightly loaded distribution transformer is connected to a
capacitive load, such as long underground power lines, an unlinear resonance
condition is sometimes set up, with very destructive results. The
transformer could be destroyed because the saturated core overheats, or
overvoltage could destroy it. This resonance condition is highly unlinear or
even chaotic, as the inductance has different values for different degrees
of core saturation, so an oscillation of any frequency could be set up, or
even completely chaotic oscillations. In power distribution systems, this
condition might occur if the mainly resistive normal load is suddenly
removed because of a fault condition, and only the capacitive load of the
distribution cables remain. The fault could cause the transformer core to
saturate momentarily, and that starts the ferroresonance condition, which
then continues until the energy is removed by MOV overvoltage protection or
remaining resistive load. 

This type of resonance is not limited to the power line frequency (or the
RSG frequency in TC use), it starts by a transient condition and resonates
at any or several frequencies, or at completely chaotic frequencies,
regardless of the steady state resonance frequency that is given by the
capacitance and the non saturated inductance of the circuit. 

I think this condition could be highly relevant to to a saturated TC ballast
loaded by the tank cap, if the resonant energy is not promptly removed by
the spark gap or safety gap. I think this underlines the importance of a
safety gap, and why a saturated ballast core is a very bad idea. Perhaps
this explains why the series resistors of a Terry filter saves the fragile
neon transformers - the resistors removes the energy from the oscillating
system if the core saturates. 

Perhaps it even explains why many coilers consider a variac as a necessary
thing to provide a soft start for the coil. Of course we want to control the
kVA input to the coil, but this could be done just as well by controlling
only the current, by a ballast with taps, or by a variable welding
transformer as ballast. If the ballast core does not saturate, it will all
by itself limit the inrush current to the transformer, no variac will be
needed for that purpose. No inrush current at start up means no transient
condition that could otherwise set up a ferroresonant condition. Many
prominent coilers in the past have recommended some resistive ballast along
with the inductive ballast, to avoid an unstable condition when regulating
power to their coils. And, of course, the use of a variac for a soft start.
Perhaps all this would have been unnecessary, if the ballast had been
designed not to saturate under any operating condition?

Do I have any proof that ferroresonance occurs in a TC system? No, but
"thumping sounds" or other irregular behavior of inductive ballasts have
been described on several occasions on this list, and coilers have
recommended a touch of resistive ballasting along with the inductive, to
avoid unstable operating conditions. That is exactely how ferroresonance
would appear, if it exists in a TC circuit. 

Is it possible to design an inductive ballast that never saturates? Yes it
is, just make the ballast core larger than the transformer core. I have 55
kg of iron in my HV transformer, and 130 kg core laminations in my ballast.
With a 12 mm air gap (or actually twice that considering the magnetic field
passes the gap twice) I can just throw the switch with the ballast tap set
for 50 amps and the transformer shorted on the secondary, without tripping
the 40 A mains fuse. And I have no variac. 

Stockholm, Sweden

Tyler and David,
Yes, as much as I hate to give up more valuable floor space in the shop to
another welder, it seems inevitable. I will get one someday, but I am
waiting for something priced so right, I can't say NO. So in the meantime,
I just need to scrounge up an additional 3.4 mH of inductance using stuff
around the shop. Hence the old NST core idea, which is not heavy, does not
take floor space, and has no additional cost involved.
~Dan (streamers on a budget)
Kansas City area

Tesla mailing list