Magnetic amplifiers revisited
I recently found a copy of Magnetic-Amplifier Circuits: Basic Principles,
Characteristics and Applications by William A. Geyger at a local used book
shop. It said something I believe is noteworthy:
"At first, it is important to stress that the usual interleaving technique
adopted in conventional transformer-core construction, with a stack of thin
laminations of the required form, is not adequate for saturable reactors in
magnetic amplifiers, owing to the presence of air gaps at the joints.
Staggering of the joints does not overcome the fundamental difficulty,
because the effective air gap is still appreciable."
This reduces the sharpness of the magnetization curve. The author goes on
to recommend single piece stampings or tape wound toroid cores. Well this
inspired me to try using a small toroidal transformer as a mag-amp. Quite
simply put, it seems to work well! I used a toroidal transformer with about
1 sq. in core area and 120vac in and 30vac out. I wired one of the primary
leads to 120vac one to a 200va NST and the other lead of the NST to back to
120 vac. On the control side I used a large inductor (the primary of another
larger transformer) and a dc power supply and fed this through the
secondary. No dc in yielded itty bity sparks from the NST. An input of .25A
would give approximately 1/2 power out and .5A full power.
There other more complex arrangements for mag-amps which may give better
performance in some way or another but this simplistic arrangement works
I imagine variacs employ tapewound cores which would make them good
candidates for mag-amps. It would be a problem if they employ cut cores, but
I'm sure someone on the list has dissected a few and could shed some light
on this. Control windings could be wound over the original windings.