"John Byrns" wrote in message
...
In article ,
Randy or Sherry Guttery wrote:

Ken wrote:

So, what is the design method for a transformer that saturates easily?

Without getting into a bunch of formulae, etc. (which I'd probably screw
up anyway)... a couple of factors - 1) absolute minimum core to couple
the windings - i.e. magnetically "starved". 2) no gaps in the core - let
the DC current's field circulate well - such that it "interferes" with
the AC field. The core can only hold so much flux - if DC is "pushing"
the field one way -- the AC (when it opposes) is only going to "reduce"
it - not reverse it - (or not fully reverse it) so that the coupling
becomes very inefficient. If you look at most output transformers
designed for single-ended use - they have a gap in the core somewhere.
Obviously - such a gap would not be appropriate for a saturable reactor.

And that sets me pondering again whether the primary "effect" is bucking
or just reactance... Let's say for the moment that bucking is not the
primary mode - and reactance is. Then why the reversed phasing (if
bucking isn't a factor)?

Well - as I just noted - in a true saturable reactor - the DC flux
"overwhelms" the AC flux. Since the AC and DC are additive half the
time - and subtractive half the time - the control isn't going to be
symmetrical. This is overcome in "the real world" by twin reactors -
with the DC "reversed" through one (compared to the other). This way
the "offset" in one reactor is "countered" by the other --- and then
they "switch roles" when the AC reverses polarity. If you look at the
circuit here - (and again - for discussion sake totally ignore bucking)
- the AC is "reversed" all the time at one end - or the other of the
primary -- as the two coils are phase reversed.

Back to saturable reactor theory - when the DC control winding drives
the core into saturation - the reactance in the AC winding drops
dramatically. That being the case with this circuit - then the two
windings would 1) loose coupling so bucking is no longer a factor - and
2) have virtually no reactance in series with the bulb. Then by 1/2 the
AC "reactance winding" reversed - both halves would contribute their
part to the overall source impedance - providing better symmetry.

Now I'm not so sure that pure reactance doesn't play a larger role than
originally thought... That perhaps control is indeed more reactance -
and "bucking" is just a happy "bonus" to the equation...

Is one or the other really necessarily a "bonus", don't they work in
opposite directions assuming the two windings are connected in a
"bucking" configuration? Of course we don't actually know they are
really connected in a "bucking" configuration, we are just speculating
they are because of the way the schematic is drawn.

without taking some measurements (esp. being able to Un-reverse phase
the two windings) - it's hard to guess...

I'm glad you are the one working all this out, I would like to see more
information on the construction of the transformer and the disposition
of the windings before speculating too much. If the transformer is of
open frame construction and is not potted, the OP should be able to
easily determine if an E-core is used and if all the windings are on the
center leg?


Aw, c'mon guys, don't you understand push-pull?

I have never seen the specific radio in question, but the general idea is a
three-legged E-core, DC winding on the center, ac windings on the outer
legs. AC windings are "in-phase." A far as the AC is concerned the AC coils
are in (magnetic) series; as far as the DC coil is concerned they are in
(magnetic) parallel, to use the electrical analogy. Because the AC field
from both coils is equal and opposite, the AC signal is not induced into the
DC coil, but the DC field goes ( in opposing directions) through both legs
of the AC circuit, saturating them both as the current increases.