| |
LW1DSE > TECH 21.08.10 21:30l 264 Lines 15300 Bytes #999 (0) @ WW
BID : 1787-LW1DSE
Read: GUEST
Subj: Magnetic Amplifiers
Path: IZ3LSV<IK2XDE<DB0RES<ON0BEL<XE1FH<HI5MLE<LW1DRJ<LW8DJW
Sent: 100821/2001Z 50018@LW8DJW.#1824.BA.ARG.SA [Lanus Oeste] FBB7.00e $:1787-L
From: LW1DSE@LW8DJW.#1824.BA.ARG.SA
To : TECH@WW
[¯¯¯ TST HOST 1.43c, UTC diff:5, Local time: Sat Aug 21 16:21:05 2010 ®®®]
ÉÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍ»
º Magnetic Amplifiers (Mag-Amp) º
º ***************************** º
ÈÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍͼ
Magnetic amplifiers, also called mag amps for short, provide an
electro-magnetic method of amplification. Mag amps were quite common prior to
the development of solid state transistors. As advances in semiconductor
technology progressed, magnetic amplifiers become a relatively expensive
component. Consequently the use of mag amps declined. A properly made mag amp
is highly reliable, hence they are still used in some applications with
demand the reliability performance criteria that a mag amp can meet. Another
feature of mag amps is the high isolation voltages that can be achieved
between windings with proper design. Mag amps may still be preferred over
semiconductor devices in safety critical applications.
A typical simple mag amp contains two identical coils, each having
identical high permeability square loop magnetic cores and each wound with an
identical winding not shared with the other coil. An alternating voltage
source is connected to one end of these windings and a load is connected to
the other end. The windings are either connected in series or in parallel
such that the cores' magnetic flux generated by the alternating voltage are
out of phase (in opposite directions). Alternating current (A.C.) will flow
through these windings. Either a shared second winding is wound on both coils
or each coil is wound with a second identical winding. In the latter case the
windings are series connected such that a D.C. (fig 1) flowing through these
windings generate magnetic flux in the cores, which are in phase (in the same
direction). These windings are connected to a variable D.C. current source
(which might consist of series connected D.C. voltage source and a variable
resistor). The D.C. winding(s) is (are) referred to as the control winding(s).
Schematic representations of two typical mag amps are given in Figs 1 and 2
further below. The mag amps shown may also be referred to in literature as a
type of saturable reactor. A mag amp may also be referred to in literature as
a type of transductor.
Air gaps within a mag amp's core structure are detrimental to mag amp
performance. Proper mag amp performance requires nearly identical symmetry in
core flux excursions; hence leakage flux should be minimized. Toroidal cores
have essentially zero air gaps and the toroidal geometry maximizes magnetic
coupling and minimizes leakage flux. Consequently, toroids are the core shape
of choice.
Other variations of mag amps exist, including a single core version
that has three core legs. The middle leg has a D.C. control winding. The
outer legs have identical A.C. windings. In theory D.C. flux generated in the
center leg divides equally and flows through both outer legs (fig 1). The A.C.
windings are connected such that their phases don't permit any A.C. flux flow
through the center leg (in theory). There are practical difficulties (in the
form of magnetic tolerances) with this type of mag amp design. More advanced
mag amp circuits use rectifying elements to isolate the load from the mag amp
during core reset. Core reset refers to the volt-second transition from
saturation flux (top flat portion of the B-H loop) to the flux value at the
opposite side of the B-H loop (bottom flat portion of the loop).
Butler winding can make (and has made) mag amps. Butler winding has
several types of toroid winding machines that can be used to wind a variety
of mag amp core sizes. This includes toroid-taping machines. For toroids, we
can (and have done) sector winding, progressive winding, bank winding, and
progressive bank winding. Butler winding also has other types of winding
machines. That includes two programmable automated machines. We can wind and
assemble various standard types of core with bobbin structures
(E, EP, EFD, PQ, POT, U and others), and some custom designs. Our upper
limits are 40 pounds of weight and 2 kilowatts of power. We have experience
with foil windings, litz wire windings, and perfect layering. Butler winding
has vacuum chamber(s) for vacuum impregnation and can also encapsulate. To
ensure quality, Butler Winding purchased two programmable automated testing
machines. Most of our production is 100% tested on these machines.
Mag Amp Theory
--------------
The following discussion isn't intended to give a detailed understanding
of mag amp operation. It isn't intended to describe all the variations of mag
amp designs or applications. It is intended to give a basic insight to how a
typical simple mag amp functions. Rectifier aided mag amp circuits aren't
discussed. Butler Winding has some but limited experience with mag amps. If
you require more information than the following discussion supplies, please
contact Butler Winding and ask to speak to an engineer about mag amps. Butler
Winding will provide whatever help we reasonably can.
Refer to the schematic of Figure 2 bearing in mind (in theory) that the
two coils have identical windings and identical cores. Because of transformer
action, the A.C. voltage impressed across the mag amp's A.C. windings will
induce a voltage across each control winding. Because of the opposite phasing
of the A.C. windings, the induced voltages in the D.C. windings will buck
each other and exactly cancel each other (in theory) resulting in zero A.C.
voltage induced across the D.C. source. Consequently, low impedance D.C.
source will not load down the A.C. windings.
Consider the impedance of the A.C. windings with no D.C. current
supplied. The core and windings are designed such that;
1) the core doesn't saturate at the maximum intended A.C. voltage, and
2) each A.C. winding has a relatively much higher impedance than the intended
load.
Because of the high impedance, very little A.C. current flows.
Consequently, there is very little voltage drop across the load. Now consider
the impedance of the A.C. windings with a D.C. current flowing through the
control winding. Both cores have a D.C. biasing flux of equal value and the
same phasing. The A.C. windings of Figure 5 are connected in parallel but with
opposite phasing. The total flux in a core is the sum of the D.C. flux and the
A.C. flux. Because of the opposite A.C. winding phasing, the A.C. voltage
increases the core flux of one core while decreasing the core flux of the
other core until saturation occurs. Eventually the alternating fashion of the
A.C. voltage causes the changing flux to reverse the direction of flux change
of both cores. Now apply enough D.C. current to cause one core to enter
saturation. The core's flux reaches its maximum values and doesn't change
(ideal theory) while in saturation; hence no induced voltage will oppose the
applied A.C. voltage. The impedance of that core's A.C. winding drops to near
zero value. There can be very little voltage drop across that A.C. winding.
The other A.C. winding is connected in parallel to this A.C. winding. This
A.C. winding shunts the current around the other A.C. winding hence the other
A.C. winding also sees very little voltage impressed across it. Consequently
the flux of the other core changes very little (essentially stays where it
is). While a core is saturated there is very little impedance between the
A.C. voltage source and the load impedance. Consequently significant load
current flows during saturation and produces a relatively large voltage drop
across the load. Because of the eventual A.C. voltage reversal, the saturated
core will eventually come out of saturation, high A.C. winding impedance will
occur again, and the load current will again drop to near zero value.
Eventually the other core saturates resulting in high load current until the
core leaves saturation. The mag amp has seen a complete A.C. cycle and will
proceed to the next cycle. For mag amps, entering saturation is like closing
a switch. The time spent in saturation is the turn-on time of the mag amp
switch. The amount of time spent in saturation is determined by the amount of
D.C. biasing current. A larger D.C. bias current causes the cores to enter
saturation earlier and exit saturation later, thereby increasing the length
of time current is delivered to the load, thereby increasing the average
amount of current delivered to the load in a given period of time. Once a
steady state condition is reached in an idealized mag amp, it can be shown
that the averaged ampere-turns of the load current are proportional to the
ampere-turns of the control current. With appropriate choices of turns
ratio, windings, and cores, one can achieve significant power amplification
gain.
The schematic in Figure 4 shows the A.C. windings connected in series.
When one core saturates both of its winding have relatively very low
impedance and can be ignored. The core's A.C. winding doesn't shunt the other
A.C. winding, but the other A.C. winding will not maintain its high impedance
level if the D.C. source has a sufficiently low impedance. With one core
saturated the low impedance D.C. source becomes a transformer-coupled load to
the unsaturated A.C. winding. The impedance on the unsaturated A.C. winding
drops to the transformer coupled reflected value of the low impedance D.C.
source. A load current flows which produces a significant load voltage.
ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿ Figure 1: Mag Amp's
a11 ³ è ------- ³ a21 windings distribution.
\³ ÚÄÄÄÄÄÄÄÄÄÄ¿ ÚÄÄÄÄÄÄÄÄÄÄ¿ ³ / using 1 core.
±±³ ³±± ²²³ ³²² ±±³ | ³±±
±±³ | ³±± ²²³ | ³²² ±±³ | ³±±
±±³ | ³±± ²²³ | ³²² ±±³ | ³±± è = AC magnetic field
±±³ | ³±± d1_²²³ | ³²²_d2 ±±³ ³±± (doesn't circulate
/ ³ è ÀÄÄÄÄÄÄÄÄÄÄÙ í ÀÄÄÄÄÄÄÄÄÄÄÙ è ³\ in center leg).
³ í - Á - í ³ a22
a12 ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙ í = DC magnetic field.
Magetic core
a11 ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿ ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿ a21 Figure 2: Mag Amp's
\³ ÚÄÄÄÄÄÄÄÄ¿ ³ ÚÄÄÄÄ¿³ ÚÄÄÄÄÄÄÄÄ¿ ³/ windings distribution
±±³ ³±± ²²³ ³²² ²²³ ³²² ±±³ ³±± using 2 similar cores.
±±³ ³±± ²²³ ³²² ²²³ ³²² ±±³ ³±±
±±³ ³±± ²²³ ³²² ²²³ ³²² ±±³ ³±±
±±³ ³±± ²²³ ³²² ²²³ ³²² ±±³ ³±±
/ ³ ÀÄÄÄÄÄÄÄÄÙ ³³ ³ ³ ÀÄÄÄÄÄÄÄÄÙ ³ \
a12 ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙd1 d2 ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙ b22
2 identical magetic cores
References:
----------- a11 d1
± = AC (controlled) ÄÄÄÄ¿ · ÚÄÄÄÄ
o ± º ² o
² = DC (control) ± º ² Fig 3: Schematic of Mag Amp.
a11 = start of winding #1 ± º ² Note the core representing
a12 = end of winding #1 a12 ± º ² the squeare histeresis loop.
a21 = start of winding #2 ÄÄÄÄÙ Ó ÀÄÄÄÄÄÄ¿
a22 = end of winding #2 a21 d2 ³
d1 = start of winding #3 ÄÄÄÄ¿ · ÚÄÄÄÄ ³
d2 = end of winding #3 o ± º ² ³
o = denotes polarity of ± º ² ³
windings ± º ² ³
a22 ± º ² o ³
ÄÄÄÄÙ Ó ÀÄÄÄÄÄÄÙ
~ a11 d1
oÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿ · ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄo +
o ± º ² o
AC voltage source ± º ² DC Controling source
± º ²
oÄÄÄÄÄÄÄ¿ ± º ² ÚÄÄo -
~ ³ ÚÄÄÄÄÙ Ó ÀÄÄÄÄ¿ ³
³ ³ a12 ³ ³
³ ³ ³ ³
³ ³ a21 ³ ³ ³i
³ ÀÄÄÄÄ¿ · ÚÄÄÄÄÙ ³
³ ± º ² o ³ DC current
³ ± º ² ° control adjust.
³ ± º ² °<ÄÄ¿
³ o ± º ² ° ³
³ ÚÄÄÄÄÙ Ó ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÙ
³ ³ a22 d2
³ ³ Fig 4: Schematic of a mag amp in
³ ° a circuit. AC windigs are
³ ° R Load AC series wired.
³ °
ÀÄÄÄÄÄÄÄÄÄÙ
<---> AC controled current
~ a11 d1
oÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄÄÄÄÄ¿ · ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄo +
³ o ± º ² o
AC voltage source ³ ± º ² DC Controling source
³ ± º ²
oÄÄÄÄÄÄÄ¿ ³ ± º ² ÚÄÄo -
~ ³ ÚÄ(ÄÄÄÄÄÙ Ó ÀÄÄÄÄ¿ ³
³ ³ ³ a12 ³ ³
³ ³ ³ ³ ³
³ ³ ³ a22 ³ ³ ³i
³ ³ ÀÄÄÄÄÄ¿ · ÚÄÄÄÄÙ ³
³ ³ o ± º ² o ³ DC current
³ ³ ± º ² ° control adjust.
³ ³ ± º ² °<ÄÄ¿
³ ³ ± º ² ° ³
³ ÀÄÂÄÄÄÄÄÙ Ó ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÙ
³ ³ a21 d2
³ ³ Fig 5: Schematic of a mag amp in
³ ° a circuit. AC windigs are
³ ° R Load AC parallel wired.
³ °
ÀÄÄÄÄÄÄÄÄÄÙ
<---> AC controled current
B Fig 6: Squeare histeresis
³ loop of core materials
ÚÄÄÅÄÄÂÄÄ for magnetic amplifiers.
³ ³ ³
³ ³ ³
<ÄÄÄÄÄÅÄÄÅÄÄÅÄÄÄÄÄ> H
³ ³ ³
³ ³ ³
ÄÄÁÄÄÅÄÄÙ
ÉÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍ»
º Compilled from various sources in the Intenet. Translatted to ASCII by º
º LW1DSE Osvaldo F. Zappacosta. Banfield (1828), Buenos Aires, Argentina. º
º Made with MSDOS 7.10's Text Editor (edit.com) in my AMD's 80486. º
º August 20, 2010 º
ÈÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍͼ
ÉÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍ»
º Osvaldo F. Zappacosta. Barrio Garay (GF05tg) Alte. Brown, Bs As, Argentina.º
º Mother UMC æPC:AMD486@120MHz, 16MbRAM HD IDE 1.6Gb MSDOS 7.10 TSTHOST1.43C º
º Bater¡a 12V 160AH. 9 paneles solares 10W. º
º oszappa@yahoo.com ; oszappa@gmail.com º
ÈÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍͼ
Read previous mail | Read next mail
| |
