U.S. patent number 3,772,617 [Application Number 05/303,137] was granted by the patent office on 1973-11-13 for method and arrangement for setting the remanent flux density in magnetic circuits and equalizer utilizing same.
This patent grant is currently assigned to Bell Telephone Laboratories, Incorporated. Invention is credited to Albert John Ciesielka.
United States Patent |
3,772,617 |
Ciesielka |
November 13, 1973 |
METHOD AND ARRANGEMENT FOR SETTING THE REMANENT FLUX DENSITY IN
MAGNETIC CIRCUITS AND EQUALIZER UTILIZING SAME
Abstract
The remanent flux density in a selected portion of a magnetic
circuit is precisely set to a desired value by comparing the flux
density in the selected portion to a predetermined reference,
generating an error signal based on the comparison, integrating the
error signal, and inducing an H field in the magnetic circuit in
accordance with the integrated error signal. Integration of the
error signal is performed only when the flux density has a
substantially remanent value, i.e., only when the H field in the
magnetic circuit is substantially zero. Thus, error in setting the
flux density due to imperfect remanence in the magnetic circuit is
substantially eliminated. In an illustrative embodiment of the
invention, the magnetic circuit comprises a magnetic core having an
air gap, and is advantageously employed in a transmission cable
equalizer. The remanent flux density in the air gap is utilized to
precisely set the resistance value of a magnetoresistor, which in
turn determines the gain of the equalizer.
Inventors: |
Ciesielka; Albert John
(Atlantic Highlands, NJ) |
Assignee: |
Bell Telephone Laboratories,
Incorporated (Murray Hill, NJ)
|
Family
ID: |
23170692 |
Appl.
No.: |
05/303,137 |
Filed: |
November 2, 1972 |
Current U.S.
Class: |
333/18; 330/52;
330/144; 361/146; 327/553; 327/511; 307/101; 330/132; 333/16 |
Current CPC
Class: |
H04B
3/04 (20130101) |
Current International
Class: |
H04B
3/04 (20060101); H04b 003/04 (); H01f 013/00 () |
Field of
Search: |
;333/17,18,16
;307/101,309 ;317/123 ;330/52,132,144 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Gensler; Paul L.
Claims
I claim:
1. A method for setting the flux density in a selected portion of a
magnetic circuit comprising the steps of,
generating an error signal corresponding to the difference between
the flux density in said selected portion and a desired value
therefor,
integrating said error signal, and
inducing an H field in said circuit in accordance with the
integrated error signal,
characterized in that said integrating step is performed only when
the flux density in said selected portion has substantially a
remanent value.
2. A method in accordance with claim 1 wherein said integrating and
inducing steps are performed independently in sequence.
3. A method in accordance with claim 2 wherein said generating step
comprises the steps of,
sensing the flux density in said selected portion,
providing a control signal having an amplitude proportional to the
sensed flux density,
providing a reference signal corresponding to said desired flux
density value, and
determining the difference between said control signal and said
reference signal to generate said error signal.
4. A method in accordance with claim 3 wherein said selected
portion is an air gap and wherein said sensing step is performed
using a magnetoresistor positioned in said air gap.
5. A method in accordance with claim 4 wherein said control signal
providing step comprises the steps of extending a predetermined
signal to the input of an amplifier and adjusting the gain of said
amplifier based on the resistance value of said magnetoresistor,
said control signal being derived from the output signal of said
amplifier.
6. An arrangement for setting the flux density in a selected
portion of a magnetic circuit comprising, means for generating an
error signal corresponding to the difference between the flux
density in said selected portion and a desired value therefor,
means operative for integrating said error signal, and means
operative for inducing an H field in said magnetic circuit in
accordance with the integrated error signal, the improvement
comprising means for operating said integrating means only when the
flux density in said selected portion has substantially a remanent
value.
7. An arrangement in accordance with claim 6 wherein said
last-mentioned means comprises means for operating said integrating
means and said inducing means independently in sequence.
8. An arrangement in accordance with claim 7 wherein said
generating means comprises means for sensing the flux density in
said selected portion, means for providing a control signal having
an amplitude proportional to the sensed flux density, means for
providing a reference signal corresponding to said desired flux
density value, and means for determining the difference between
said control signal and said reference signal to generate said
error signal.
9. An arrangement in accordance with claim 8 wherein said selected
portion is an air gap and wherein said sensing means includes a
magnetoresistor positioned in said air gap.
10. An arrangement in accordance with claim 9 wherein said control
signal providing means comprises an amplifier, means for extending
a predetermined signal to the input of said amplifier and means for
adjusting the gain of said amplifier in accordance with the
resistance value of said magnetoresistor, said control signal being
derived from the output of said amplifier.
11. In an equalizer including circuit means adapted for insertion
into a transmission path and operative for varying the gain of said
path the combination comprising, a magnetic core having an air gap
interrupting its continuity and providing remanent magnetic flux in
said air gap, means for operating said circuit means in accordance
with the density of magnetic flux in said air gap, means coupled to
said circuit means for detecting a pilot signal applied to said
path, means for comparing the detected pilot signal to a
predetermined reference signal and for generating an error signal
based on the comparison, means operative for integrating said error
signal only when the flux density in said gap has substantially a
remanent value, and means operative for applying a magnetic field
to said core to adjust the density of the flux in said air gap,
said magnetic field being proportional to the output of said
integrating means.
12. An equalizer in accordance with claim 11 wherein said
last-mentioned means comprises means for operating said integrating
means and said applying means independently in sequence.
13. An equalizer in accordance with claim 12 wherein said means for
operating said circuit means includes magnetoresistive means
positioned in said air gap.
14. An equalizer in accordance with claim 13 wherein said detecting
means includes means for providing a d.c. control signal
proportional to said detected pilot signal and where said comparing
means includes means for determining the difference between said
control signal and said error signal.
15. An equalizer in accordance with claim 14 further comprising
means for amplifying the output of said integrating means, said
magnetic field being proportional to the output of said amplifying
means.
Description
BACKGROUND OF THE INVENTION
My invention relates generally to magnetic circuits and, more
particularly, to a method and arrangement for precisely setting
remanent flux density in magnetic circuits.
As herein defined, a magnetic circuit is an arrangement in which
magnetic flux is produced in response to an applied magnetic field.
A particular magnetic circuit may include one or more paths of
magnetic material and a coil for applying the magnetic field
thereto. The magnetic circuit may also include an air gap or gaps
in one or more of the paths. The flux density in a particular path
or portion of the circuit is determined principally by the circuit
geometry, the strength of the applied magnetic field and the
previous magnetic history of the circuit.
In the prior art, the flux density, or B field, in a selected
portion of a magnetic circuit is generally set to a desired value
by inducing a magnetic field, or H field, in the circuit such that
the desired flux density is attained. Many magnetic materials, for
example, ferromagnetic materials, exhibit substantial remanence, so
that the H field in the magnetic circuit can be reduced to zero
without the flux density being reduced to zero as well.
Advantageously, then, a nonzero flux density can be permanently
established in a selected portion of a magnetic circuit (for
example, an air gap or a section of magnetic material) without
expending energy to maintain same.
However, remanence in magnetic materials is not perfect and the
flux density does not remain precisely at a desired, originally set
value once the H field is removed. Rather, the flux density drops
somewhat, retaining on the order of 80 to 90 percent of its
original value. In certain applications, an error of this magnitude
may be tolerable. In others, more precise flux density control is
necessary.
One known arrangement for precisely setting flux density in a
magnetic circuit maintains an H field in the circuit at all times.
This arrangement precludes advantageous utilization of magnetic
remanence, as described above. Moreover, a failure in the H field
supply will change the flux density from the desired value.
Another prior arrangement utilizes so-called square hysteresis loop
materials which have remanences typically on the order of 98
percent. These materials are relatively expensive and thus not
commercially viable for many applications.
Moreover, in certain magnetic circuit applications, an error of
even several percent in a flux density setting may be intolerable.
For example, a magnetic circuit comprising a magnetic core having
an air gap interrupting its continuity, can be utilized in an
arrangement for controlling the gain of a transmission cable
equalizer. A magnetoresistor is positioned in the air gap, its
resistance value thus being determined by the remanent flux density
in the gap. The equalizer gain is determined by the resistance
value of the magnetoresistor. Thus, precise control of the
equalizer gain requires that the remanent flux density in the air
gap be precisely set.
SUMMARY OF THE INVENTION
Accordingly, an object of the invention is to provide a method and
arrangement for precisely setting remanent flux density in magnetic
circuits.
A more specific object of the invention is to provide a method and
arrangement for precisely setting the remanent flux density in an
air gap in a magnetic circuit.
A further object of the invention is to provide a method and
arrangement for precisely setting the flux density in the air gap
of a magnetic core and thus the resistance value of a
magnetoresistor positioned in the gap.
Another object of the invention is to provide a method and
arrangement for precisely setting the flux density in the air gap
of a magnetic core to precisely set the resistance value of a
magnetoresistor to, in turn, precisely set the gain of a
transmission cable equalizer.
These and other objects are achieved by comparing the flux density
in a selected portion of a magnetic circuit to a predetermined
reference, generating an error signal based on that comparison,
integrating the error signal, and inducing an H field in the
magnetic circuit in accordance with the integrated error signal to
thereby change the flux density in the selection portion. In
accordance with the invention, integration of the error signal is
performed only when the flux density has substantially a remanent
value; that is, only when substantially no H field is being induced
in the magnetic circuit. In this way, the invention substantially
eliminates error in setting the flux density due to imperfect
magnetic remanence. In the prior art, as described above, a desired
flux density is established while some H field is being induced in
the magnetic circuit so that subsequent reduction of the H field to
zero causes an unwanted change in the flux density.
An important feature of the invention is that the above-mentioned
integrating and field inducing steps are performed independently in
sequence; when the error signal is being integrated, no H field is
induced in the magnetic circuit and when the H field is being
induced in the magnetic circuit, no integration of the error signal
is performed. The sequence of integrating and field inducing is
performed repetitively until the flux density in the selected
portion of the magnetic circuit attains the desired value.
An illustrative embodiment of the invention precisely sets the flux
density in the air gap of a magnetic core. The flux density is
sensed by a magnetoresistor positioned in the gap, and the
resistance value of the magnetoresistor establishes the amplitude
of a control signal. The above-mentioned error signal is generated
by determining the difference between the control signal and the
above-mentioned predetermined reference.
In the illustrative embodiment, the magnetoresistor is
advantageously utilized as a gain-determining element in a
transmission cable equalizer. A pilot frequency tone is transmitted
down the cable to be equalized, and is thereby extended to the
input of an amplifier included in the equalizer. The pilot tone,
detected at the equalizer output, is rectified and filtered to
provide the above-mentioned control signal. A subtractor determines
the difference between the control signal and a predetermined
reference voltage proportional to that value of air gap flux
density which provides a predetermined equalizer output amplitude.
The resultant error signal is processed in accordance with the
invention, as described above, to set the remanent flux density in
the air gap, by which the resistance value of the magnetoresistor
and hence the gain of the equalizer are set.
BRIEF DESCRIPTION OF THE DRAWING
A clear understanding of the invention and of the preceding and
other objects and features thereof may be gained from a
consideration of the following detailed description and
accompanying drawing in which:
FIG. 1 depicts an arrangement according to the invention,
illustratively employed in a transmission cable equalizer; and
FIGS. 2A and 2B show graphs useful in describing the operation of
the arrangement of FIG. 1.
DETAILED DESCRIPTION
FIG. 1 depicts a magnetic circuit comprising a magnetic core 378,
an air gap 379 interrupting the continuity thereof, and core
windings 377. FIG. 1 further depicts circuitry in accordance with
the invention for precisely setting the remanent flux density in a
selected portion of the magnetic circuit, viz. gap 379. The
magnetic circuit is utilized illustratively in an automatic
transmission cable equalizer 30 which includes an operational
amplifier 31. As described in detail hereinafter, the magnitude of
the flux density in gap 379 determines the gain of amplifier 31 and
hence of equalizer 30 for a predetermined pilot frequency. Thus the
present invention, by providing precise setting of the flux density
in gap 379, assures precise setting of the gain of equalizer 30 at
that pilot frequency.
The gain of equalizer 30 is a function of input frequency and is
adjusted at each of (illustratively) four pilot frequencies
f.sub.1, f.sub.2, f.sub.3 and f.sub.4 (which are distributed across
the frequency band over which equalizer 30 is to operate) by
adjusting the impedance of one of four equalizer input networks
34-37. For this purpose, networks 34-37 include variable
resistances 34a-37a, respectively. Equalizer 30 is adapted for
equalizing a transmission path comprising transmission cable 20
over which signals are extended from an information source 10 to a
utilization circuit. The frequency-dependent gain characteristic of
equalizer 30 is adjusted to correspond to frequency-dependent
attenuation in cable 20. Thus, the frequency characteristic of
signals from source 10 is substantially preserved at the input of
the utilization circuit. When equalizer 30 is to be aligned, that
is, when its gain characteristic is to be adjusted in accordance
with the frequency-dependent attenuation in cable 20, tones at each
of the above-mentioned pilot frequencies are transmitted down cable
20 from a pilot tone source 15. The impedances of networks 34-37
are adjusted by adjusting variable resistances 34a-37a,
respectively such that the signal at output terminal 38 of
amplifier 31 has a predetermined amplitude at each pilot frequency.
Advantageously, this adjustment of variable resistances 34a-37a is
effected automatically. For this purpose, connections between
terminal 38 and each variable resistance are provided via taps off
lead 33.
Each of the variable resistances 34a-37a comprises a magnetic core
having an air gap, circuitry in accordance with the invention for
precisely setting the flux density in the air gap, and a
magnetoresistor positioned in the air gap. The magnetoresistor in
each variable resistor comprises the resistive element thereof
while the magnetic core and its associated circuitry control the
resistance value of the magnetoresistor.
Thus, variable resistance 37a, for example, includes magnetic core
378, magnetoresistor MR positioned in gap 379, and circuitry for
setting the flux density in gap 379. The latter comprises
rectifying and filtering circuit 370, subtractor 371, error signal
lead ER, switch 372, integrator 373, amplifier 374, switches 375
and 376 all connected in series in the order named. Switch 376 is
connected to core windings 377. Variable resistance 37a also
includes a pilot tone detector 380 which is coupled to output
terminal 38 of equalizer 30, as is rectifying and filtering circuit
370. Detector 380 closes switch 376 when a tone at pilot frequency
f.sub.4 is present at terminal 38.
Setting the flux density in gap 379 in accordance with the
above-discussed prior art can be illustrated by assuming that when
cable 20 is being equalized at pilot frequency f.sub.4, switches
372 and 375 are both closed. Circuit 370 provides a d.c. control
signal proportional to the amplitude of the a.c. pilot signal (of
frequency f.sub.4) at output terminal 38. Subtractor 371 determines
the difference between the control signal and a predetermined d.c.
reference. The resultant error signal is extended along lead ER to
integrator 373 and therefore to amplifier 374. The voltage at
output terminal R of amplifier 374 generates a current in windings
377. The H field induced in windings 377 and thus in core 378
adjusts the flux density in gap 379, thus the resistance value of
magnetoresistor MR and thus the gain of equalizer 30 at pilot
frequency f.sub.4.
The gain of amplifier 374 is chosen, based on the parameters of the
other components of variable resistance 37a to ensure that the
system is stable and that the error signal on lead ER will converge
to zero. At that time, equalization of cable 20, i.e. alignment of
equalizer 30 at pilot frequency f.sub.4, is complete. The signal at
terminal R and, accordingly, the H field induced in core 378 become
constant.
Thus, in the prior art, when the flux density in gap 379 and the
resistance value of magnetoresistor MR are such that cable 20 is
equalized at pilot frequency f.sub.4, the H field in core 378 is
not zero. If the H field in core 378 is thereafter reduced to zero
to take advantage of the magnetic remanence of core 378, the flux
density in gap 379 drops (assuming first quadrant operation)
somewhat because the remanence in core 378 is not perfect. This, of
course, lowers the resistance value of magnetoresistor MR and
destroys the equalization. The remanence even of so-called square
hysteresis loop materials, is generally too imperfect for precise
equalization.
This problem in the prior art is alleviated in accordance with the
present invention by integrating the error signal only when the
flux density in the air gap has a remanent value; that is, only
when the H field in the core is substantially zero. Thus,
integration of the error signal on lead ER and induction of an H
field in core 378 on the basis of the signal at terminal R, are
performed independently in sequence in two distinct steps. During
the first step, switch 372 is closed and switch 375 is opened so
that a voltage proportional to the error signal on lead ER, and
thus to the equalization error, is built up at terminal R with no
current flowing in windings 377. During the second step, switch 372
is opened and switch 375 is closed so that current flows in
windings 377 proportional to the equalization error that obtained
when the flux density in gap 379 was at its last remanent value.
The resultant H field induced in core 378 changes the flux density
in gap 379, thereby adjusting the resistance value of
magnetoresistor MR and reducing the equalization error.
This two-step cycle of integration and induction, hereinafter
referred to as a "magnetization cycle", is performed repetitively
until the desired flux density is achieved. Once the error signal
on lead ER is zero at the start of a cycle, i.e. with switch 375
open, the flux density in gap 379 is assured to have the desired
remanent value which provides alignment of equalizer 30 at pilot
frequency f.sub.4. In the prior art, this same value of flux
density would be attained with some nonzero H field induced in core
378. Thus it is seen that, advantageously, the present invention
substantially eliminates equalization error due to imperfect
remanence in core 378.
Although switches 372 and 375 are illustratively shown in FIG. 1 as
mechanical elements, other known arrangements, for example,
electronic switches, as well as circuitry for operating them in the
manner described above, will be apparent to those skilled in the
art.
FIGS. 2A and 2B show graphs exemplifying the operation of equalizer
30 and, in particular, of variable resistance 37a, during alignment
of the former at pilot frequency f.sub.4. FIG. 2A depicts
variations in the gap 379 flux density in response to H field
pulses induced in core 378 via windings 377. FIG. 2B depicts a time
graph of those H pulses, as well as the signal at terminal R which
generates them. The H scale along the ordinate of FIG. 2B is
aligned with the H scale along the abscissa of FIG. 2A.
As seen from FIGS. 2A and 2B, the voltage at terminal R is
initially zero and the flux density in gap 379 is B.sub.i.
Assumptively, the flux density necessary to equalize cable 20 at
pilot frequency f.sub.4 is B.sub.o.
In the example of FIGS. 2A and 2B, ten magnetization cycles,
designated I, II...X, are needed to change the remanent flux
density in gap 379 from B.sub.i to B.sub.o. The integration steps
of cycles I, II...X are performed in time periods designated Ia,
IIa...Xa, respectively. The induction steps are performed in time
periods designated Ib, IIb,...Xb, respectively. During periods Ia,
IIa,...Xa, switches 372 and switches 375 are closed and opened,
respectively. During periods Ib, IIb,...Xb, the opposite
relationship obtains.
Since B.sub.i is less than B.sub.o (the d.c. reference provided to
subtractor 371 corresponding to the latter), the error signal on
lead ER is negative during period Ia and a positive ramp is
generated at terminal R. In period Ib, then, positive current (as
indicated by an arrow) flows into windings 377, inducing an H field
pulse of magnitude H.sub.C in core 378.
Although the rise time of the H field pulses in FIG. 2B are shown
as substantially zero to simplify illustration, some nonzero rise
time elapses before the pulses attain their full magnitude. It will
be appreciated that this rise time is not an important
consideration as long as periods Ib, IIb...Xb have sufficient
duration.
As seen in FIG. 2A, the flux density in gap 379 rises to point C in
response to the H field pulse of period Ib. Since core 378 has
imperfect remanence, the flux density in gap 379 falls back to
point D when switch 375 opens at the start of period IIa. The flux
density at point D is less than B.sub.o so that the error signal on
lead ER is still negative. Thus, the voltage at terminal R further
increases during period IIa. In period IIb a pulse of magnitude
H.sub.E is induced in windings 377, driving the flux density in gap
379 to point E.
The flux density falls back to point F as period IIIa begins. The
flux density is now greater than B.sub.o so that the error signal
on lead ER is positive and the voltage at terminal R decreases
during period IIIa. The pulse induced during period IIIb has
magnitude H.sub.G, the flux density rising to point G but dropping
back to point F as period IVa begins.
During period IVa, the terminal R voltage drops still further. The
flux density rises to point I during period IVb and drops back to
point F as period Va begins.
In period Va the voltage at terminal R becomes negative and a
negative H field pulse of magnitude H.sub.J is induced in windings
377 during period Vb. This reduces the flux density in gap 379 to
point J. Due to imperfect remanence in core 378, the flux density
rises to point K as period VIa begins. The flux density is reduced
to point L during period VIb.
As period VIIa begins, the flux density rises to point M at which
point the flux density is again less than B.sub.o and the error
signal on lead ER is again negative.
Accordingly, the voltage at terminal R increases during period
VIIa. The flux density thus goes to point N in period VIIb, returns
to and remains at point M during cycle VIII and rises to point O in
cycle IX, from which it falls back to point P and finally rises to
point Q in cycle X.
At the end of cycle X the flux density in gap 379 falls to B.sub.o.
Thereafter, the error signal on lead ER is zero when switches 372
and 375 are closed and open, respectively, and accordingly, the
voltage at terminal R stays constant. Thus, the flux density
oscillates between points Q and B.sub.o during any magnetization
cycles which follow cycle X, such as cycle XI. Generally, the
number of magnetization cycles which will be needed to attain a
particular remanent flux density, such as B.sub.o, is not known
beforehand, and some number of magnetization cycles corresponding
to an anticipated maximum will occur during alignment at each pilot
frequency.
The above described arrangement is merely illustrative of the
invention. Thus, although use of the invention in conjunction with
a particular magnetic circuit is disclosed herein, it will be
appreciated that the invention can be utilized to set the remanent
flux density in any magnetic circuit, whether or not it has an air
gap, as long as the circuit has provision for detecting the flux
density in the portion of the circuit of interest. An example of
such a magnetic circuit is the transfluxor, described for example
by J. A. Rajchman and A. W. Lo in Proceedings of the IRE, Vol. 44,
page 321, March 1956.
It is to be understood therefore that numerous and other varied
arrangements in accordance with the principles of the invention may
be devised by those skilled in the art without departing from the
spirit and scope thereof.
* * * * *