Abstrict An electromagnetic flow meter includes a preamplifier with extremely
long time constant to achieve wide band operation. To alleviate
undesirably long recovery from transients and the like, an effective
dc bias is obtained by periodically resetting the bias levels. The
invention is applicable outside the field of electromagnetic flow
meters.
Claims I claim:
1. An electromagnetic flow meter comprising:
a clock producing a cyclic output,
magnetic field means driven by said clock to produce a magnetic
field of cyclic form divided into two magnet half cycles,
a transducer including a pair of electrodes and means to support
said electrodes in a fluid flow field subject to a magnetic field
produced by said magnetic field means,
wide band amplifier means, with a pair of outputs, coupled to said
electrodes for changing impedance levels of signals derived from
said electrodes,
said wide band amplifier means having a bias control including
switching means driven in synchronism with said clock to stabilize
a bias level of said wide band amplifier means to thereby limit
a period of saturation in said wide band amplifier means,
combining means to combine signals from said pair of outputs, and
sampling switch means coupled to an output of said combining means
an driven in synchronism with said clock to sample an output of
said combining means in timed relation to operation of said switching
means.
2. The apparatus of claim 1 wherein, during each magnet half cycle,
said sampling switch means effects a sampling operation prior to
operation of said switching means and said switching means resets
a voltage at a terminal in said wide band amplifier means.
3. The apparatus of claim 1 wherein said wide band amplifier means
includes a FET source follower with a gate terminal coupled to one
of said pair of electrodes, a voltage regulator coupled to said
FET, said voltage regulator including said switching means for at
times resetting a voltage at a terminal in said voltage regulator.
4. The apparatus of claim 3 wherein said FET is housed in said
transducer.
5. The apparatus of claim 1 wherein said pair of electrodes each
includes a sensing electrode and a guard electrode, said wide band
amplifier means includes a pair of wide band amplifiers, one for
each of said sensing electrodes, each said pair of wide band amplifiers
coupled between a sensing electrode and a guard electrode.
6. The apparatus of claim 5 wherein each of said pair of wide band
amplifiers includes an FET input stage.
7. The apparatus of claim 6 wherein said FET input stage includes
a source resistor node which is biased by an output of one of said
pair of wide band amplifiers.
8. The apparatus of any of claims 1-7 wherein said combining means
comprises a differential amplifier with a pair of inputs, each connected
to a different one of said wide band amplifier means outputs.
Description DESCRIPTION
Technical Field
The invention relates to electromagnetic flow meters, and more
particularly to a preamplifier designed to process signals produced
by the electrodes of an electromagnetic flow meter.
Background Art
The electromagnetic flow meter includes a magnetic field source
which is arranged to subject a flowing fluid field to a predetermined
magnetic field. At least a pair of electrodes are also supported
in the flowing fluid field which is subjected to the magnetic field,
and voltages are induced into the electrodes which are related,
at least in part, to the fluid velocity. Associated apparatus is
designed to process the signal induced on the electrodes so as to
discriminately amplify the flow related voltages at the expense
of other spuriously induced voltages. Some flow meters have conductive
or galvanic relationship with the metered fluid. Others are capacitively
coupled to the metered fluid. The latter type of flow meters, sometimes
called electrodeless, actually also have electrodes. For purposes
of this application any device or element that conveys the flow
induced voltage to signal processing elements is considered to be
or include an electrode regardless of the mechanism (conductive,
galvanic or capacitive) employed.
The burden placed on the electronics is indeed significant because
the flow related voltage is measured in microvolts whereas typically
the voltage required to produce the magnetic field for example can
be measured in tens of volts. Thus the associated electronics has
to deal with a number of noise sources. A first noise source is
derived from unwanted coupling between the magnetic field source
and the sensing electrodes and its associated electronics. A second
noise source is derived from power line induced voltages.
The prior art has dealt with these same problems; a common technique
for reducing the contribution of these noise sources is common mode
rejection, i.e. a differential amplifier is used in the signal processing
chain to amplify the difference in the voltages induced in two sensing
electrodes. This is particularly effective for diffuse sources such
as power line noise. Another useful technique for rejecting power
line signals is described in my prior U.S. Pat. No. 3855858 entitled
"Self-Synchronous Noise Rejection Circuit for Fluid Velocity
Meter".
The other type of noise contribution, derived from the magnetic
field source itself is termed "transformer effect", because
it is induced in the electronics wiring via magnetic induction.
A substantial reduction in this transformer effect voltage is obtained,
as described in my prior U.S. Pat. No. Re. 28989 by arranging the
magnetic field to switch between two levels, i.e. alternate. Since,
as is known to those skilled in the art, a voltage induced via magnetic
coupling is related to the time derivative of magnetic flux, the
"transformer effect" voltage induced as a result of an
alternating magnetic field, is at a maximum just after the change
in magnetic field and dies out as a function of time.
Accordingly, as described in my U.S. Pat. No. Re. 28989 a sampling
technique is employed to sample the electrode generated voltage,
and the sampling incident is chosen as long after the alternation
in magnetic flux, as is possible, i.e. just before the next alternation.
However, parasitic effects in the electronics have, in the past,
been a limiting factor in reducing this "transformer effect"
voltage or noise. Furthermore, the very techniques used to minimize
transformer effect coupling have introduced other, undesirable characteristics
into the equipment. For example, a preamplifier is typically capacitively
coupled between electrodes and the differential amplifier. The main
function of the preamplifier is impedance transformation, i.e. change
the megohm (or higher) level of the electrode signal to something
more reasonable for coupling to the differential amplifier. While
prior art workers have used preamps with time constants of one or
a few seconds, I prefer to use time constants of much longer duration,
e.g. many hundreds of seconds, a thousand seconds or even more.
This is advantageous in minimizing the effects of the "transformer
effect". The use of square wave magnetic field waveforms and
sampled sensing is predicated on the rapid decay of spurious signals.
However, the shorter time constant preamps introduce a phase shift
which can undercut the flow meter's noise handling technique. While
the spurious voltages do decay rapidly the phase shift can skew
the noise timing so that more noise is present at the sampling time
than would otherwise be expected. This affects the flow meter's
signal vs. flow velocity characteristic. By using long time constants
the preamp takes on wide band characteristics and eliminates (or
significantly reduces) the unwanted phase shifting. The term wide
band must be interpreted relative to the predominant frequency.
In the case of electromagnetic flow meters the predominant frequency
is dictated by the magnetic field switching which may be in the
range of 3-10 Hz.
However, this very same technique, i.e. the relatively long time
constant has an undesirable side effect in that any relatively large
voltage pulse or spike (which can be produced as a result of a number
of different natural effects) pushes the amplifier into saturation,
and the long time constant means that the amplifier remains in saturation
for that very same long time. For example, a wide band amplifier
with time constant on the order of 1000 seconds will, once saturated,
remain in saturation for a time on the order of the same 1000 seconds.
During that period of time the amplifier, as well as all of the
remaining apparatus, is effectively disabled.
It is therefore an object of the present invention to provide an
electromagnetic flow meter which is not subjected to the disadvantages
of long time constants, notwithstanding the fact that it includes
a wide band preamplifier, wherein the wide band characteristic is
obtained by using elements with a relatively long time constant.
It is another object of the present invention to provide an electromagnetic
flow meter in which transformer effect coupling is minimized without
paying the price of subjecting the instrument to long periods in
saturation in response to a spurious noise pulse. Another object
of the invention is to provide a wide band amplifier for use with
periodic or simulated periodic signals which reduces the disadvantages
of other wide band amplifiers, i.e. saturation and long recovery
times. The manner in which the and other objects of the invention
are met is described hereinafter.
Summary of the Invention
In accordance with one aspect of the invention a preamplifier is
provided, coupled between the sensing electrodes of an electromagnetic
flow meter, and a sampling switch, which is timed in relation to
alternations of the magnetic field, which preamplifier is of the
wide band variety but which includes a switching means to stabilize
the operating point of the preamplifier to prevent the preamplifier
from remaining saturated for inordinately long periods of time in
response to excessive voltages. In accordance with this aspect,
the invention thus provides:
an electromagnetic flow meter comprising:
a clock,
a varying magnetic field means driven by said clock to produce
a varying magnetic field,
a pair of electrodes and means to support said electrodes in a
fluid flow field subject to a magnetic field produced by said magnetic
field means,
wide band amplifier means, with a pair of outputs, coupled to said
electrodes for changing impedance levels of signals derived from
said electrodes,
said wide band amplifier means including switching means driven
in synchronism with said clock to stabilize a bias level of said
wide band amplifier means,
combining means to combine signals from said pair of outputs, and
sampling switch means coupled to an output of said combining means
and driven in synchronism with said clock to sample an output of
said combining means in timed relation to operation of said switching
means.
In accordance with another aspect, the invention provides a switched
wide band amplifier which includes a high impedance input stage
such as, for example, an FET or vacuum tube, and a regulator stage
for minimizing loading on the input stage; this is achieved by AC
bootstrapping together nodes of said input stage. To stabilize the
bias level of the input stage, at least one terminal in the regulator
stage is, at times, switched to a selected potential level. Since
the switched terminal is electrically coupled to the input stage,
this ensures that a bias level of the input stage is controlled
to prevent the input stage from saturating. Thus, in accordance
with another aspect, the invention provides:
a signal source of a varying signal with a period P,
a high impedance active device with a control node coupled to said
signal source and two other nodes, AC bootstrapping means coupling
said nodes together, and
switching means in said AC bootstrapping means for stabilizing
a DC bias of said active device by driving at least one of said
nodes to a fixed potential at a rate fixed with respect to said
period P.
Brief Description of the Drawings
The present invention will now be further described in the following
portions of the specification, when taken in conjunction with the
attached drawings, to enable those skilled in the art to make and
use the same. In the drawings, like reference characters identify
identical apparatus and
FIGS. 1-3 illustrate respectively, longitudinal and cross-section
and a developed view of a probe type electromagnetic flow meter
in which the invention finds application;
FIG. 4A is a schematic and block diagram of a preferred embodiment
of the invention applied to the electromagnetic flow meter of FIGS.
1-3;
FIG. 4B is a timing diagram illustrating the temporal sequence
of selected signals in the circuit of FIG. 4A; and
FIG. 5 is a block diagram of the invention.
Description of Preferred Embodiments
FIGS. 1 2 and 3 illustrate respectively a typical longitudinal
section, cross section, and developed view of the probe portion
of an electromagnetic flow meter of the probe type with capacitively
coupled sensing electrodes. Although in this application the invention
will be illustrated as applied to a probe type flow meter which
is arranged to be inserted in a flow field to sense fluid velocity,
those skilled in the art will be aware, after reviewing this description,
that the invention can be applied to spool type flow meters as well
(in which the flow meter surrounds a flow defining conduit).
FIG. 1 is a longitudinal section of the transducer portion 20 of
the probe type electromagnetic flow meter. A housing or sleeve 22
which for example can be fiberglass is closed off at one end by
an end cap 24. Interior of the sleeve 22 is an electromagnet having
a core 50 and a winding 52. Leads Z1 and Z2 are connected to the
terminations of the winding 52. The winding 52 is potted and thus
a body 54 of insulating material surrounds the winding. A thin layer
56 of electrically conductive material such as a silver paint is
provided on the outer surface of the insulation 54 to serve as a
shield to shield the electromagnet from the detecting electrodes,
described below. The power leads Z1 and Z2 are a shielded twisted
pair including a cylindrical shield 60 of electrically conducting
material. The shield 60 is fared into the shield 56 to thoroughly
shield the electrodes and associated electronics.
For the electromagnetic flow meter shown in FIG. 1 an outer sheath
10 of dieletric material completely surrounds the probe. Interior
of the sheath 10 are a pair of detecting electrodes 30 and 32 disposed
opposite each other. Accordingly, the sensing electrodes are capacitively
coupled to the signal source. Flow meters of this type are sometimes
referred to as electrodeless. Each of the electrodes 30 and 32 is
surrounded by an associated guard electrode, 80 and 82 respectively.
Each sensing electrode and its associated guard electrode are connected
respectively to a conductor, for example sensing electrode 30 is
connected to a conductor X1 and the associated guard electrode 80
is connected to a conductor Y1; similar remarks apply to the sensing
electrode 32 and its associated guard electrode 82 connected to
conductors X2 and Y2 respectively. Each of these pairs of conductors
actually comprises a shielded pair wherein the Y conductor comprises
the shield for the associated X conductor.
FIG. 2 is a cross section of the electromagnetic flow meter taken
through the line 2--2 in FIG. 1. FIG. 3 is a developed view of the
shell 22 interior of the sheath 10. For the configuration shown,
the flow meter includes only a pair of sensing electrodes; however,
as described in my U.S. Pat. No. Re. 28989 two pairs of sensing
electrodes may be provided, each 90.degree. apart. Referring briefly
to FIG. 3 note that the guard electrodes, 80 and 82 are separated
from the associated sensing electrodes, 30 and 32 by an insulating
region 50. The angular distance between the center lines of the
two sensing electrodes can be 180.degree., although that is not
required; see my copending application entitled "Skewed Electrodes"
filed simultaneously herewith. Furthermore, as described in my cited
reissue patent, the flow meter can be produced without the sheath
10.
Power for producing the described electromagnetic field is coupled
over the conductors Z1 and Z2 and the signal conditioning electronics
is connected to the conductors X1-Y1 and X2-Y2.
FIGS. 4A and 4B are respectively a schematic of a preferred embodiment
of the preamplifier and an associated timing diagram.
Referring first to FIG. 4A, the preamplifier includes a fore amplifier
100 associated with one sensing electrode, a fore amplifier 200
associated with the other sensing electrode, and a differential
amplifier comprising operational amplifier 140 and 150. The differential
amplifier (140-150) has inputs coupled to the outputs of the fore
amplifiers 100 and 200 and an output coupled to other processing
circuitry. FIG. 4A illustrates a schematic of a preferred embodiment
of the fore amplifier 100 fore amplifier 200 is not illustrated
inasmuch as it is substantially identical to fore amplifier 100.
As shown in FIG. 4A, the sensing electrode and guard electrode
leads X1 and Y1 respectively, are coupled to the gate and source
of an FET Q1 respectively. Coupled in parallel is a gate leak resistor
R.sub.g. In addition to the lumped elements, a series capacitance
C.sub.o represents the capacitance between the sensing electrode
and the fluid being monitored, and a gate leak capacitance C.sub.g
represents the capacitance between sensing and guard electrodes.
In the "electrodeless" configuration of FIGS. 1-3 one
embodiment actually constructed had C.sub.o of about 100 picofarads
and C.sub.g was on the order of a few hundred picofarads. With an
R.sub.g of 10 terohms the time constant, R.sub.g C.sub.o is 10.sup.3
seconds. It should be apparent to those skilled in the art that
we use FET as a generic which accomplishes a function. The FET Q1
functions as an impedance changing element, changing the relatively
high impedance signal at its input, to a low impedance signal at
its output. Those skilled in the art will be aware that other active
devices exhibiting this impedance changing characteristic can also
be used, one such device is a vacuum tube in which plate, cathode
and grid correspond respectively to drain, source and gate. The
FET Q1 should be as close as possible to the transducer's electrode
and guard, and best incorporated in the transducer.
In practical implementations, the FET Q1 can be separated from
the rest of foreamp 100 by up to several hundred feet. The connecting
transmission line (connecting DR.sub.1 and SR.sub.1) may represent
a sizable capacitance to ground.
Amplifier 120 must drive this load. In addition, to prevent ringing
the damping resistor R13 is used.
As shown in FIG. 4A, the drain of FET Q1 is labelled DR.sub.1
and the source is labelled SR.sub.1. FIG. 4A illustrates the apparatus
connected to those terminals. More particularly, DR.sub.1 is connected
via resistor R.sub.13 (usually needed for stability if the cable
is long between Q1 and the remainder of the foreamp) to the output
terminal of an operational amplifier 120. That same output terminal
is coupled through a resistor R.sub.5 to the inverting input. The
non-inverting input is coupled through a resistor R4 to the output.
A voltage regulating diode D.sub.1 has a cathode coupled to the
non-inverting input of operational amplifier 120 and an anode coupled
to a resistor R6 whose other terminal is coupled to the inverting
input.
Those skilled in the art will recognize a conventional DC voltage
stabilizer or regulator but our regulator is driven by connecting
the anode of diode D1 to the output of an operational amplifier
110 which is directly coupled to its inverting input, the non-inverting
input is coupled to the terminal SR.sub.1. A resistor R1 (the source
resistance of FET Q.sub.1) is connected to the terminal SR.sub.1
and the other terminal of the resistor--the source resistor node--is
connected to the output of an operational amplifier 130. The output
of operational amplifier 130 is coupled via a capacitor C to its
inverting input and one terminal of a switch S.sub.1. The other
terminal of the switch S.sub.1 is connected via a resistor R.sub.3
to the anode of diode D1 and through a resistor R.sub.2 to the non-inverting
input of operational amplifier 130. The non-inverting input is also
coupled to one terminal of a switch S.sub.2 whose other terminal
is grounded.
Switches S.sub.1 and S.sub.2 are controlled via a switching signal
R, the development of which is discussed hereinafter. At this point,
it is sufficient to note that the switches S.sub.1 and S.sub.2 are
operated in synchronism with each other. When closed switch S.sub.2
grounds the non-inverting input of operational amplifier 130 and
switch S.sub.1 connects the inverting input of the operational amplifier
130 through resistor R.sub.3 to the anode of diode D.sub.1 and the
output of voltage follower 110.
The terminal SR.sub.1 is also coupled to the non-inverting input
of operational amplifier 140(which is one input of the differential
amplifier 140/150; the second input is the non-inverting input to
operational amplifier 150). To provide adjustment of differential
balance, the inverting input of amplifier 150 is coupled to a potential
divider comprising resistors R10-R12 between ground and the output
of the operational amplifier 140 and the inverting input of operational
amplifier 150 as well as R9.
For differential action, the output of operational amplifier 140
is connected through a resistor R8 to the inverting input of operational
amplifier 150. The output of operational amplifier 150 is connected
via resistor R9 to the inverting input of operational amplifier
150. The non-inverting input of operational amplifier 150 is coupled
to the output of the unillustrated fore amplifier 200. Finally,
the output of the differential amplifier 140/150 is coupled to the
input of a sampling switch 90 under the control of a control signal
CS, the development of which is discussed hereinafter.
The fore amplifier 200 is identical to fore amplifier 100; its
output is connected to the second input of the differential amplifier
140/150 i.e., connected to the non-inverting input of operational
amplifier 150.
The preamplifier (shown as a single operational amplifier in U.S.
Pat. No. Re. 28989) in accordance with the present invention comprises
fore amplifier 100 (for one channel or one sense electrode), a fore
amplifier 200 (for the other sense electrode), each coupled to the
differential operational amplifier 140/150. Each of the fore amplifiers
includes an input FET stage to achieve an impedance changing function.
The operational amplifiers 110 120 and 130 in fore amplifier 100
(and the corresponding amplifiers in fore amplifier 200) ensure
that the overall preamplifier exhibits an exceedingly long time
constant yet is capable of being synchronously reset to its correct
electrical operating point, and thus can be turned on and recover
from transients without a long delay. Before discussing the operation
of the fore amplifier reference is made to the timing diagram of
FIG. 4B. FIG. 4B shows four relevant waveforms. An upper clocking
waveform (clock) of FIG. 4B which for example can comprise the output
of an oscillator 310. The second line of FIG. 4B, labelled CS, is
the sampling strobe or control signal which controls the sampling
switch 90. The last line in FIG. 4B comprises the drive for the
electromagnet 52. The period of the clock is arranged such that
each polarity of the magnet drive occupies an equal and whole number
of clock cycles. The magnet drive is periodic in that it is of one
polarity for a first duration and rapidly switches to another polarity
for an equal duration. The sampling switch is closed (indicated
by the pulse in the waveform CS) just prior to the change in magnet
drive polarity and accordingly the sense electrode is sampled as
long after the alternation in magnetic polarity, as is possible.
The control signal R is located between the termination of the sampling
operation and the change in magnetic polarity, that is the switches
S.sub.1 and S.sub.2 are closed during the period of time the waveform
R is high, and the switches are open at other times. Thus, as shown
in FIG. 4B, the falling edge of CS is coincident with the rising
edge of R and the falling edge of R is coincident with the change
in magnetic polarity. As a precaution the timing strobes may be
modified such that the "coincidences" are separated by
several microseconds to guarantee: (1) sampling switch 90 is fully
opened before transient owing to reset R commences; and, (2) magnet
drive does not change until the reset transient in the fore amplifier
has died away. With the switches S.sub.1 and S.sub.2 in an open
condition the non-inverting input of amplifier 130 is coupled via
resistor R2 to the output of operational amplifier 110 which performs
a unity gain buffering function. As a result the non-inverting input
to operational amplifier 130 faithfully follows the potential on
SR.sub.1 the source of FET Q1. Since FET Q1 is coupled as a source
follower, this terminal follows the gate potential or the potential
on the sensing electrode.
On the other hand, when the switches S.sub.1 and S.sub.2 are closed,
the same non-inverting input terminal of operational amplifier 130
is connected to ground. Accordingly, operational amplifier 130 then
performs as an inverting integrator with a time constant of R.sub.3
C. As a result, its output (coupled to SR.sub.1 via resistor R1)
will reach equilibrium when the output of operational amplifier
110 has been reduced to ground potential. In other words, equilibrium
is reached when the output voltage of amplifier 130 has been driven
to that voltage which is required to place the source of FET Q1
at electrical ground voltage. Accordingly, at the conclusion of
the pulse R the FET is driven to an operating point such that, after
completion of each flow signal (coincident with the pulse CS) the
source of FET Q1 is reset to ground. This ensures that the FET is
not bootstrapped into saturation, and means that if saturated, the
period of this saturating interval will terminate rapidly, usually
with the next pulse R.
Following the reset strobe R, the magnet alternation takes place
and note that this occurs when the switches S.sub.1 and S.sub.2
are open. As a result, Q1's drain and its source resistor R1 are
fully bootstrapped and the sensing electrode is fully guarded; therefore,
fore amplifier 100 faithfully follows changes in the flow voltage,
even through the high impedance of the coupling capacitor C.sub.o.
By reason of the switching action of switches S.sub.1 and S.sub.2
the pertinent time constant for the fore amplifier 100 is infinite
(corresponding to the infinite resistance seen at the open switches).
In the preferred embodiment shown in FIGS. 4A and 4B the duration
of the pulse R is about 15-20 milliseconds. In this embodiment of
the pulse R is produced twice for each magnet cycle (or sampling
cycle). Since any perturbation of the sense signal caused by this
double frequency switching is symmetric, it should be rejected by
the sampling system. Under these conditions the R pulse duration
can be significantly shorter, e.g. 1-3 milliseconds or even less.
While the pulse duration R can be lengthened beyond 15-20 milliseconds,
there is no advantage. Furthermore, while in the preferred embodiment
the R pulse is produced twice per cycle, that rate can be reduced
so long as it remains symmetric. In other words, it can be produced
twice per two cycles, twice per four cycles, etc. Those skilled
in the art will understand how the R pulse of FIG. 4B is generated
by dividing the output of oscillator 310 and how other R pulse timing,
as recited above, can be effected by appropriate dividing and counting.
Accordingly, no further description of counter 320 is necessary.
Although the amplifier 140/150 has been described as a differential
amplifier, it should by now be apparent that the differential function
is not essential to the invention. In that regard, the amplifier
140/150 can be regarded as effecting a combining function, combining
the input signals derived from the sensing electrodes and producing
an output representing a combination of the inputs.
FIG. 5 is a block diagram of the invention. As shown in FIG. 5
the terminal DR.sub.1 is connected to the output of a voltage regulator
REG1 with inputs connected to + and SR.sub.1 respectively. A second
regulator, REG2 has an output terminal connected to R.sub.1 and
input terminals connected to SR.sub.1 and B-, respectively. Since
the regulators REG1 and REG2 faithfully output changes across their
inputs, the terminals DR.sub.1 and SR.sub.1 are AC bootstrapped.
A switch S is controlled to close by the reset pulse R to establish
a bias point for the active device connected to DR.sub.1 and SR.sub.1.
Comparing FIGS. 4A and 5 note that operational amplifier 110 is
a unity gain buffer (preferably gain of 1.000 or better) so that
REG1 corresponds to operational amplifier 120 and D.sub.1 and REG2
corresponds to operational amplifier 130 switches S.sub.1 and S.sub.2
in FIG. 4A correspond, in FIG. 5A to switch S. The advantages of
the invention are not restricted to the field of electromagnetic
flow meters but can be applied wherever a high impedance signal
must be faithfully followed via a wideband preamplifier exhibiting
rapid turn on and recovery from transients. The periodicity of the
input signal is necessary so that the perturbations caused by the
reset can be rejected. The flow meter signal is periodic since the
driving field is periodic, however in other applications the periodicity
requirement can be artificially introduced as is well known to those
skilled in the art.
While FIGS. 4A and 5 show the switches S.sub.1 S.sub.2 and S as
mechanical, those skilled in the art will realize that electronic
analog switches could be used so long as the transition from open
to closed (and vice versa) is suitably short and the impedance ratio
between open and closed conditions is suitably high.
The electromagnetic flow meter field is particularly demanding
of the electronics in that the source impedance can vary over a
wide range. The source impedance varies with the conductivity of
the metered fluid. Typical fluids and their conductivities are:
sea water 4 mhos/m; alcohol 10.sup.-3 mhos/m; hydrocarbon 10.sup.-13
mhos/m. In other less demanding applications the high impedance
input device, the FET described as exemplary, could be replaced
by bipolar transistors. |