Solutions for Torque and Speed Measurement on
Electric Machine Controllers Test Benches
Soluciones para la medición de par y velocidad en bancos de ensayos de controladores
de máquinas eléctricas
A. Veyrat Durbex
†*1
, Y. Nachajon Schwartz
†2
, H. Tacca
†3
†
Laboratorio de Control de Accionamientos Tracción y Potencia (LABCATYP), Departamento de Electrónica, Facultad de
Ingeniería, Universidad de Buenos Aires, Paseo Colón 850, C1063ACV, Buenos Aires, Argentina
.
1
aveyrat@fi.uba.ar
2
ynachajon@fi.uba.ar
3
htacca@fi.uba.ar
*
Departamento de Energía, Facultad de Ingeniería, Universidad de Buenos Aires
Paseo Colón 850, C1063ACV, Buenos Aires, Argentina
Abstract— The continuous search for improvement in
electromechanical developments requires a clear
understanding of the torque measurement required in each
application. The decisions made will have a profound impact
on the quality and cost of the results. This article presents the
principles normally used for sensing mechanical torque on a
shaft and the reasons for the widespread use of strain gauges.
The importance of differentiation between static and dynamic
measurements and the current technologies used in each case
are discussed. The relevant factors that define a transducer are
described and a comparative analysis between them is carried
out, to then examine the possible mounting methods, with their
benefits and limitations. Finally, different low-cost solutions
are proposed for the design of the torque and speed
measurement section without sacrificing system performance,
for different test benches for electrical machine controllers,
including the necessary signal conditioning electronics.
The article aims to be a tutorial compendium of topics to
study to successfully implement a test bench without damaging
the torque transducer or introducing measurement errors,
since the information on these topics is scattered and it is
difficult to access knowledge of the selection criteria and
project procedures.
Keywords: Electric machines; torque transducer; torque
measurement.
Resumen— La continua búsqueda de la mejora en los
desarrollos electromecánicos impone una clara comprensión de
la medición de par necesaria en cada aplicación. Las decisiones
tomadas tendrán un profundo impacto en la calidad y el costo
de los resultados. En este artículo se presentan los principios
comúnmente empleados para el sensado de par mecánico en un
eje y el porqué de la utilización extendida de las galgas
extensiométricas. Se discuten la importancia en la
diferenciación entre las mediciones estáticas y dinámicas, y las
tecnologías actuales utilizadas en cada caso. Se describen los
factores relevantes que definen a un transductor y se realiza un
análisis comparativo entre ellos, para luego examinar los
posibles métodos de montaje, con sus beneficios y limitaciones.
Finalmente se proponen distintas soluciones de bajo costo para
el diseño de la sección de medición de par y velocidad sin
sacrificar el rendimiento del sistema, para diferentes bancos de
ensayos para controladores de máquinas eléctricas, incluyendo
la electrónica de acondicionamiento de señal necesaria.
El articulo pretende ser un compendio tutorial de temas a
estudiar para implementar con éxito un banco de ensayos sin
dañar al transductor de par ni introducir errores de medición,
ya que la información sobre esos temas está dispersa y es difícil
acceder al conocimiento de los criterios de selección y
procedimientos de proyecto.
Palabras clave: Máquinas eléctricas; transductor de par;
medición de par.
I. INTRODUCTION
The importance of rotating machines in the modern world
is well known. Their usefulness is evidence in almost all
areas of production, industry, commerce and daily life. In
order to satisfy the demands of increasingly efficient,
reliable and durable machines, the ability to accurately
measure the mechanical variables that determine the
performance of machines such as: power, torque and
angular speed is essential.
Online measurements of these quantities allow real-time
monitoring, help ensure consistency in product quality, and
can provide early indications of impending problems.
Torque and power measurements are used to test advanced
new machine designs and develop new machine
components [1].
In order to perform torque sensing, various physical
properties are used to convert, for example, rotations,
variations in mechanical stresses, or magnetic properties
into a proportional electrical signal. In general, the
modification of this property is manifested in a section of
the shaft on which the torque is exerted and measured.
Two types of torque transducers can be distinguished,
static torque transducers, which measure torque without
Recibido: 06/04/21; Aceptado: 08/06/21
Creative Commons License - Attribution-NonCommercial-
NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
https://doi.org/10.37537/rev.elektron.5.1.131.2021
Original Article
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
20
rotating or with limited angular movements, and rotary
torque transducers, which rotate as part of the system and
are suitable where dynamic torque measurements are
required.
For speed measurement, several methods are used, these
originate instruments such as mechanical or optical
tachymeters, strobe lamps and tachometric dynamos, but the
pulse rotary generator (encoder) is the most widely used,
due to its accuracy and resolution.
The test benches developed for various powers in the
Laboratory for Control of Drives, Traction and Power
(LABCATYP) dependent on the Department of Electronics
of the Faculty of Engineering of the University of Buenos
Aires, are intended to use a technique for developments
called RCP (Rapid Control Prototyping) that allows
software-implemented control algorithms to interact in real
time with variable drives for induction motors, application
examples can be seen in [2], [3].
Traditional methods of torque measurement based on
magnetic or hydraulic brakes, or dynamometric balances
provide static torque measurements. Even the transducers
that carry out dynamic measurements usually include a
display that gives the reading of the torque value.
The development of the solutions proposed in this work
is motivated by the need to have real-time torque
measurement on the RPC platform to fully exploit the
potential of the test bench, since the use in the control loop
of this torque signal allows the developer to evaluate,
compare and optimize his algorithms.
In this paper a comparative analysis of the different
technologies that are currently used to measure torque, is
carried out. The factors to be considered in the selection of
the indicated transducer, the mounting methods and
different design solutions are presented for the
implementation of the mentioned benches.
II. TYPES OF TORQUE TRANSDUCER
There are different techniques for measuring torque, one
of them is based on two toothed discs mounted on the ends
of a torsion bar on which inductive, capacitive or optical
sensors collect two signals whose phase will be proportional
to the torsion angle of the bar and therefore to the torque [1].
Another method uses a diagonally grooved bar on its
surface with two superimposed windings, one for excitation
and the other for sensing, which will detect the variation in
the reluctance of the grooved zone as a consequence of the
torsion of the shaft [4].
However, the use of strain gauges is the most widely
used method. The transducer consists of a flexible element,
on which are applied gauges arranged to flex in a
Wheatstone bridge configuration. The torque applied to the
sensor causes tension or compression of the gauges, varying
their resistance in proportion to the deformation and
generating an output voltage signal proportional to the
torque [5] when the bridge is energized. The benefit of
using this method lies in the negligible mass of the strain
gauges, which allows high accelerations and working
frequencies greater than 50kHz, ideal for both the
measurement of static and dynamic torques. In combination
with the properties of the bridge, it allows excellent linearity,
hysteresis and repeatability characteristics, as well as a
minimal effect of temperature and compensation for
interference caused by parasitic pairs [6].
The static or reaction torque transducer uses the reaction
torque, which is the turning force or moment, imposed on
the stationary part of a device by the rotating part, as energy
is delivered or absorbed. Power can be transmitted from a
rotating member to a fixed one by various means, such as
the magnetic field of a motor or generator, brake shoes or
pads on drums or rotors, or the lubricant between a bearing
and a shaft. Therefore, reaction torque sensors become
useful tools for measuring properties such as engine power,
braking effectiveness, lubrication and viscosity [7].
A typical example can be found with an electrodynamic
balance such as the one presented in Fig. 1. A brake,
whether by friction, magnetic or hydraulic, presents a
resistant torque on the motor shaft, the torque produced by
the motor generates an equal and opposite reaction torque in
the machine casing that is transmitted to the base, it is at
that fixing point where, installing the transducer by reaction,
the measurement is made.
Fig. 1. Measurement of reaction torque in a motor.
Rotating or dynamic torque transducers rotate as part of
the system, it can be said that they are reaction transducers
that have the freedom to rotate, they acquire the torque in a
rotating train of shafts, which is known as in-line torque
measurement. This method allows locating the sensor as
close as possible to the torque of interest and avoiding
possible errors in the measurement, such as parasitic torques
(bearings, etc.), extraneous loads, and components that have
large rotational inertias that would dampen any dynamic
torques [8]. Fig. 2 shows an example of a test bench for
electrical machines.
Fig. 2. Online torque measurement.
The transducers come prepared with different coupling
mechanisms, the most common being: cylindrical shaft
couplings and flange-type couplings. Those with a
cylindrical shaft, in turn, can be smooth or with a single or
double key.
In reaction transducers, the electrical connection
necessary to feed the strain gauge bridge, as well as to
collect the output voltage proportional to the coupling, is
trivial since the sensor does not rotate. In the case of rotary
transducers, there are several methods to connect the sensor
mounted in the rotary application with the fixed part where
the electrical connections are made. The system adopted to
make this link essentially determines the performance of the
transducer.
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
21
http://elektron.fi.uba.ar
A. Slip ring transducer
It has four conductive slip rings (composed of a copper
and silver alloy, with a 90% silver content) that rotate with
the sensor, and brushes that rub on the rings and provide
electrical contact with the fixed part. The bridge is excited
by two of the rings with the external power supply (in the
order of a few Volts) and the output signal is collected
through the other two (in the order of a few mV). The
brushes used in an instrument grade slip ring set are made of
a compound of 80% silver and 20% graphite, two brushes
per ring are used to ensure a positive electrical contact and
the same signal quality in both directions of rotation of the
shaft [9]. They are generally not recommended for
continuous use or for high speeds, due to the wear of the
brushes, which generates dust. This wear added to that of
the rings makes periodic cleaning and replacement of
brushes necessary, which discards them for use in
applications where maintenance is difficult.
B. Transducers with rotating transformers
Rotary transformers provide a means of electrical
connection without physical contact between the fixed part
and the rotating shaft where the sensor is located. These
transformers are similar to the conventional ones except that
one of the windings is rotating with respect to the other, the
windings are wound concentrically, with a coil rotating
inside or next to another stationary coil. For these
transducers, two rotary transformers are needed, one serves
to transmit the excitation to the strain gauge bridge, while
the second transfers the output signal to the non-rotating
part of the transducer. Due to the nature of transformers, AC
amplifiers are required for signal conditioning, increasing
the cost of the system.
C. Transducers with rotating analog electronics on the
shaft
This method uses rotating electronics on the shaft. As in
the previous case, there is no contact between rotating and
fixed parts, but in this case the power that is transmitted
with a rotating transformer is rectified and stabilized on the
shaft to excite the sensor. In addition, the torque-
proportional output signal of the bridge is amplified and
converted to variable frequency signal for transfer to the
stator. A more advanced version uses a single link between
rotor and stator, a fixed antenna transmits an alternating
voltage of the order of 20kHz from the stator that is
collected by a winding in the rotor, it is rectified and
stabilized to feed the bridge, the output voltage from the
bridge controls the frequency of the square-wave voltage, it
switches a carrier of the order of MHz that feeds a second
winding in the rotor. The antenna receives the carrier,
amplifies, filters and demodulates it to recover the
measurement signal [6].
D. Transducers with rotating digital electronics on the
shaft
A more complex variant involves digitizing the torque
signal on the rotor. This signal generated by the bridge is
filtered, amplified and digitized, then passed to a processor
that encodes it in the form of a serial word for transmission
to the stator through a rotating transformer [10]. Compared
to the analog signal, the measurement signal in digital form
is more robust against possible disturbances. This signal
enters a microprocessor and then, depending on the type of
transducer, it is converted into a voltage, current or digital
signal and reaches the output of the transducer where it can
be measured directly at the connector.
A separate rotor feature collects the power from another
transformer or other type of induction link, conditions it,
and then supplies it to the other electronics in the rotor.
Other variants of these transducers use an antenna to
transmit power to the rotor and telemetry through inductive
or infrared coupling to transfer the signals to the stator.
E. Special transducers
The dual-range transducer is an option when the
application demands a precise measurement of both the
operating torque and the torque peaks, these are generally
based on two bridge sensors mounted on the same axis,
where the one with the lower range receives an
amplification and filtering, which affects its bandwidth.
Realizing a true dual range transducer is physically
impossible as it would require the same torsion shaft to have
two different diameters to mount each gauge bridge on and
this would mean that the lower range one would be
destroyed by using the higher range.
The clamp-on type transducer consists of a pre-calibrated
flex bar mounted between two collars that are clamped
around the shaft, the precisely spaced edges of the collars
provide a measurement of shaft torque without modifying
the shaft train. Naturally this transducer uses telemetry to
communicate the measurement information to the fixed part.
Another alternative that manufacturers provide is the
transducer with only the sensing element on the shaft, but
without the signal conditioning electronics on the stator.
This offers a convenient solution from the point of view of
maintenance and supply of replacement for that user in a
position to develop this stage.
Usually, all types of rotary transducers have a variant that
includes an encoder. Convenient when there is a need to
measure speed.
III. RELEVANT FACTORS FOR THE SELECTION OF A TORQUE
TRANSDUCER
The transducer is the mechanically weakest link in the
shaft train, therefore it will fail in the first instance as a
consequence of a bad installation or bad choice of
equipment, and this could mean an extremely costly mistake.
To correctly select the torque transducer, the
specifications of the equipment must be evaluated according
to the mounting environment so that the entire system is
appropriate. In some cases, the installation scenario may be
redesigned to adapt it to the desired characteristics of the
transducer, at least in part, but this will not always be
possible, so the following factors should be considered:
Torque range. The choice of the range should consider
the behavior of the application. In addition to the nominal
torque of the driving machine and the load, which generally
represent the average value that it can deliver or receive
over time, the dynamic behavior must be considered. The
starting and braking couplings, as well as the pulsating
couplings in the case of reciprocating motors, coupling
jumps in electrical machines such as the typical star-delta
starting must be observed not only for the purposes of the
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
22
http://elektron.fi.uba.ar
measurement but to guarantee the integrity of the equipment.
Typically the safe overload range is 2 times the full scale
value (but the sensor life cycle is reduced) and the
catastrophic overload range is 4 times that value (sensor
failure). Manufacturers provide a simple calculation method
to determine a safety multiplier with respect to mean torque,
from the characterization of the loads and the driving
machines through a service factor.
Speed range. The nominal speed of the transducer must
be greater than the maximum speed of the system regardless
of the direction of rotation, to ensure that it does not suffer
damage. In general, the higher the torque range, the greater
the diameter and weight and the lower the rotation capacity.
Accuracy and resolution. When considering
measurement errors, non-linearity, hysteresis and
temperature effects should be evaluated individually. If the
temperature is stable during the test, the temperature errors
can be ignored, in the same way if the measurements are
made with increasing loads only, the hysteresis error can be
ignored, and for loads measurements near the full capacity
can depreciate the non-linearity [7]. If there is the
possibility of correcting the data using a calibration table,
the error in measurement can be minimized. An advanced
user can get rid of all errors except the non-repeatability
error. In relation to the resolution, the analog measurement
theoretically has infinite resolution limited only by the
signal-noise ratio of the system. Also, in the case of digital
transducers, the resolution in bits of their analog-digital
converters must be considered.
Mass and moment of inertia of the transducer. These are
decisive in the stresses that the bearings must tolerate, both
static, the weight they must support, and dynamic, due to
the higher mass the bending vibrations of the shaft are
greater. As a consequence of this, the natural frequencies of
bending of the shaft will be lower. While the moment of
inertia affects the angular acceleration that the system can
reach and the higher its value, the lower the natural
frequencies of shaft torsion.
Torsional stiffness and in other dimensions. The fact that
the torque measurement is performed from the conversion
of a mechanical variation to an electrical one requires the
existence of a certain flexibility of the shaft. In general,
flange torque transducers (method of coupling and
transmission of torque) of more compact construction are
more rigid in all directions than those with a cylindrical
shaft [6]. A higher stiffness is related to a higher overload
capacity, but at the cost of loss of resolution, it is also
associated with higher natural frequencies.
Maintenance requirements. In slip-ring based transducers
the brush wear must be checked and cleanliness is essential
to avoid scratches on the silver rings. Although the
accumulation of particles on the brushes is not desirable,
care must be taken when cleaning because a thin film of
graphite on the ring helps the lubrication process [9]. In the
cases of non-contact torque transducers, maintenance is
significantly less, especially if they do not have bearings.
Environmental factors. The application of the equipment
will determine the temperature conditions to which it will be
subjected, in some cases forced ventilation over heat
sources or a thermal shield over the transducer and in cold
environments, covers or heating enclosures must be
considered. Protection against foreign bodies and fluids
defined by the IP degree of the equipment must be
considered to guarantee correct operation. The
electromagnetic radiation from the environment must be
minimized to avoid interference in the sensor electronics,
the high voltage from the ignitions of internal combustion
engines, large electrical machines and AC and DC variable
speed drives are some typical sources of interference, a
successful installation should include shielded cables and
adequate grounding to improve immunity.
Torsional vibrations and in other dimensions. In addition
to vibrations in the sense of measurement, parasitic, bending
and axial vibrations, caused by imbalance, buckling or
pulsating acceleration of masses in reciprocating motors,
must be taken into account. In many cases these factors are
not relevant or can be solved by adopting a simple modeling,
but if the application requires greater precision or the
geometry of the system is very complex, a computer is used.
IV. ADVANTAGES AND DISADVANTAGES OF EACH
TECHNOLOGY
Reaction torque sensors: Advantages of this sensor: It is
suitable for any application speed. It does not require the
intervention of the shaft. It does not have the problem
(existing in rotating applications) of the electrical
connection to the sensor and it is also a low cost solution.
Disadvantages: Low dynamic response. Greater error when
measuring couplings caused by its own mass.
Slip rings: As advantages can be mentioned their low
inertia. Good dynamic response. Possibility of fixed or
floating mounting. Economical for low torque ranges.
Disadvantages: Low rigidity. Speed range limited by
contact. High maintenance due to brushes and bearings.
Important electrical noise with increasing speed. Low
accuracy. Difficulty in measuring low ranges as a result of
the drag torque generated by the brush-ring friction.
Backlash, since in general it uses a shaft with a key system.
Rotary transformers: The contactless transmission stands
out in this case. Low inertia. Possibility of fixed or floating
mounting. The limitation of the drag torque and the high
maintenance that the slip ring sensor had has disappeared. It
improves the speed range, however the use of bearings and
the brittleness of the transformer cores still limits the
maximum RPM. Economical, but more expensive than slip
rings. Disadvantages: You must use an alternating current
source. Limited bandwidth. Low stiffness. Still requires
bearing maintenance. Electrical noise and low accuracy,
consequence of the alignment of the transformer primary
and secondary. The existing air gap makes it sensitive to
vibration. Backlash, since it generally uses a shaft with a
key system.
Analog telemetry: It has the advantage of contactless
transmission. High stiffness. The stator can be disassembled
without disassembling the shaft train (versions with
antenna). Very low backlash since it generally uses a flange
coupling system. Low maintenance. Suitable for high speed
applications. High immunity to noise. Compact and
lightweight. Disadvantages: High inertia due to its larger
diameter. Limited dynamic response (better than rotary
transformers). Telemetry can be susceptible to interference
from nearby metals.
Digital telemetry: It has the advantage of contactless
transmission. High stiffness. The stator can be disassembled
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
23
http://elektron.fi.uba.ar
without disassembling the shaft train (versions with
antenna). Very low backlash since it generally uses a flange
coupling system. Low maintenance. Suitable for high speed
applications. High immunity to noise. Compact and
lightweight. High immunity to interference. Less
uncertainty. Great bandwidth. Possibility of software
configuration [11]. Disadvantages: High inertia due to its
larger diameter. High cost for low torque ranges.
Clamps: The main virtue of this type of sensor is that it is
not required to intervene on the shaft. Useful in applications
where quick installation and uninstallation are required. It is
useful for large shaft diameters. Economical for high torque
ranges. High speed operating capability. Its main
disadvantage is that it measures indirectly, so knowledge of
the shaft characteristics and mathematical calculations are
required. It has low accuracy and limited bandwidth.
V. MECHANICAL CONSIDERATIONS FOR ASSEMBLY
For a correct operation of the test system, the mechanical
installation must meet a series of requirements that
guarantee that the stresses on the transducer are kept within
the limits specified by the manufacturer. If not, the
measurements will be tainted by parasitic pairs, and what is
worse, there is a risk of damaging the transducer [12].
These requirements include the correct selection, for each
class of transducer, of the type of coupling, hardness and
resistance of the coupling materials, the proper mounting
method, the balancing and alignment of the rotating system.
A. Compensation element
The function of the couplings is to extend shaft
transmission lines or connect sections of different shafts,
whether or not they are aligned with each other [13], and
they play an important role in minimizing undesirable
stresses on the shafts by compensating for the inevitable
geometric errors of the assembly.
Flexible single joint couplings, also called half coupling,
are used to compensate for geometric errors. Static offset
resulting from alignment errors or errors associated with
imperfections in the components that generate off-centering
of the shafts. In general, these couplings are elastic in
relation to flexural deformation, but are rigid for torsional
deformation. Simple joint couplings can compensate for
angular misalignments, they fulfill the function of a
universal joint but unlike an ideal joint, they cause restoring
forces or couplings [6]. The limit of angular misalignment
they allow is low, generally much less than one degree. A
representation of this type of compensation element can be
seen in Fig. 3(a).
Flexible double joint couplings (Fig. 3(b)), also called
full coupling, can compensate for angular and parallel
misalignments, they fulfill the function of two
interconnected universal joints. Like simple joints, they
generate restoring forces or couplings. The limit of angular
and parallel misalignment they allow is low. The application
of this category of couplings is limited to compensating for
the minimum inaccuracies that are unavoidable in any
assembly process.
Joint shafts are made up of two joints and a straight shaft
between them (Fig. 3(c)), and can compensate for angular
and parallel misalignments. The principle of operation is
like the double joint, but does not generate restoring forces
or couplings. This kind of joint allows greater displacement
than the elastic elements of the couplings, which makes it
possible to compensate for large angular and parallel
misalignments.
Fig. 3. (a) Single joint. (b) Double joint. (c) Joint shaft.
To select the correct coupling capacity, it is not only
necessary to know the power and speed of the application,
but also its severity. To determine this severity, the coupling
manufacturers recommend a weighting that considers the
type of load, the type of machine drive, the number of starts
per hour, and the number of hours of daily operation [13].
B. Types of shaft couplings
Two large groups can be distinguished, rigid and flexible
couplings.
The former are made up of pieces that rigidly join the
shafts in such a way as to prevent relative movement
between them. They require lubrication frequently. Some
common types are plate or flange type, tapered clamp, and
stud sleeve couplings (Fig. 4(a)). These are highly
discouraged for the installation of torque transducers, since
they cannot absorb vibrations and taking into account that
both shafts must be perfectly aligned, correct operation of
the transducer, or even worse, its integrity cannot be
ensured.
Flexible couplings allow certain misalignments between
shafts, providing the necessary compensation for geometric
errors in the installation. There are a wide variety of these,
which including sliding metal parts or elastic elements.
(a) (b)
Fig. 4. (a) Rigid coupling [14]. (b) Bellows coupling [15].
Beam couplings: they are manufactured from a single
piece of aluminum or stainless steel with helical grooves.
They can be used as couplings with double joint
functionality for low torque applications, in the order of a
few tens of Nm, have low inertia and are free of backlash.
They have a relatively low torsional stiffness, due to the
cuts they roll and unwind, so in demanding conditions of
torque transmission, they generate positioning errors. They
do not require maintenance, they do not carry lubrication.
They compensate for important angular misalignments of
the order of 5º and radial and axial misalignments of less
than 0.5 mm.
Bellows Couplings: Consists of anodized aluminum or
stainless steel hubs and stainless steel bellows, an example
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
24
http://elektron.fi.uba.ar
can be seen in the Fig. 4(b). They are used in motion control
applications, servos, encoders, resolvers or stepper motors,
due to their backlash-free operation, low inertia and high
torsional rigidity. They cover torque ranges from tenths of
Nm to a few kNm. They do not require maintenance.
Because they are totally metallic, they excel in
environments with high temperatures. Their ability to
absorb vibrations, in general, is inferior to couplings with
polymeter inserts. They can be used as double joint, with
angular compensation capacities of the order of 2º and
radial and axial misalignments of tenths of a mm.
Flexible elastic band coupling: they consist of a rubber
band that is mounted on the coupling hubs and fixed by
means of clamping crowns and screws, (Fig. 5(a)). The
coupling can be made of natural rubber or chloroprene. It is
used in applications from a few Nm to tens of kNm for
uneven torques, vibrations or large misalignments. It has
double joint functionality with angular and radial and axial
offsets of several mm, but has very low torsional stiffness
and high rotational inertia.
(a) (b)
Fig. 5. (a) Flexible elastic band coupling [16]. (b) Spider coupling [16].
Star or spider couplings: it is formed by two identical
cast iron cubes connected by an elastomer ring elastic to
bending, but rigid to torsion, (Fig. 5(b)). The degree of
torsional rigidity (soft in general) and the capacity to
transmit power varies according to the material used. Nitrile
is the softest, Urethane has a 50% higher power transfer
capacity than Nitrile and Hytrel 200% higher. Reliable for
power transmission across the full range of services in
reciprocating motor or electric machine applications.
Available in a torque range of several tens of kNm and
speeds up to 20000rpm. They have low inertia. Its ability to
compensate for misalignments is low, less than 1º and tenths
of an mm. It has simple joint functionality.
Universal joints: a joint is made up of two yokes
connected to each other with a central cross, the movement
of the drive shaft is transmitted through the cross to the
driven shaft, allowing large angular misalignments of up to
25º in 1800rpm applications. It suffers from a major
problem, the transmission of speed and coupling from one
shaft to the other includes an oscillatory component, which
causes vibrations and wear. They require minimal
lubrication. Joining two universal joints on a straight shaft
(joint shaft functionality) mitigates the effect of non-
uniform rotary motion. If the deflection angles of both joints
are the same, the non-uniformity arising from each of the
two joints is canceled and the same drive shaft torque and
angular velocity occurs on the driven shaft. Although the
oscillatory component is maintained in the intermediate
shaft. Joint shafts are used when the design foresees that
individual sections of the shaft train will be offset or move
relative to each other. Also when greater compensation is
required than that provided by elastic couplings. Intended
for applications up to hundreds of kNm. They have backlash.
They are high cost.
Chain Couplings: These couplings consist of two hubs
that include a steel sprocket each and are connected by a
double chain, (Fig. 6(a)). Suitable for low speeds (due to
balancing difficulties) and aggressive environments.
Compact and simple. Very low inertia. Torsional stiffness
limited by deformation of the chain. They require
lubrication. High torque transmission capacity. They have
backlash. It can compensate for angular misalignments of
the order of 2º but radial of tenths of an mm.
Grid couplings: they consist of two cubes slotted on the
perimeter where a serpentine-shaped steel tape is inserted,
(Fig. 6(b)). The flexibility and torsional resilience of these
grating couplings help reduce vibration and dampen shock
loads, but provide you with reduced torsional stiffness.
Torque ranges from hundreds of Nm to hundreds of kNm.
Limited speed. They require lubrication. Very low angular
compensation capacity of less than 0.1º and radial of tenths
of an mm.
(a) (b)
Fig. 6. (a) Chain coupling [17]. (b) Grid coupling [18].
Pin and bush couplings: Consists of two cast iron or steel
hubs, connected by steel bolts sheathed with elastomer
shock absorbers, (Fig. 7(a)). They are especially suitable for
drives with special safety and reliability requirements.
Torque shock loads and changing loads are not a problem
for these flexible, compact and robust couplings. The steel
variant is also particularly suitable for high-speed drives.
Torque ranges higher than MNm. They have low rotational
inertia. Simple joint functionality. Angular compensation
capacity less than 0.1º and radial of tenths of an mm.
(a) (b)
Fig. 7. (a) Pin and bush coupling [19]. (b) Steel disk coupling [16].
Steel disk couplings: These are made up of packs of thin
high grade stainless steel discs. The plates are joined
together by pressed bushings and are attached to the cubes
with special screws. Some models are capable of spinning at
high speeds (40000rpm). They can transmit high torques
(2000kNm). They have high torsional stiffness, no backlash
and low inertia. Excellent balancing characteristics. Steel
multi-disc couplings require two sets of plates or plate packs
to emulate the functionality of double joints or joint shafts,
(Fig. 7(b)). The angular compensation allowed per pack of
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
25
http://elektron.fi.uba.ar
discs is up to 1º in models with 6 screws (0.2º with 10
screws) and the radial and axial compensation is a few mm.
Gear Couplings: Each hub has external steel gear teeth
that engage internal gear teeth from a flanged sleeve. The
flanges are screwed together. The coupling is normally
sealed. They require lubrication, except for some models
made of nylon that transmit low torque. They have very low
inertia and high torsional stiffness. High power density and
low weight. They have wear that can cause mechanical
backlash. They are capable of transmitting high torque
(MNm) with high speed. Capable of compensating angles of
more than 1º per gear and some mm of radial slippage. They
can function as a double joint. They are high cost.
In all cases the specified maximum offset values cannot
occur simultaneously. The allowed values are reduced if
angular, parallel and axial misalignments are combined
simultaneously.
C. Geometric errors in the shaft train
Due to the manufacturing and assembly tolerances that
are always present, the various shaft train components are
never fully aligned and centered with each other.
As mentioned, the transducer is the weakest element of
the set and must be able to withstand all the mechanical
power of the system, so any parasitic stress generated by
misalignment is undesirable. To avoid or at least minimize
these effects, the shaft train elements must be correctly
aligned and the corresponding compensation elements used.
The system must be balanced to reduce the harmful
effects of centrifugal force, which can mean forces on the
transducer bearings and machines, as well as vibration.
Manufacturers declare that, with careful machining of the
couplings, and considering the transducer is factory
balanced, no additional balancing is required for speeds
below 3000rpm.
VI. SHAFT TRAIN DESIGN TO INCLUDE A TORQUE
TRANSDUCER
For the design of the shaft train a basic requirement is the
static determination of the bearing configuration. This is a
requirement of good mechanical design, rather than torque
measurement. By definition, in mechanical engineering, a
bearing configuration is statically determined when all the
forces and moments on the bearings can be determined from
the knowledge of the external forces and moments applying
the laws of statics, without the need to determine the state of
deformation [6].
There is also a requirement for kinematic determination,
no shaft section or other component should move or tilt
without deformation required for this to happen.
In torque measurement applications (test benches), the
drive and load machine have a bearing configuration with
no degrees of freedom in relation to tilting angles. A
coupling of these two machines on a single shaft train can
only be statically determined by inserting compensating
elements with joint functionality that allow tilting
movements.
On the other hand, the inclusion of a bearing element in
the shaft train always requires additional couplings. The
installation of torque transducers with bearings can be done
in two possible ways, floating or fixed.
In floating installations the transducer is supported only
by its links to the machines, through the bearings, as
illustrated in Fig. 8. With the floating shaft there is only one
degree of freedom between each end of the torque sensor
shaft and the adjacent coupling shaft, therefore, couplings
with single joint functionality are sufficient to ensure a
statically determined bearing configuration. Some device,
usually a flexible conductive belt, prevents the stator from
rotating. This is also used to connect the transducer housing
to electrical system ground.
Fig. 8. Floating installation of a transducer with bearings.
This type of mounting is more tolerant of misalignment
errors between the driving machine and the load, since the
transducer installed between two simple joints behaves like
a joint shaft, and the allowable radial misalignment is
directly proportional to the distance between the points of
flexion. Another advantage is the reduction of undesirable
loads on the sensor, since the loads are transferred entirely
through the sensor with no effect on the bearings [5].
In fixed installations the transducer is fixed to the base or
bench. If the driving machine and the load are provided with
bearings and supported on their legs, then couplings with
double joint functionality or joint shafts are necessary to
ensure a statically determined bearing configuration, this
can be seen in Fig. 9.
Fig. 9. Fixed installation of a transducer with bearings.
This installation reduces the mass in suspension in the
couplings and can increase the critical speed of the shaft,
respecting the speed rating of the torque sensor. Half of the
weight of each coupling is supported by the torque sensor
shaft and the other half is supported by the drive and load
shafts. It is best for high rpm applications, and also for test
benches where machine changes may be required, as the
transducer remains supported and aligned. On the other
hand, it is necessary to avoid vibrations of the transducer
housing when it incorporates an encoder, to avoid errors in
the speed measurement.
The disadvantage of this method is that parasitic loads on
the bearings can cause heating and premature wear [5], [6].
When it comes to installing torque sensing flanges
lacking bearings, single sided double joint flexible coupling
or joint shaft is used. The radial and bending load due to the
weight are very low in this sensor, for this reason, in general
the assembly of Fig. 10 is used where the torque measuring
flange supports the joint shaft. However, the load limits
must always be respected with respect to parasitic loads,
and if necessary add a support with bearings to support the
weight of the joint.
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
26
http://elektron.fi.uba.ar
Fig. 10. Installation of a torque sensing flange.
VII. ANGULAR VELOCITY MEASUREMENT
The importance of accurately measuring the determining
mechanical variables in the performance of the machines
has already been mentioned, in addition to torque, angular
speed is essential in any application that involves a rotating
shaft and obtaining both in real time allows to know the
transferred power at all times.
According to the method used to detect a change in a
physical magnitude, for example, electromagnetic radiation,
magnetic field, electrical voltage or according to the type of
mounting used, different types of encoders arise with
different capacities, advantages and disadvantages. [20].
Mechanical: The mechanical encoder or potentiometer,
limited to one or several turns according to its construction,
makes it easy to obtain an absolute encoder, however, its
wear and permanent adjustment limit its use.
Optical: The optical encoder, built from an
electromagnetic radiation emitter, such as a diode or a laser
diode, a slotted disk and at least one photo-receiver, is
special for environments with a high degree of
electromagnetic noise, such as for example large engines,
since it is completely immune. Resolution and robustness
are not an issue when properly selecting disk materials or
light emitting diode. They can also be used in places with
tight spaces, using their reflective variant where the beam
does not pass through the slot (transmissive variant) but is
reflected off the disk and is collected by the photoreceptor
located in the same plane as the emitting diode. They can be
used in applications where absolute or relative velocity
measurements are needed, since they can be shaped
according to the number of photo receptors included, being
able to achieve a pseudo absolute realization using two
photo-receptors and a third photo-receptor that will collect
only one change per shaft turn. Among other options to
select them, access to information should also be mentioned,
which for different applications may be useful. In absolute
encoders, the signals of all the photo-receivers can be
accessed at the same time or you can also choose to obtain a
single serial signal, which delays obtaining the complete
data, although it also simplifies its processing because a
single line of information will have to be manipulated.
Other types of output for each channel can be push-pull,
open-collector or line driver, depending on the cabling to be
used or the system to which it will be integrated.
Magnetic: The magnetic type encoder, in contrast to its
optical pair, allows it to be used in environments where
there is dirt such as grease, dust, oils and even water.
Basically they can be found in two large groups, magnetic
or electromagnetic. The first deals with those that use a
permanent magnet on the rotating axis and then one or two
hall-type sensors collect the changes in the magnetic field as
the movement develops. Several problems of constructive
origin can occur in those of this type, the location of the hall
sensor with respect to the center of rotation is important,
therefore its perpendicularity also as they affect the
alignment and the quality of the final measurement.
Arrangements where a hall sensor that detects vertical
magnetic changes is usually preferred, since it is less
sensitive to the location, compared to the other type that
measures magnetic field changes in the horizontal. Due to
its inherently analog nature, the resolution it can achieve
depends on the digitization circuit, being able to have very
high resolution, accuracy and reliability performance,
although for very demanding applications the optical
encoder is still preferred, since when manipulating beams
lower latencies are handled. Then, the second group of
electromagnetic induction encoders work by detecting
changes in the magnetic field between an inductor coil and
another detector, depending on a disk attached to the rotor
that rotates jointly with the shaft to be measured. When the
disk acquires an irregular shape, the device varies the
reluctance of the resulting magnetic circuit. This class is
also preferred in applications where sealing is not needed, in
turn its bandwidth of up to 1.5 MHz makes it suitable for a
wide variety of uses.
VIII. PRACTICAL REALIZATION OF THE TORQUE AND
SPEED MEASUREMENT SYSTEM (1ST SOLUTION)
As part of a test bench for electrical machine controllers,
the design and implementation of the torque and speed
measurement system was carried out. In the case of torque,
it was required to appreciate transients and ripples, which is
why a dynamic torque transducer was chosen. Low drive
machine speed and moderate utilization factor eased
demands on sensor technology, this combined with a narrow
torque range reduced the options to traditional sensors
without electronics on the shaft. The classic slip ring
transducer was chosen, favoring its superior dynamic
response compared to the rotary transformer sensor,
accepting the disadvantages of a higher maintenance
requirement.
A. Mechanical installation of torque transducer
As a first step, the commitment made for maintenance
was evaluated. Equation (1) shows an empirical formula for
calculating brush life based on real field conditions and
laboratory tests [21]. At 20% of the original length the
spring can no longer be relied upon to hold enough tension
on the brush to ensure good electrical contact with the slip
ring.
RPM
1015.5
totalof 80% toHrs ;Brush ware
6
, (1)
Considering that the speed of the test bench machines is
around 1000 rpm, the wear of 80% of the brushes would
occur after 15500 accumulated hours of operation. Time
more than adequate for a laboratory application.
To define the measurement capacity of the transducer, the
typical compromise was presented, on the one hand
choosing a range that is too large, would mean that the
accuracy and resolution might not be sufficient for the
application, if instead, a range that is too low, the sensor
could be damaged by overload.
To select the correct capacity, the initial step is to
determine the normal running torque that should be
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
27
http://elektron.fi.uba.ar
measured. This can be done from the data of the
characteristic plate of the bench machine, which has a
nominal power of 3 kW at 715 rpm.
Knowing these data, the average torque in nominal
operation will be:
40Nm
2715rpm
60s3000WP
T
Nom_1
, (2)
The next step is to decide the factor of safety necessary to
avoid overloading the transducer. The type of load and
driving machine must be considered to determine the
maximum peak torque of the system.
In general, sensor manufacturers assign a service factor to
loads that is between 1 and 4, with 1 being the least severe,
applied to regularly operating devices. While factor 2 is
assigned to irregularly running machines or start-stop
devices. Factor 3 for loads that mean big shocks and / or
greater inertia. Factor 4 being the most severe and involving
heavy torque reversals.
For drive machines the service factor ranges from 0 for
regularly running machines to 4 for gas or diesel engines of
a few pistons. The three-phase motor controlled by a
frequency converter corresponds to a service factor equal to
1.
The maximum peak torque expected can be obtained by
affecting the average torque by the sum of the two service
factors [5]. The test bench used in this work will have as a
drive machine a three-phase motor controlled by a variable
frequency drive (service factor 1), and as load a direct
current generator (service factor 1), this means that the
maximum peak of expected torque could reach twice the
average torque in nominal operation.
Considering that the priority is torque measurement with
the highest accuracy and resolution (these qualities would
be halved by doubling the sensor capacity), that the systems
are in a controlled laboratory environment and that the
sensors withstand 100% of overload without damage, a
compromise was made to select a transducer with a range
only 40% greater than nominal torque. A 500 lbf-in
(56.5Nm) transducer was selected.
Fig. 11. Fixed mounting of the torque transducer.
In the choice of the installation method of the transducers
for this work, the decision to keep vibrations on the sensor
to a minimum prevailed, which is why it was decided to
make a fixed installation with a double joint on both sides,
also achieving the benefit of being able to change machines
without having to realign the entire shaft train.
To achieve this, a design was made in our laboratory
based on low-cost flexible couplings with an elastomer stars
[22]. Using two additional cubes, which were turned to
create two spacers and with two complementary elastomer
stars, the couplings with double joint functionality were
created, which ensure, for a test bench, a very adequate
durability, in a simple way and at low cost (Fig. 11).
The jaw couplings mentioned provide an economical
solution for standard power applications, absorbing
moderate shock loads and light vibrations. The material of
the elastic element is Nitrile that allows an angular
misalignment of 1º and parallel of 0.38mm [23]. Their
capacity is 105 Nm, resulting from affecting the nominal
power by the average service factor corresponding to
electric motors, for less than 10 hours of daily service,
according to the manufacturer's manual.
B. Torque signal conditioning
The environment in which the transducer will work is
subject to noise from various sources, AC and DC variable
speed drives. The noise generated by this equipment
radiates through the connecting cables and is coupled from
the windings of the machines to the shafts and through the
bearings to the housings of the machines. To minimize these
effects, the machines were grounded and the electronics to
condition the sensor signal were mounted in a metal box in
the vicinity of the transducer. The connection was made
with shielded cable and in turn the box was grounded.
Fig. 12. Conditioning of the torque signal.
The controller test bench uses a development technique
called RCP (Rapid Control Prototyping) which allows
software-implemented control algorithms to interact in real
time with the controlled plant hardware. A controller board
housed in a PC is responsible for providing the link. The
torque signal conditioner adjusts the signals delivered by the
transducer to match the board. It provides offset regulation
and an adjustable gain amplifier, since the transducer
provides an output between -5 V and +5 V proportional to
the coupling, CUPIN (CUP) in Fig. 12, and the board
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
28
http://elektron.fi.uba.ar
housed in the PC has inputs for voltages between -10 V and
+10 V (CUPIN).
Finally, the cabling from the conditioner to the test bench
was done with shielded cable to guarantee the integrity of
the coupling measurement.
C. Processing of angular velocity measurement.
The motor-generator group includes the possibility of
measuring rotation speed by various methods. One of these
is based on a permanent magnet direct current generator
(tachometer dynamo), another method is based on a pulse
rotary generator (encoder), both are coupled to the set shaft
and each generate a signal, analog in a case and digital in
another, which are conditioned to use in real time on the test
bench.
Fig. 13. Tachometric dynamo coupled to the shaft.
The analog measurement with the dynamo (Fig. 13)
provides a fast responding DC voltage signal, but contains
the ripple of the mechanical rectification of the collector.
Being common the use of this method in the industry, it is
interesting to evaluate the behavior of the tested control
models when it is applied, and to compare them with the
behavior that results from the use of digital measurement.
The 3 kW bank is fitted with a 60V / 1000 rpm dynamo [24]
with a ripple factor of 0.2%.
Fig. 14. Conditioning circuit for the signal from the dynamo.
In Fig. 14 you can see the conditioning circuit for the
signal coming from the dynamo. It basically contains filters
for electromagnetic interference, both common mode and
differential mode, and an adjustable attenuator stage to scale
the voltage level to that allowed by the input A / D
converters of the controller board.
The digital speed measurement in this 3 kW set is
performed by an encoder included in the dynamic torque
transducer [25]. The signals delivered are conditioned by
the circuit of Fig. 15 which then supplies them to the test
bench where they are treated by a circuit similar to that used
in solution 3 (see below in section X).
Fig. 15. Recomposition of encoder signals.
IX. PRACTICAL REALIZATION OF THE TORQUE AND SPEED
MEASUREMENT SYSTEM (2ND SOLUTION)
Using criteria similar to those mentioned in the previous
case, a slip ring type transducer was selected and a fixed
installation was carried out on a bench with a nominal
power of the impulse machine of 1CV (736W) at 910 rpm,
the nominal torque was compute in (3).
7.7Nm
2910rpm
60s736WP
T
Nom_2
, (3)
The safety factor was determined and a transducer with a
range approximately 50% greater than the nominal torque
was selected, a transducer of 100 lbf-in (11.3Nm).
Fig. 16. Tachometric generator, encoder and metal box with electronics.
Jaw couplings were used and for installation of the
transducers a fixed mounting with double joint couplings
was made.
Fig. 17. Encoder with flexible beam coupling.
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
29
http://elektron.fi.uba.ar
The rotation speed measurement was carried out with an
old tachometric generator [26] of 100V / 1000rpm. (Fig. 16).
As the transducer did not include an integrated encoder, the
digital speed measurement was carried out with a rotary
pulse generator of 2500 p / r [27] linked by means of a
flexible beam type coupling as can be seen in Fig. 17.
The electronics used are similar to that of the previous
case.
X. PRACTICAL REALIZATION OF THE TORQUE AND SPEED
MEASUREMENT SYSTEM (3RD SOLUTION)
This time the drive machine is a squirrel cage three-phase
induction motor, ½ HP, 50Hz, 4 poles (Siemens) of
1370rpm. The nominal coupling is determined in (4).
2.6Nm
21370rpm
60s370WP
T
Nom_3
, (4)
The torque transducer is rated at 50 lbf / in (5.64Nm),
max. 3000rpm (Sendev). In this case, the equipment does
not include the electronics to manage the link with the
sensor bridge. From the datasheet the sensitivity available is
2mV / V at full scale [28]. In other words, for each Volt the
Wheatstone bridge is supplied with, 2mV will be obtained
at the output, if the maximum torque is applied.
Simple toothed couplings with an elastomer of
intermediate hardness were used in our laboratory for
assembly to allow power to be transmitted without
sacrificing the measurement bandwidth [29], [30]. The
installation used fixed anchor and double joint couplings as
can be seen in Fig. 18.
Fig. 18. Assembly with double couplings and fixed anchor.
Since the transducer signal is weak to be digitized by the
driver board, a conditioner was designed to amplify it,
which also supplies the voltage to energize the Wheatstone
bridge. Always avoiding electromagnetic interference from
the environment, which is important.
The first design factor is the location of the conditioner as
close as possible to the sensor, a housing was installed
inside which the entire circuit was housed. This guarantees
that the path of weak signals is minimal, protecting them
from possible interferences. The aforementioned casing acts
as a Faraday cage that covers the entire conditioner circuit,
thus protecting it from electromagnetic radiation that could
disturb the operation of the circuit.
Then, despite having amplified the weak signals, the
cabling to the controller's AD channel was done with
shielded cable to further protect against interference.
The electronic circuit includes a dedicated power supply,
which draws power from a transformer with electrostatic
and electromagnetic shields, to prevent interference
conducted in the power supply from affecting the
measurement.
Fig. 19. Conditioner for torque signal.
The embodiment is based on an application specific
monolithic integrated circuit (INA125) (see Fig. 19). The
circuit provides the bridge with a stable, precise and
temperature compensated power supply while its integrated
instrumentation amplifier allows to modify the gain with a
single adjustment. In this case the gain is adjusted by RV5
according to (5) [31].
RV5
R4
60K
4Gain
, (5)
As a margin of safety, the conditioner gain was adjusted
by simulating a signal greater than the maximum of the
instrument by about 10%. By varying RV5, a gain of 426
was obtained to obtain a maximum voltage of 9.5V at the
output. To achieve the adjustment of the conditioner gain,
all the jumpers were connected and the variable resistor
RV2 was regulated, previously disconnecting the transducer.
For speed, two measurements were used, an analog one
similar to that addressed in the first development and a
digital one from an encoder of 1024 pulses per turn.
Fig. 20. Conditioner for the encoder signal.
As it is common to find in the industry measurements of
rotary pulse generators of lower resolution, as can be the
typical case of a gear with a magnetic sensor, a circuit
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
30
http://elektron.fi.uba.ar
capable of emulating the operation of said poorer
measurement was made, with a divider of pulses made with
a configurable binary counter so that by means of jumpers
two signals can be chosen to send to the control platform,
one of them is the full resolution measurement and the other
is the reduced resolution one. Then, the voltage levels are
adapted to those compatible with the digital inputs of the
controller board (Fig. 20).
To detect the direction of rotation, a D flip-flop is used
that uses the direct signal as input and the quadrature signal
as clock. Thus, when the motor rotates forward, Sd leads Sq
causing, on each rising edge of Sq, the flip-flop copies a
logic "one" at its output. On the other hand, when turning in
reverse, Sq leads Sd and therefore the output is a logical
“zero”.
Comparing the three solutions, it can be seen that the 1st
solution is the one with the highest cost because it includes
the amplification electronics and the encoder in the
transducer. It requires less implementation effort and fewer
resources since they are mostly mechanical realizations,
electronic implementations are few. The required expertise
is high. For the 2nd solution, the cost is reduced by using an
external commercial encoder, instead of purchasing a
transducer with an included encoder. Increases the demand
for work and resources used in assembly. The required
expertise is high. In the 3rd solution the cost is the lowest
since the torque transducer only includes the sensor. The
amplification and signal conditioning are user-developed.
The demand for labor and resources used in the assembly
are maximum. It requires a very high degree of knowledge
of the subject. The dynamic response in all cases was
greater than that demanded by the time constants inherent to
mechanical systems.
XI. CONCLUSION
The reliability of the results of a torque measurement and
the budget allocated radically depend on the understanding
of the conditions demanded by the application, accuracy,
speed, bandwidth; as well as the choice of the transducer
capable of satisfying them. Generally, a compromise will
have to be appealed by selecting the transducer that meets
the least cost.
In addition to the characteristics of the equipment,
environmental factors and details of the installation must be
considered. Dirt in the environment, high temperature,
interference, can be some of the problems to consider.
Space limitations can be decisive in the choice of transducer,
as well as restrictions in mechanical mounting. The design
of auxiliary electronics must always consider
electromagnetic interference from the environment.
The torque transducer is the weakest link in the axle train,
not addressing all the variables can mean, not only faulty
measurements, but also equipment breakdown.
In this work, low-cost solutions were implemented for
the design of test benches for using Rapid Control
Prototyping. Various proposals for future work can be made
taking advantage of this platform. Performance analysis of
different PWM algorithms or study of torque disturbance
recognition algorithms that make it possible to distinguish
between transients and ripples are some of the possible
suggestions.
ACKNOWLEDGMENT
This work has received funding from the University of
Buenos Aires with funds from the subsidy UBACYT
20620170100006BA
REFERENCES
[1] I. J. Garshelis, “Torque and Power Measurement”, in The
Measurement, Instrumentation and Sensors Handbook, J.G. Webster
(Ed.). Boca Raton: CRC Press, 1999.
[2] J. Khan, "Rapid Control Prototyping (RCP) solutions for the
validation of motor control applications," 2016 International
Conference on Emerging Technological Trends (ICETT), 2016, pp.
1-6, doi: 10.1109/ICETT.2016.7873699.
[3] E. Quintero-Manriquez, E. N. Sanchez, R. G. Harley, S. Li and R.
A. Felix, "Neural Inverse Optimal Control Implementation for
Induction Motors via Rapid Control Prototyping," in IEEE
Transactions on Power Electronics, vol. 34, no. 6, pp. 5981-5992,
June 2019, doi: 10.1109/TPEL.2018.2870159.
[4] Fleming, William J. “Magnetostrictive Torque Sensors -
Comparison of Branch, Cross, and Solenoidal Designs.” SAE
Transactions, vol. 99, 1990, pp. 393–420.
[5] K. Skidmore, “Torque Measurement Primer. Interface Advanced
Force & Torque Measurement, Interface Inc”, 2010.
[6] R. Schicker, G. Wegener, Measuring Torque Correctly, Hottinger
Baldwin Messtechnik GmbH, 2002.
[7] “Technical Information-Torque Sensor, Test & Measurement
Sensors & Instrumentation”, PCB Load & Torque, Inc., PCB
Piezotronics, 2011.
[8] D. Schrand, “The Basics of Torque Measurement”, Technical Notes
and Articles. Sensor Development Inc., 2006.
[9] “High Quality Instrument Grade Slip Ring Assemblies A Technical
Discussion”, Technote 9504/N022, SensorData Technologies, 2014.
[10] “Plug and Play” USB T25 Torque Sensor, Operation Manual,
Interface, 2009.
[11] M. Minda, “How to Choose a Torque Sensor”, Hottinger Brüel &
Kjaer, 2020. [Webinar]. Available:
https://www.hbm.com/en/9016/webinar-how-to-choose-a-rotating-
torque-sensor/
[12] Rotary Torque Transducer Installation Guide, Sensor Technology,
TorqSense.
[13] V. Quilodrán Jopia, “Acoplamientos Mecánicos”, Ingeniería de
Ejecución Mecánica en Mantenimiento Industrial, Universidad
Tecnológica de Chile. INACAP.
[14] (2021) Mecapedia-Acoplamiento de manguito [Online]. Available:
http://www.mecapedia.uji.es/acoplamiento_de_manguito.htm
[15] Backlash-free applications-easily solved, Siemens, 2015.
[16] Flexible couplings, Grupo Oria, 2016.
[17] Acoplamiento de cadena 10B Z16, de Gier Drive Systems.
[18] J. Piotrowski, Shaft alignment handbook, 3rd ed., CRC Press, 2006.
[19] Flexible couplings, Rupex Series, Flender Couplings, 2020.
[20] Asahi-Kasei, “Basic Knowledge of Encoder”, Tutorials, Industry 4.0.
[Online]. Available: www.akm.com/global/en/technology/technical-
tutorial/basic-knowledge-encoder/type-mechanism-1/
[21] “Slip Ring and Slip ring Brush Maintenance”, Technote 9812/N049.
SensorData Technologies, 2014.
[22] A. F. Veyrat Durbex, “Cargas activas para un banco de ensayos de
control de motores de inducción trifásicos,” thesis, Universidad de
Buenos Aires, march of 2015.
[23] Acoplamientos de mandíbula, Catalogo General, SKF.
[24] “Dínamo taquimétrica de C.C. Mocbos Modelo DT60/10, Chapa
característica”, Motortech .
[25] “Model 01224-052, S/N 173852, Calibration data sheet”, Sensor
Development.
[26] “Tachometer generator, Chapa característica”, General Electric.
[27] “Rotary encoder model OEW2-25-2MD, Hojas de datos”, Nemicon,
Nidec Nemicon Corp.
[28] “Installation – Model 01192, Data sheet”, Sensor Development.
[29] Y. Nachajon, P. Witis, G. Bongiovanni, H. Tacca, and F. Ferreira,
“Banco de ensayos para algoritmos de control para motores de
inducción trifásicos”, SAAEI, Guijón-España, 2006.
[30] Y. Nachajon Schwartz, P. Witis, and G. Bongiovanni, “Banco de
ensayos para algoritmos de control para motores de inducción
trifásicos”, AADECA, 2006.
[31] “INA125, Data sheet”, Burr-Brown Corporation, February 1997.
Revista elektron, Vol. 5, No. 1, pp. 20-31 (2021)
ISSN 2525-0159
31
http://elektron.fi.uba.ar
Enlaces de Referencia
- Por el momento, no existen enlaces de referencia
Copyright (c) 2021 Alejandro Fabio Veyrat Durbex
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Revista elektron, ISSN-L 2525-0159
Facultad de Ingeniería. Universidad de Buenos Aires
Paseo Colón 850, 3er piso
C1063ACV - Buenos Aires - Argentina
revista.elektron@fi.uba.ar
+54 (11) 528-50889