[0001] The present invention relates to a method and a device to recognize coins in a validation
unit. Such a unit is a part of a coin-handling device which function is to determine
if an inserted coin is acceptable or not.
[0002] A coin-handling device is an apparatus, which validates and accepts coins used as
payment mean usually integrated in a vending machine or a payphone. The coin-handling
device comprises a housing equipped with several openings including the coin entry
slot, which is accessible from the front panel of the machine. A coin inserted in
this slot gets into a runway generally inclined downwards before reaching a specific
region or part called validation unit or validator where the coin is analysed according
different parameters such as the dimensions and the material. In function of the analysis
results, the coin comes through another opening into an escrow and finally in a cashbox
(coin accepted), otherwise it is deviated to a coin return receptacle (coin rejected).
[0003] The invention concerns in particular a coin analysis method carried out by an electromagnetic
sensor specially designed for applying this method.
[0004] In the prior art, the coin passes close by a sensor made of coils mounted in the
wall of the runway which impedance is measured by a microprocessor driven unit. In
fact the coil impedance is influenced by the nature of the coin metallic alloy and
its volume (thickness and diameter). In some devices, the diameter could also be measured
with optic cells, which are masked the time of the coin passage.
[0005] Other sensors are based on a variable frequency oscillator, which provides a signal
through a coil. The presence of the coin changes the signal shape in function of its
characteristics. This response could also be collected by another coil coupled to
the first one like in a transformer.
[0006] In both type of sensors, the response signals are compared to references stored in
a non-volatile memory (EPROM, Flash memory). When the comparison results are within
or out of a predetermined window the coin is respectively accepted or rejected.
[0007] The sensor devices need a high precision mechanical and electronic hardware, which
is driven by programs loaded in a non-volatile memory. A coin validation unit has
usually to recognize several kinds of coins and to distinguish them from other ones
with similar characteristics. For example, a coin of a defined value in a country
must not be confused with another coin from another country having very close dimensions
and alloy, but with a different value. In order to minimise such errors, the device
becomes rather complex and requires high performances.
[0008] The aim of the present invention is to propose a coin recognition method and device
which main features lead to a high reliability, stability, and repeatability with
the lowest manufacturing costs.
[0009] The reliability is defined here as the capacity to have a small failure rate. The
stability is the resistance to environmental parameters changes such as temperature,
humidity and pressure. The repeatability is the ability to provide results as close
as possible on measures made successively on series of same coins.
- The aim is achieved by a method to recognize coins in a coin validation unit comprising
a runway for coin passage where is placed at least one inductive sensor in form of
a pair of coils between which a coin is going through, to said coils are applied input
electric signals which responses are analysed by an electronic circuit, characterized
in that
- the input electric signals comprises a sum of single sinusoidal signals with frequencies
having each a value f1, f2, f3, f4, ... ,fn according to the formula f2/f1=f3/f2=
... =fn/fn-1=2m or (2m+1)/2 where m is a positive integer, the first value f1 is determined
by the measurement cycle period,
- the responses corresponding to each input signal frequency represent after calculation,
values of the sensor conductance and of the sensor susceptance, said values are an
image of the coin characteristics.
[0010] The method is essentially based on calculations made on signal responses received
in presence of a coin traversing the area controlled by the sensor.
[0011] An advantage of such method is that the influence of the hardware components variation
is reduced. In fact, the generally used high precision analog or digital filters are
no more necessary therefore they are replaced by various mathematical operations performed
on signals.
[0012] The invention will be better understood with the help of the following detailed description,
which makes reference to the annexed figures, which are given as nonlimiting examples,
and describe the structure in which the method of the invention is implemented:
- Figure 1
- represents a block schematic of the device operation
- Figure 2
- illustrate the signals used in calculation
- Figure 3
- represents the sensor admittance complex diagram
- Figure 4:
- exploded view of the validator
- Figure 5:
- view of the validator right chassis
- Figure 6:
- view of the validator right chassis with the sensor mounted
- Figure 7:
- view of the right halves of the sensors
- Figure 8:
- view of the printed circuit supporting the sensors
- Figure 9:
- view of the left chassis with sensors mounted, without cover
[0013] The figure 1 shows the principle of operating of the coin-validating device. A quadratic
sinusoidal oscillator QO generates a pair of 90 degrees phase shifted sinusoidal signals
V1 and V2 of frequency f.
[0014] The first signal V1 is applied to the coil of the sensor SE via a power amplifier
PA in order to create a current in it. The current-voltage converter IUC converts
this current into a voltage V3. The oscillator output signals V1, V2 enter into a
comparator CP one after the other through a switch SW to provide a rectangular signal
according to the sign of each sinusoidal signal V1 and V2. The converter IUC output
V3 is multiplied respectively, through the synchronous detector SD, by each sign signal
issued from V1 and V2. Finally, the resulting signal is integrated during the input
signal period 1/f providing a number proportional respectively to the conductance
g and to the susceptance b of the sensor SE.
[0015] The conductance g is the real part of the complex number representing the admittance
Y (inverse of the impedance
Z,
Y=1/
Z) of the sensor, while the susceptance b represents the imaginary part.
[0016] Both values are directly influenced by the presence of the coin travelling on the
runway between the coils of the sensor. Each kind of coin differentiated by the dimensions
and the material will provide a specific pair of conductance and susceptance values.
[0017] The figure 2 shows the different signals at each step of the calculation. The oscillator
QO output V1 and V2 have the same frequency f but 90° phase shifted. The rectangular
signals SGN(V1) and SGN(V2) are provided by the comparator CP following the switch
SW. These signals have fixed positive and negative amplitudes in function of the sign
of the input signals V1 and V2.
[0018] The signal V3 is an image of the current flowing in the coil of the sensor SE. Its
phase, relatively to the input signal V1, which is a voltage, depends on the admittance
of the sensor in presence of a coin. It is known that a current in an inductance is
phase shifted compared to the voltage applied at its ends. The admittance has then
the same phase compared to the voltage than the current, according to the Ohm law.
[0019] This signal is then multiplied once by the sign signal determined by the first input
signal V1 and once by the second signal V2 determined by the second signal V2.
[0020] The value proportional to the sensor conductance g is then calculated by the integrating
analog / digital converter ADC providing the average of the produce of V3 by the sign
of the input signal V1 (SGN(V1).

[0021] The integration is made on one period T = 1/f corresponding to the frequency f of
the input signals V1 or V2.
[0022] In a similar way, the value proportional to the sensor susceptance b is calculated
with the second input signal V2.

The calculation of the sensor admittance
Y components (g, b) brings more possibilities to define the admittance than the simple
determination of the admittance modulus as it is usually performed in the prior art.
In fact, for a same modulus value, there is infinity of pairs of real and imaginary
part (g, b) combination.
[0023] This can be observed in the complex diagram of figure 3 representing a quarter of
circle (positive conductance and susceptance values) for a same admittance modulus
value Y, but with different arguments Φ (or phase) from 0° to 90°. The components
(g, b) are the projections of the vector
Y representing the complex admittance respectively on the real axis (Re) and on the
imaginary axis (Im) of the diagram.
[0024] In order to guarantee high precision coin measurements, the number of parameters
has to be adapted so that the results obtained can be differentiated with the ones
of other coins measurements. This is realised by the measuring of the sensor admittance
components g and b with several input signal V1, V2 frequencies f1, f2, ... fn. Each
coin passing close to the sensor provides a set of conductance and susceptance values
(g, b), one pair for each frequency. These values (g, b) are stored in a memory and
compared with a reference table or with a reference diagram showing each value of
the sensor admittance components g and b in function of the frequency.
[0025] In such case, the calculation circuitry needs n quadratic oscillators, 2n comparators
CD, 2n synchronous detectors SD and 2n integrating analog digital converters ADC.
[0026] One can demonstrate that when each frequency value f1, f2, f3, f4...fn of the input
signal (V1, V2) comply to the formula f2/f1=f3/f2= ... =fn/fn-1 =2m or (2m+1)/2 where
m is a positive integer, the circuitry can be drastically simplified. The number of
the mathematical operations as well as the number of calculation circuit elements
such as comparators, detectors and integrators are reduced.
[0027] For example for the fn/fn-1= 2m case, if the integration period is one quarter of
the period of maximum frequency and the integration is made on the signals that is
the response to the sum of input frequency, the combination of the integration's results
allows the separation of the different frequencies contribution. This is due when
contributions of a frequency are summed for a time equal or multiple of its period,
the result is equal to 0. For example in a case of 4 frequencies (f1, f2, f3, f4)
used, the summing of the first 8 integrations results with its sign, will eliminate
frequency f3 and f4 and remain only contributions of frequency f1 and f2; in the same
way it is possible to eliminate contributions of f2 and then obtain only f1 response.
[0028] The first value f1 (minimum value) is determined by appropriate response of the sensor
during the coin passage; maximum value is determined on the circuit accuracy, i.e.
Analog / Digital converter (ADC) speed.
[0029] The amplitude of the oscillators output signal varies in function of each frequency
because the current flowing in a coil decreases when the frequency increase. In the
aim to adapt the signal amplitude values used for calculations to the measuring range,
the oscillator signals are generated with an amplitude proportional to the frequency,
according to f
β where β is a parameter depending on the sensor design.
[0030] The present invention relates also to a coin validation device comprising a coin
runway, at least one sensor constituted by two symmetrical halves forming a support
disposed on each lateral side of the coin runway, a coil is mounted on each said half
in the way that one is facing the other and let a gap between the coils faces having
a width allowing the passage of a maximum thickness coin, calculation electronic circuitry
characterized in that
- both halves of the support form a closed magnetic circuit surrounding the runway,
the coils of the sensor are connected in serial mode,
- the electronic circuitry has means to calculate the sensor conductance and susceptance
at different frequencies in presence of a coin passing through the gap between the
coils of said sensor
[0031] The sensor constitutes the heart of the validator that is specially designed to recognize
coins according to above described method.
[0032] The figure 4 shows an exploded view of a coin validation device or coin validator
comprising two sensors placed one beside the other. Each sensor is made by two symmetric
support halves (3a, 3a', 3b, 3b'), carrying each one a coil, and mounted on the right
and left lateral side of the coin runway (1). The validator is disposed vertically
inside the host automate. Left (L) and right (R) sides are defined here as viewed
from the front part of the validator near the coin input. The coins enter into the
runway (1) through a slot on the topside (coin input Cl) of the validator and leave
the validator through another slot at its rear side (coin output CO). The validator
is constituted mainly by a right chassis (5) supporting the coin runway (1) and the
right sensor halves (not visible on the figure), a left chassis (4) mounted above
the coin runway (1) supporting the left sensor halves (3a, 3b). The sensor area of
the left chassis (4) is protected by a cover (6).
[0033] The figure 5 represents the right chassis (5) alone comprising the coins runway (1).
On both edges of the runway, openings (7a, 7b) are provided for the junction of the
right and left halves of each sensor.
[0034] Figure 6 shows the right chassis (5) with the sensors mounted in position. The coil
(2a, 2b) is mounted on each support half (3a, 3b) in the way that a coil is facing
the other mounted on the other half of the sensor located behind the runway wall.
The coin passes on the runway through a gap between the two coils having a width allowing
the passage of the maximum thickness coins. Both support halves of each sensor thus
assembled form a closed magnetic circuit surrounding the runway. The material used
for these coil supports is in general ferrite.
[0035] Figure 7 shows the right sensor halves (3a', 3b') alone with the coil (2a', 2b')
formed by an ovoid shaped kernel where the larger part is directed to the bottom edge
of the runway. This specific shape is adapted to the diameter range of the coins to
be validated.
[0036] Figure 8 shows a printed circuit board (8) supporting the two complete sensors. This
board is placed behind the runway wall of the right chassis (5) so that the runway
(1) passes at the level of the junction of each sensor support halves (3a, 3a', 3b,
3b').
[0037] Figure 9 shows the left chassis (4) with the mounted left sensor halves (3a, 3b).
This chassis (4) forms the left wall of the runway and supports the left halves (3a,
3b) of each sensor by maintaining them facing respectively each corresponding right
half (3a', 3b').
[0038] The two coils of one sensor are connected in serial mode, i.e. one end of the first
coil is connected to one end of the second coil. The two other ends are connected
to the measuring circuitry in order to be supplied by the signals coming from the
oscillator as described above.
[0039] In a preferred embodiment of the validator according to the present invention, the
two sensors are connected in a balanced or differential mode. The aim of such configuration
is to compensate the output signal measurement errors due to environmental parameters
changes as temperature or electronic component ageing.
[0040] When two sensors are connected in a balanced mode and environmental parameters have
the same influence on both sensors, the steady state response does not change with
environmental parameters. Thus, this configuration allows automatic compensation of
temperature changes or component's ageing.
1. Method to recognize coins in a coin validation unit comprising a runway for coin passage
where is placed at least one inductive sensor in form of a pair of coils between which
a coin is going through, to said coils are applied input electric signals which responses
are analysed by an electronic circuit,
characterized in that
- the input electric signals comprises a sum of single sinusoidal signals with frequencies
having each a value f1, f2, f3, f4, ... ,fn according to the formula f2/f1=f3/f2=
... =fn/fn-1=2m or (2m+1)/2 where m is a positive integer, the first value f1 is determined
by the measurement cycle period,
- the responses corresponding to each input signal frequency represent, after calculation,
values proportional respectively to the sensor conductance and to the sensor susceptance,
said values are an image of the coin characteristics.
2. Method according to claim 1, characterised in that multiple sinusoidal input signals are generated where all the frequencies start with
a well known phase relation, said signal is applied to the coil of the sensor creating
a current in said coil, this current is converted into a voltage by a current - voltage
converter,
3. Method according to claim 1 and 2, characterised in that active and reactive current components for each frequency is obtained by integrating
on a time interval that is a multiple of a quarter of the period (T/4) for each frequency
and combination of the partial results.
4. Method according to claim 1 to 3 characterised in that the relation between the frequencies is chosen according fn/fn-1 = 2, the output
signal is integrated on a quarter of the period Tk/4 corresponding to the maximum
value of the frequency fk for a duration corresponding to the period of lowest frequency
f1.
5. Method according to claims 1 to 4 characterized in that the integration on a quarter of period Tk/4 is numerically controlled, an analog
/ digital converter converts integration results in numbers proportional to real and
imaginary parts of conductance for all input frequencies after calculation, these
values are modified by the coin passage through the sensor generating a diagram
6. Method according to claims 1 to 5, characterised in that the amplitude of each input signals at used frequency is proportional to the frequency,
according to f β where β is a parameter depending on the sensor design, the output signal amplitude
being inside a determined measuring range.
7. Method according to claims 1 to 6, characterised in that each pair of conductance and susceptance values calculated with each input signal
frequency are introduced respectively in a diagram, particular points on each said
diagram are compared with previously memorised reference values corresponding to coins
accepted by the validation unit, a coin is then validated when the particular points
on each diagram match with their reference values.
8. Coin validation device comprising a coin runway (1), at least one sensor constituted
by two symmetrical halves forming a support (3a, 3b, 3a', 3b') disposed on each lateral
side of the coin runway (1), a coil (2a, 2b, 2a', 2b') is mounted on each said half
in the way that one is facing the other and let a gap between the coils faces having
a width allowing the passage of a maximum thickness coin, calculation electronic circuitry
characterized in that
- both halves of the support (3a, 3b, 3a', 3b') form a closed magnetic circuit surrounding
the runway (1), the coils (2a, 2b, 2a', 2b') of the sensor are connected in serial
mode,
- the electronic circuitry has means to calculate the sensor conductance and susceptance
at different frequencies in presence of a coin passing through the gap between the
coils (2a, 2b, 2a', 2b') of said sensor.
9. Coin validation device according to claim 8 characterised in that each coil (2a, 2b, 2a', 2b') of the sensor is formed by an ovoid shaped kernel with
the larger part directed to the bottom edge of the runway (1), said shape being adapted
to the diameter range of the coins.
10. Coin validation device according to claims 8 and 9 characterised in that both halves of the coil support (3a, 3b, 3a', 3b') of the sensor are made in ferrite
material.
11. Coin validation device according to claims 8 to 10 characterised in that it comprises two sensors connected in a balanced or differential configuration and
means to provide automatic compensation of temperature changes or component's ageing.