Table of contents
3.Introduction
4.Specification requirements
5.Solutions suggestions
6.The magnetic fields sensors
7.The inclination measurement system
8.The gyroscope
9.The data acquisition system
10.Communication system
11.The power supply
12.Realisation of the PCB
13.The embedded system
14.Static Library Util.a
15.ViewPort
16.Xcompass
17.Sensors controller commands
18.Test
19.Future improvements
20.Conclusion
21.References
Navigation’s system has been designed for
the SMR and the MMR. For that reason, it will be supply by a 12V (SMR) or of
24V (MMR) battery. Consequently, we need to develop a power supply unit that
will handle these two cases.
All components but the flip coil that we
choose can be supply by a 5V voltage. Indeed, to work correctly, the Set /
Reset coil driver requires supply voltage between 16V and 20V. Consequently,
from a 12V or 24V, we have to supply a voltage of 5V and another one between
16V and 20V.
To fulfill this requirement, we envisaged
the following possibilities:
11.2.1. Suggestion 1
The first idea was to develop a
conversion’s system for each robot. The first one converted the 24V to 5V and
the voltage for the coil compensation will be get directly to the battery. The
second one converted the 12V to 5V that will be then converted to 20V.
Figure 20: Suggestion 1
This solution present more disadvantages
than advantages. This system will require two separated inputs and three
converters that will occupy a lot of place on the circuit and that are
expensive.
11.2.2. Suggestion 2
Another solution is to do every time two
conversions as illustrate the
Figure 21:
Figure 21: Suggestion 2
In this configuration, it is possible to
externalise the 24V to 12V converter. We would then have only one input plug.
Nevertheless, this type of conversion will not be efficient: to get the 18V
from the MMR, the 24V has first to be converted to 12V. In addition, this solution
presents the same problem of the previous suggestion: Use of three expensive
DC/DC converters on a small printed board
11.2.3. Suggestion 3
A third solution is to gather the two
systems to economise one conversion and to reduce the cost and the place occupied
by the supply circuit.
Figure 22: Suggestion 3
The principle of this operating system is
almost same as the suggestion 1 but with only two converters. A 24V regulator,
which is less expensive than a converter will replace the third one. The
conversion is almost the same when we use the SMR. Instead of converting the
voltage of the battery in 5V, we first convert it to 18V (to supply the coil
driver). The regulator is then also use to protect the circuit against the voltage
variation of the battery (it can vary between 21 and 29V).
The disadvantage of this circuit is it
requires two plugs (one for each robot) and that the efficiency would be low
for the MMR: the regulator will dissipate 6V every time.
11.2.4. Suggestion 4
The last possibility is to have only one
conversion’s system that commutes according to the input voltage.
Figure 23: Suggestion 4
This circuit directly converts the input
voltage of the SMR or the MMR in 5V. The supply voltage is then compared to 16V
(the minimum voltage requires for the coil driver). If it is not sufficient, a
second converter is switch on to provide the necessary voltage. A last
advantage of this suggestion is that the whole system could also be supplied by
5V.
This solution appears as the best
solution: it will just need one converter and one input plug.
The two last configurations have been
privileged to realise the power supply circuit because they require fewer places
on the board. The DC/DC converters are the biggest components of the system and
the dimension of the PCB is limited to 80 mm by 100 mm.
For the configurations 1 and 4, we need a
system to convert for the SMR the 5V to a voltage superior to 16V. In the application
note[2] of the
magneto-sensors, it is suggested to use the MAX662 for the 5V to 20V
conversion. This component usually provides 12V, but in a specific
configuration, we can get 20V. Its price is cheap (inferior to 5$) and avoid to
buy a true 5V to 20V converter which is more expensive.
It is also for this reason that the
solutions 2 and 3 have been abandoned: we did not found a cheap converter that
realizes directly the conversion 12 to 24V or 12 to 18V.
Then, the next choice concerns the way to
obtain the 5V output from 12V and 24V. We have selected the products of two
companies.
Name
|
NDY1205
|
NDY2405
|
NME1205D
|
NME2405D
|
Input voltage range [V]
|
9-18
|
18-36
|
10,8-13,2
|
21,6-26,4
|
Output current max [mA]
|
600
|
600
|
200
|
200
|
Noise [mVpk-pk]
|
100
|
100
|
110
|
150
|
Voltage set accuracy [%]
|
±1
|
±1
|
±1
|
±1
|
Efficiency [%]
|
82
|
82
|
80
80
|
Size (length*
Width*height) [mm]
|
32,26*14,66*11
|
32,26*14,66*11
|
11,5*9,8*10,8
|
11,5*9,8*10,8
|
Product
|
C&D
TECHNOLOGIES
|
C&D
TECHNOLOGIES
|
C&D
TECHNOLOGIES
|
C&D
TECHNOLOGIES
|
Price [Kr]
|
152
|
152
|
83
|
95
|
Table 10: DC/DC converters from C&D
TECHNOLOGIES
The first one provided the different
DC/DC converter for a great part of the robots in IAU. Consequently, they have
proved the quality of their components. Precisely they use the NDY series that
offers a good efficiency (82%) and a wide input range (18-36V or 9-18V). This
component can be implemented on the suggestions 1 and 3. We have looked for
another series less expensive. We have found the NME series (80Kr for each).
The efficiency is closed to the NDY series but the input range is narrower and
the output current is low for our application: We may need 3*50mA for the
compensation coil of the magneto sensor.
We found a third companies available on
the Farnell supplier. They present a series of component with interesting
characteristics:
Name
|
TEL 3-1211
|
TEL 3-2011
|
TEL 3-2411
|
Input voltage range [V]
|
9-18
|
10-30
|
18-36
|
Output current max [mA]
|
600
|
600
|
600
|
Noise [mVpk-pk]
|
60
|
60
|
60
|
Voltage set accuracy [%]
|
±1
|
±1
|
±1
|
Efficiency [%]
|
76
|
76
|
77
|
Size (length*Width*height) [mm]
|
32*20*14,5
|
32*20*14,5
|
32*20*14,5
|
Product
|
TRACO POWER
|
TRACO POWER
|
TRACO POWER
|
Price [Kr]
|
196
|
241
|
196
|
Table 11: DC/DC converters from TRACO POWER
The cost to realize the suggestion 1 will
be more expensive than with the previous converter. Nevertheless, the TEL
3-2011 converter has a wide input range to design the last suggestion.
To conclude the last solution appeared as
the best for the price, the size and the power consumed.
A 24V regulator and a 500mA fuse are used
to protect the system. The L7824C regulator (already available at IAU)
guarantees a fixed output voltage close to 24V from an input voltage in the
range 27-38V. In addition, two diodes have been included:
·
The first one between the regulator and the
converter to avoid negative voltage
·
The second one between the 20V and the input of
the Tel3-2011 to use this last one only if necessary
We also added different capacitors to
stabilise the supply voltage as well as a diode to signify that the system is
5V powered.
The 5 to 20V conversion is assured by the
MAX662AEPA.
Figure 24: Power supply unit
A comparator between the voltage V1 and
V2 (realised with a LM307N amplifier) commands the conversion.
V1 = VSupply/4.5
V2 = Vcc/2
The microcontroller could also be used to
make that comparison (see Figure 24: Power supply unit).
To avoid confusing the two input plugs,
we have implemented two different connectors:
-A connector with two pins
for the 5V supply (facultative)
-A connectors with three pins
for the other voltage
We now have the two voltage needed to
supply the flip coil driver and the reset of the components.
|