Wireless control and measurement system for a hydropower generator with brushless exciter

(1)

TVE15056

Examensarbete 30 hp Juni 2015

Wireless control and measurement system for a hydropower generator with brushless exciter

Fredrik Evestedt

(2)

Teknisk- naturvetenskaplig fakultet UTH-enheten

Besöksadress:

Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0

Postadress:

Box 536 751 21 Uppsala

Telefon:

018 – 471 30 03

Telefax:

018 – 471 30 00

Hemsida:

http://www.teknat.uu.se/student

Abstract

Wireless control and measurement system for a hydropower generator with brushless exciter

Fredrik Evestedt

Hydropower has been around for more than a century and is considered a mature technology, but with recent advancements in power electronics and simulation capability new exciting ways to increase efficiency and reliability is possible. At Uppsala University a brushless exciter has been constructed for the experimental test rig, SVANTE. Power electronics are mounted on the shaft for control of the

generator's excitation current. In addition a wireless control and measurement system is needed to provide the desired switching patterns to the power electronics and to evaluate performance of the system.

In this thesis a shaft mounted embedded system for control and measurement is constructed as well as magnetic field sensors with measurement range up to 700mT.

The computational power comes from a National Instruments sbRIO-9606. The system has 14 individual totem pole power electronics driving channels, 48 analog input channels for current signals and it communicates wirelessly through a bluetooth connection.

The system is tested and works satisfactory but has not been mounted on the rotating side of the generator due to delays in the manufacturing.

Ämnesgranskare: Urban Lundin Handledare: José Perez

(3)

Sammanfattning

Vattenkraft har det senaste ˚arhundradet varit och är fortfarande en av Sveriges främsta källor till elektrisk energi. Tekniken som vattenkraft bygger p˚a är mogen och driftsäker, men med den senaste tidens utveckling inom kraftelektronik och simuleringsverktyg öppnas nya möjligheter att

¨

oka effektivitet, styrbarhet och drifts¨akerhet.

P˚a Uppsala Universitet finns en experimentgenerator, SVANTE, d¨ar nya tekniker kan utv¨arderas.

P˚a generatorns axel har en borstlös matare monterats och för att kunna köra denna krävs kraftelektronik för att kontrollera exciteringsströmmen till rotorn. Kraftelektroniken behöver i sin tur re- gleras samt att olika storheter s˚asom spänning, ström och magnetfält behöver mätas.

Detta examensarbete handlar om konstruktionen av ett inbyggt system för att tr˚adlöst hämta in mätdata fr˚an sensorer samt styra kraftelektronik som sitter monterat p˚a axeln i ett vattenkraftverk.

Detta implementeras i programvaran LabVIEW fr˚an National Instruments p˚a en sbRIO-9606.

Utöver detta konstrueras kretskort för mätning av magnetfält upp till 700mT.

Det konstruerade systemet fungerar tillfredsställande och testas genom att en växelriktare samt en DC-DC omvandlare styrs fr˚an systemet. Magnetfältsensorerna fungerar bra över hela mätomr˚adet med bra linjäritet och mätnoggrannhet. Allt som allt har fyra kretskort designats och utvärderats dessutom har LabVIEW-kod skrivits.

(4)

1 Introduction

The first modern hydropower plant in Sweden was constructed at the end of the ninetieth century, it was situated in the river Viskan and could produce 2.2kW. An extremely small generator by today’s standards but it was the start of an enormous exploitation of the Swedish rivers. Today, the total installed power is 16.15GW and in 2013, hydropower supplied 60.8TWh of energy to the Swedish grid. This corresponds to 41% of the total energy consumption [1].

The majority of today’s big hydropower plants were built in the 1950s and 1960s as a result of the construction of a 400kV transmission line from Harspr˚anget to Hallsberg. The power plants are getting old so refurbishment and upgrades are required to keep up with today’s standards. In this process new technologies can be implemented to potentially increase reliability, controllability and efficiency.

1.1 Background

At the Division of Electricity, Uppsala University, an 185kVA experimental generator called SVANTE is available. Specifications of the machine can be seen in Tab. 1 [2].

Table 1: Main specifications of SVANTE.

Frequency 50Hz

Number of pole pairs 6

Speed 500rpm

Slots per pole and phase 3 Number of stator slots 108 Stator inner diameter 725mm

Stator length 303mm

Air gap length 8.3mm

Power of driving motor 75kW

Rotor weight 900kg

Stator weight 700kg

At the moment upgrades are done to facilitate new research projects, below is a list of the new additions.

• New shaft lathed to fit the new additions.

• A six-phase brushless exciter.

• Permanent magnet thrust bearing.

• Electromagnetic thrust actuator.

• Power electronics for active control of the exciter and rotor currents.

• Sensor and control electronics.

(7)

A CAD-drawing of the experimental rig with the new components installed can be seen in Fig. 1.

Figure 1: CAD-drawing of SVANTE with the new shaft [3].

1.2 Project description

With the installation of a brushless exciter on the shaft, there is a need for a way to rectify and control the excitation current in the main generator. An embedded control and measurement system mounted on the shaft is needed for for this purpose, the system shall communicate wirelessly to a main control unit and be able to drive power electronics devices.

(8)

about the construction of the control and measurement systems while the other part is about the power electronics and the related control scheme.

A specification of requirements is presented below.

• Simultaneous sampling of currents and voltages relevant to the control of the active rectifier and the buck converters.

• Enough processing power to implement space vector modulation.

• Accurate measurement of magnetic field up to 700mT on the rotor poles.

• Magnetic field measurements of 28 positions on the shaft.

• Wireless communication to the system.

• LabVIEW-programmable hardware

The thesis describing the power electronics can be found here [4].

(9)

2 Theory

2.1 Excitation systems

Excitation systems can be classified into three categories based on the power source for the excitation.

DC excitation systems

DC generators are used as power source and provide the current to the rotor through slip rings. They were used in early hydropower systems but got superseded by AC exciters in the mid 1960s.

AC excitation systems

A second generator is used as power source, generally the generator is on the same shaft as the main generator. The AC output is then rectified and fed into the rotor windings. The rectification can either be stationary with current being fed through slip rings to the rotor, or it can be rotating and no slip rings is needed.

Static excitation systems

Static excitation systems supply the excitation current through slip rings and take their power directly from the main generator [6].

2.2 Power electronics

Power electronics is defined as the application of solid-state electronics to the control and conversion of electric power. It is based primarily on switching power semiconductor devices to generate a desired voltage or current [7].

2.3 Bluetooth

Bluetooth is a wireless technology for exchanging data, invented by Ericsson in 1994. It uses the 2.4GHz ISM band with gaussian frequency shift keying (GFSK) as modulation scheme. In GFSK the frequency of the carrier is shifted to carry the modulation, a binary 1 is represented by a positive deviation in frequency while a binary 0 is represented by a negative frequency deviation.

Communication is based on a master-slave principle. One master device can control 7 slaves in a piconet, see Fig. 2.

(10)

Figure 2: Bluetooth master-slave architecture with one master and four slaves.

Bluetooth uses frequency hopping techniques to avoid interference. A transmission changes channel within the 2.4GHz ISM band 1600 times per second in a random pattern, this makes it more immune to interference. Version 2.1 + enhanced data rate (EDR) supports a bit rate of 3Mbps and utilizes phase shift keying (PSK) as well as GFSK as modulation schemes. Through the use of the protocol RFCOMM a wireless asynchronous serial port can be established between two devices, a useful feature for sending data between devices.

In Tab. 2 the different classes of bluetooth devices are listed. It is sorted based on the transmit power, class 1 is mainly for industrial applications while class 2 is the standard for mobile phones and similar items.

Table 2: Bluetooth classes and their corresponding transmit power and range.

Class Transmit power (dBm) Range (m)

1 20 100

2 4 10

3 0 0.1

To start a bluetooth communication between devices a procedure known as pairing must take place.

The process of pairing is as follows.

1. The devices look for other devices in range.

2. The user requests to pair with a specific device.

3. The device prompts for a passkey which is then compared with the other device.

4. If the keys are the same the connection is established.

This is only done once, afterwards the devices are paired until a user deletes the pair [8, 9].

2.4 Serial peripheral interface

Serial peripheral interface is a serial data transfer protocol developed by Motorola. It is used for communication between devices in full-duplex mode. Generally a bus has one master and an arbitrary amount of slaves connected to the same data lines. The following lines are available.

SCLK - Clock for the bus, controlled by the master.

(11)

MOSI - Master Out Slave In, output from master to slave.

MISO - Master In Slave Out, output from slave to master.

SS - Slave Select, used to select which peripheral that is allowed to use the bus.

In Fig. 3 a block diagram of a typical SPI connection with two slaves is shown.

SCLK MOSI MISO SS1 SS2 Master

Slave 1

Slave 2 SCLK

SCLK MOSI MISO SS1

MOSI MISO SS2

Figure 3: Block diagram of a SPI bus with one master and two slaves.

Since SPI is full duplex, data is exchanged simultaneously from and to the master. This is accom- plished by using a circular buffer consisting of one shift register in the master and one in the slave, see Fig. 4, data exchange is done by shifting the bits between these two registers [10].

A0 A1 A2 A3 A4 A5 A6 A7 B0 B1 B2 B3 B4 B5 B6 B7

MASTER SLAVE

MOSI

MISO

Figure 4: Circular buffer between master and slave.

In Fig. 5 the timing diagram for a transfer of one byte is shown. The communication starts by pulling SS low and the first bit is outputted from the master and the slave on MOSI and MISO. On the rising edge of the SCLK the bit on MISO is read by the master while the bit on MOSI is read by the slave. At the falling edge of SCLK a new bit is outputted on MISO and MOSI respectively and the process repeats until all bits are sent and SS is pulled high.

(12)

Figure 5: Timing diagram for one byte transfer.

2.5 Aliasing and anti-aliasing filters

In sampled measurement systems the aliasing effect has to be considered. The Nyquist sampling theorem says that if you have a signal that is band limited to a bandwidth of f_o then you can collect all information in that signal as long as your sample rate is higher than 2 ∗ fo, see Eq. (1).

f_o < f_sample

2 (1)

If you sample a signal that does not fulfill (1), the frequency content above f_sample

2 will be aliased back into the original signal. This leads to distortion of the sampled signal [11].

2.6 Successive approximation analog to digital converters

Succesive approximation register (SAR) ADCs implements a binary search algorithm to sample the signal. In Fig. 6 a functional block diagram of the architecture is available.

Track&Hold V_In

DAC SAR Data out

Comparator

Figure 6: Functional diagram of a SAR ADC.

When a sampling starts the analog input is held in a track/hold and the digital to analog converter (DAC) is set to, V_REF/2. A comparison between the analog input and the DAC output is done to determine if VIn is less, or greater than VDAC. The SAR-logic then saves the result in the first position of the register. The control logic then moves to the next bit and repeats the process all the way down to least significant bit (LSB), when finished it outputs the N-bit digital word in the output register [12]. In Fig. 7 the operation of a 4-bit SAR ADC can be seen.

(13)

Figure 7: Operation of a 4-bit SAR ADC.

2.7 The Hall effect

The hall effect can be used to get the magnitude of a magnetic field. The effect can be observed when an electric current is passed through a conductive material which is in a magnetic field with an orthogonal component to the current. A force is then exerted on the charged electrons according to Eq. (2)

F = q~~ v × ~B (2)

This leads to a charge build up on one side and charge depletion on the other. The voltage that arises is called the hall voltage. See Fig. 8 for a sketch.

Figure 8: Sketch that shows the principle of the hall effect [13].

(14)

2.8 Current measurement techniques

2.8.1 Resistive current sensing

A shunt resistor is inserted in series with the circuit in which the current shall be measured. A low resistance is important to ensure a negligible impact on the original function of the circuit. Eq. (3) is used to get the current.

I = U

R (3)

Since Rshunthas a low resistance, the voltage drop over it will be small so a good voltage measurement is needed to get good performance. See Fig. 9 for a circuit diagram.

i

R_shunt +

V −

V

Circuit to be measured

Figure 9: Current measurement of an arbitrary circuit with a shunt resistor.

2.8.2 Current transformer

Non intrusive measurement of alternating current can be done by means of a current transformer, it consists of a magnetic core with a hole through it. The wire in which the current of interest flows is passed through the hole making the primary side of the transformer one turn whilst the secondary winding has lots of turns, see Fig. 10.

Figure 10: Basic sketch of current transformer.

(15)

The output current of the secondary is measured. See Eq. (4) for the ratio between I_s and I_p. I_s= I_pNp

N_s (4)

Where N_p and N_s is the number of turns on the primary side of the transformer and N_s is the number of turns on the secondary side.

2.8.3 Hall effect based current measurement

Non intrusive measurement of both DC and AC can be done by utilizing the hall effect. It works in the same way as a current transformer except that the secondary winding is replaced by a small air gap in which a hall element is placed, see Fig. 11. The hall element will output a voltage proportional to the magnetic flux through it. This is directly proportional to the current passing through the core.

Figure 11: Basic sketch of a hall effect based current measurement.

2.9 Voltage measurement techniques

2.9.1 Resistive divider

For measuring high voltages a resistive divider can be used, Eq. (5).

Vm= R2

R1 + R2V_supply (5)

It provides an easy way to measure high voltages with a low voltage ADC. In Fig. 12, a basic circuit diagram of a voltage divider used for measuring an arbitrary circuit is presented.

(16)

+ − V_supply

R1

R2 V V_m

Figure 12: Voltage measurement of an arbitrary circuit with a resistive divider.

2.9.2 Hall effect based voltage measurement

Voltage can be measured with a hall effect based sensor. Rm is dimensioned to provide a specific current at different voltage levels. This current is passed through a hall element and an output voltage is generated at its output terminals, Fig. 13. The main benefit of this compared to a resistive divider is that the measurement system and the high power system is galvanically isolated.

+ −

V

im

R_m

Hall Vm

Figure 13: Voltage measurement of an arbitrary circuit with a hall element.

2.10 LabVIEW

LabVIEW is a graphical programming platform from National Instruments. It is most commonly used for control systems and data acquisition.

Programs are created by connecting functional blocks together and associating these to a front panel with controls. The programming language is by nature parallel since multiple loops can run simultaneously, this makes it powerful as a programming language for standard applications as well as programming of devices capable of running many threads in parallel, such as FPGAs [14].

(17)

3 Method

3.1 System overview

In Fig. 14, a schematic for the power electronics system is presented. Measurements of voltage and current are done in six locations.

The brushless exciter is represented by three voltage sources, the output current is measured and then fed to an active rectifier constructed with MOSFETs. A DC voltage is generated at VDC,1, the next stage is a buck converter which lowers the voltage and creates another DC rail at node VDC,2, both these voltages are measured. An H-bridge topology is used to control the current through the rotor, LRotor, for this a current measurement is needed as feedback to the current controller that generates switching patterns of the H-bridge. An extra current measurement is done at L_buck to analyse the inductor currents in the buck circuit.

V_DC,1

V_DC,2 C1

L_buck

C2

L_Rotor

Figure 14: Full high power system with ideal switch representation [4].

In Fig. 15 a simplified block diagram of the electronics is presented.

(18)

Figure 15: Block diagram of the control and measurement system.

3.2 Single Board RIO, sbRIO-9606

The processing unit used is sbRIO-9606 from National Instruments, see Fig. 16.

Figure 16: Single board RIO, sbRIO-9606.

(19)

It features a 400MHz real-time processor, a Xilinx Spartan-6 LX45 FPGA and I/O on a single printed circuit board. Access to 96 DIO-lines is available through a high-speed connector, this enables connection of custom made daughter boards. Integrated Ethernet, CAN, RS232 and USB is also available. The system is programmed from LabVIEW. For RS232 specifications see Tab. 3 [15].

Table 3: RS232 DTE Serial port specifications.

Baud rate support Arbitrary Maximum baud rate 230.4kbps Data bits 5, 6, 7, 8

Stop bits 1, 2

Parity Odd, Even, Mark, Space, None

Flow control RTS/CTS, XON/XOFF, DTR/DSR, None

3.3 General purpose inverter controller, NI 9683

The general purpose inverter controller is an off the shelf daughter board for sbRIO. It contains I/O for control and monitoring of power electronics. 14 channels in push-pull configuration are available for driving power electronics, the voltage level is set by providing a voltage at the V_ext-pin. In Tab. 4 specifications are presented [16].

Table 4: Half-bridge digital output.

Number of channels 14

Maximum continuous output current 10mA

Output impedance 100Ω

External power supply voltage range 5 - 30V

Minimum pulse width 500ns

Maximum switching frequency (50 pF) 500kHz

There are 32 DIO channels directly linked to the FPGA available. In Tab. 5 specifications are presented.

Table 5: LVTTL digital input/output.

Maximum tested current (per channel) 3mA Maximum total current (all lines) 96mA

TTL voltage level 3.3V

The GPIC also provides 16 pseudo-differential analog input channels. In Tab. 6 specifications are presented.

(20)

Table 6: Simultaneous analog input specifications.

ADC resolution 12 bits

Input range ±10V, ±5V

Sample rate (per channel) 100kS/s

Bandwidth 210kHz

Scanned analog channels are available for monitoring of slow processes, such as temperature. In Tab. 7 specifications are presented.

Table 7: Scanned analog input specifications.

Input range 0 - 5V

Sample rate (per channel) 1kS/s

Bandwidth 130kHz

3.4 Rotor main board

The main board connects to the GPIC through 2.54mm headers and provides bluetooth connectivity, 32 extra ADC channels, power supply and DSUB-cable connectivity.

3.4.1 Power supply

The power for the system is supplied from the high voltage DC rail of the active rectifier. A TDK-Lambda HWS100A-24/A is used to convert the high voltage DC to 24V [17]. This 24V rail is connected to the ”Power in” connector in Fig. 34.

The on board power supply accepts an external voltage between 18 - 36V and provides ±15V, 5V and 3.3V. The external voltage is passed to the sbRIO connector. The main DC/DC converter, Fig. 17, is a Traco Power TEN60-2423N, it is a switching regulator that can provide 60W. The efficiency is up to 92% and the output ripple is 125mV peak-to-peak maximum when measured with 20 MHz bandwidth [18].

Figure 17: The main DC/DC converter.

(21)

The 5V rail is supplied by a Texas Instruments LM340-N, which is a 5V linear regulator with 75µV specified output noise and the 3.3V rail is supplied by a ST LD1117, a low drop out regulator with 0.003% of Vout as output noise [19, 20].

3.4.2 RN41XV, bluetooth module

The bluetooth module RN41XV from Roving Networks is used, see Fig. 18. It supports bluetooth version 2.1 and is a Class 1 module. Communication is done through a serial interface via UART at a maximum rate of 240kbps. An external antenna can be connected, this is essential since the chip will be mounted inside and aluminium chassis [21].

Figure 18: RN41XV, bluetooth module.

3.4.3 ADC input signal conditioning

All analog signals in the system use current as the signal carrier to increase noise immunity. The current signal is converted to voltage by passing it through a resistor R_s close to the ADC. The value of Rs is determined by Eq. (6) and the maximum input voltage to the ADC.

U = RI (6)

The voltage signal is then buffered into an anti-aliasing filter implemented with a Sallen-Key topology. In Fig. 19 the input stage to the ADC can be seen.

(22)

−

+

R1 R2

C2 C1

VOut

−

+

R_s I_In i

Sallen-Key low pass filter

Figure 19: Input stage to ADC.

With R1, R2, C1 and C2 exchanged to impedances, see Fig. 20, the transfer function of the filter can be calculated.

−

+

Z1 Z2

Z4 Z3

v_out v_in

v₁

v+

v₋

Figure 20: Sallen-Key with impedances.

In this analysis all components are assumed to be ideal, this leads to Eq. (7).

v₊= v−= v_out (7)

Kirschoffs current law applied at v₁, Eq. (8).

vin− v₁

Z₁ = v1− v_out

Z₃ +v1− v_out

Z₂ (8)

Kirschoffs current law applied at v+, Eq. (9).

v1− v_out Z2

= vout

Z4

(9) From Eq. (9) the expression for v1 is obtained, Eq. (10).

v₁ = v_out Z₂ Z4

+ 1

(10)

(23)

The transfer function, Eq. (11) is found by combining Eq. (8) and (10).

v_out vin

= Z₃Z₄

Z1Z2+ Z3(Z1+ Z2) + Z3Z4

(11) The cut-off frequency for the anti-aliasing filter is set to 10.6kHz with Q = 0.5 this reduces it’s impact on the signal of interest while filtering out high frequency noise. The following values are used Z1 = Z2 = 15kΩ and Z3 = Z4 = 1

s1 × 10⁻⁹. Eq. (12) shows the transfer function of the filter.

H(s) = 44.44 × 10⁸

s²+ 13.33 × 10⁴s + 44.44 × 10⁸ (12)

3.4.4 AD7490, analog to digital converter

AD7490 is used as the ADC for the 32 additional analog input channels. It is a 16 channel, 12 bit converter that uses the succesive approximation for conversion. For key specifications see Tab. 8.

Table 8: Key specifications of AD7490.

Input range 0 - 5V

Sample rate (per channel) 1MS/s Signal-to-noise+distortion ratio 70.5dB

Serial communication with the integrated circuit (IC) is SPI compatible. The conversion clock and SPI clock is shared and comes from the SPI master, therefore conversion speeds is fully controllable from the software in the master [22].

3.4.5 Relay control

Four relays can be controlled from the main board. The output can handle 500mA continuous current and has integrated flyback diodes. These can be used to connect and disconnect different parts of the circuit when necessary.

3.5 Current measurement

For current measurement LEM LA55-P is used. It utilizes the hall effect to measure the current in a cable passing through the sensor, see Fig. 21.

(24)

Figure 21: Current measurement board.

The sensor output is a current with conversion ratio 1:1000. A current of 50A in the main cable generates a 50mA output from the sensor, this current is then passed through a resistor close to the measurement ADC and then sampled [23].

100nF

−15V 22µF

+15V

iout

I_Out LEM LA55-p

Figure 22: Circuit diagram of a current sensor board.

Table 9: Specifications for LA55-P.

Overall accuracy ±0.65%

Linearity error < 0.15%

Response time < 1µs Bandwidth (-1 dB) 200kHz

3.6 Voltage measurement

For voltage measurement LEM LV25-P is used, see Fig. 23. The measurement is based on the hall effect thus it provides galvanic isolation between the high voltage side and the measurement side.

(25)

Figure 23: Voltage measurement board.

It is connected in parallel with the load and the desired measurement range is set by choosing resistor values on its input. The resistor should be set so that 10mA is passed into the device at the nominal voltage. The current is then converted with a conversion ratio of 2500:1000, a 10mA current into the terminal corresponds to a 25mA current on the output terminal [24].

i_m 10kΩ 10kΩ

10kΩ 10kΩ

V_m

100nF

−15V 22µF

+15V

iout

I_Out LEM LV25-p

Figure 24: Circuit diagram of a voltage sensor board..

Table 10: Specifications for LV25-P.

Overall accuracy ±0.9%

Linearity error < 0.2%

Response time 40µs

3.7 Magnetic field measurement

The magnetic field measurement need to be able to measure magnetic fields up to 700mT, for this purpose the hall element, ChenYang CYSJ166A was used. Its main features is 0-3T measurement range, good linearity and good temperature stability [25].

The output of the hall element is buffered then fed in to a XTR117 from Texas Instruments. It is a 4 - 20mA current loop transmitter that enables a voltage signal to be converted to current and fed through the same wire as the power supply [26]. In this way only two wires for each sensor board is required and the signal’s noise immunity is increased. See Fig. 25 for a complete circuit diagram

(26)

Hall-

element ₋

+

10nF

i_out 100nF

VIn

N C I_In I_RET I_O

VREG

V + B E XTR117 MCP6021

15kΩ +V_H

−VH

+V

−V

+4.7 V

npn

Figure 25: Circuit diagram for the magnetic field measurement board.

3.8 Serial communication in LabVIEW

LabVIEW’s integrated libraries for serial communication are used to communicate with the bluetooth modules. A state machine is implemented that continuously checks the RX port for data and then either reads the incoming data or writes measurement data back to the other device, see Fig. 26.

Figure 26: State diagram for the serial communication loop.

This state machine is used on the rotating side of the system. It should send measurement data all the time until a new set point value for the rotor current is sent from the stationary system, when this becomes available it should be handled right away and then go back to sending measurement data again. For the complete program see Appendix A.

3.9 SPI in LabVIEW

The SPI communication is implemented as a state machine, see Fig. 27. It initializes the transfer by pulling SS low. Then it writes MOSI and reads MISO in 16 clock cycles before returning to the initial state. In this way the clock frequency is configurable in LabVIEW [27]. For the complete program see Appendix A.

(27)

Figure 27: State diagram for the SPI communication loop.

(28)

4 Results

For detailed schematics and layouts for the presented circuit boards, see Appendix B.

4.1 Magnetic field sensor

The hall element and the magnetic field sensor was characterized by applying a magnetic field from -700mT to 700mT. The magnetic field was generated by a C-core, see Fig. 28.

Figure 28: The magnetic C-core used for characterization of the magnetic field sensor [28].

The sensor was placed in the air gap of the C-core along with the measurement probe from a LakeShore 410 Gaussmeter. The output voltage from the sensor was measured with a Fluke 175 multimeter, the accuracy of the two instruments are as follows.

• Gaussmeter, 2% of reading 0.1% of full scale at 25^◦C.

• Multimeter, 0.15% + 2 counts.

Two hall elements were characterized, both elements was measured from -700mT to 700mT three times, the result can be seen in Fig. 29. The output voltage as a function of magnetic field is presented.

(29)

Figure 29: Linear curve fit of output voltage at different magnetic fields with residual plot for the hall element.

A picture of a fully populated magnetic field sensor board can be seen in Fig. 30. It uses only two wires for signal transmission and for powering the circuit. The dimensions of the board is 10x18mm and the maximum height is 2mm.

Figure 30: Fully populated hall sensor board.

The magnetic field sensor board was characterized with the same amount of measurement points as the hall element. The current signal was converted to voltage by passing its output current through a resistor with R = 200.2Ω. In Fig. 31 the output voltage as a function of magnetic field can be seen.

(30)

Figure 31: Linear curve fit of output voltage at different magnetic fields with residual plot for the complete hall measurement system.

Mounting positions for the magnetic field sensors can be seen in Fig. 32.

Figure 32: Mounting of the magnetic field sensors on the rotor.

(31)

4.2 Rotor main board

The main board has connectors to fit the GPIC board from National instruments. It extends its functionality by adding power supply for the complete system, 32 extra ADC channels, bluetooth connectivity, and distribution of signals to the rest of the system. Pictures of the fully populated main board with the function of each section marked can be seen in Fig. 33.

Figure 33: Top side of main board with sections marked.

The bottom of the main board is populated with all the connectors necessary to distribute the signals in the system. The main cable standard for signals is DSUB. See Fig. 34 for an overview of the bottom side of the board with connectors and their function marked.

(32)

Figure 34: Bottom side of main board with individual connector’s function marked.

4.3 The finished main unit

In Fig. 35 the fully assembled main system is shown.

(33)

Figure 35: Fully assembled system.

The system in Fig. 35 was put in an aluminium box with holes milled for the connectors, see Fig. 36.

Figure 36: Aluminium box with the circuit boards mounted.

(34)

4.4 Rotor distribution boards

To distribute the signals from the main board two circuit boards were created to fit around the shaft close to the rotor. The boards are mounted on a bakelite piece used for connecting the rotor windings, see Fig. 37.

Figure 37: Distribution boards mounted on a bakelite piece meant for connection of rotor windings.

4.4.1 Distribution board for rotor power electronics and SPI

A fully populated distribution board for rotor power electronics and SPI can be seen in Fig. 38.

(35)

Figure 38: Distribution board for power electronics.

4.4.2 Distribution board for rotor sensors

A fully populated distribution board for sensor connections can be seen in Fig. 39.

Figure 39: Distribution board for sensors.

(36)

5 Conclusions

The control and measurement system works well and has been tested. Voltage rails are within specifications, all ADC channels are fully functional and the half bridge output drivers works satisfactory. The system’s capability for driving power electronics was tested by mounting everything in the box and drive a buck converter and a three phase inverter with SVM control scheme. This was successful and current and voltage measurements were working well.

Figure 40: Box for control, measurement and power electronics (control board not mounted) [4].

The distribution boards fits nicely and no problems were discovered when analysing and testing these boards.

The software for the system has basic functionality, it can sample the ADC channels via SPI and it can communicate wirelessly between two bluetooth modules. No elaborate programming to automatize the bluetooth connection setup or the sending of data between the two units has been done.

Magnetic field sensors with high linearity and a large measurement range has been constructed.

The sensors have low noise floor, 24mVp−p at 0T and increased noise immunity since current is used as signal carrier. A first degree polynomial fit, y = 9.741 × 10⁻⁴x + 2.850, was done and shows good consistency and linearity with R² = 0.995. For further improvement the vias for mounting the cables, see Fig. 30, should be changed to pads so that the whole bottom side is free from exposed conducting surfaces. A list of minor errors in the layout of the board were discovered, see list below.

• Hole size for 10 pin connector too small.

• Protection diodes for TTL pins were incorrectly positioned.

• Pad size for flyback diodes in relay channels too small.

(37)

References

[1] El˚aret, 2013. Svensk energi. http://www.svenskenergi.se/Global/Statistik/El%C3%

A5ret/Sv%20Energi_el%C3%A5ret2013_versJUNI2014.pdf. [Accessed 27 May 15].

[2] M. Wallin. Measurement and modelling of unbalanced magnetic pull in hydropower generators.

2013. ISSN 1651-6214; 1029.

[3] J. Jos´e P´erez-Loya (personal communication, 2015)

[4] T. Johansson. Active rectification and control of magnetization currents in synchronous generators with rotating exciters. 2015.

[5] P. Schavemaker and L. van der Sluis. Electrical power system essentials. Wiley, 2008.

[6] P. Kundur. Power Systems Stability and Control. McGraw-Hill, 1994.

[7] Rashid, M.H. (2004). Power electronics : Circuits, devices and applications, 3rd edition. Upper Saddle River, N.J. ; London: Pearson Prentice Hall. 2004.

[8] H. Labiod, H. Afific, C. de Santis. Wi-Fi, Bluetooth, ZigBee AND WiMax. Springer, 2007.

[9] Ian Poole. Radio Electronics. Available at: http://www.radio-electronics.com/info/

wireless/bluetooth/bluetooth_overview.php. [Accessed 01 June 15].

[10] Mayank Prasad. Available at: http://maxembedded.com/2013/11/

serial-peripheral-interface-spi-basics/. [Accessed 01 June 15].

[11] T. Wescott. Sampling: What Nyquist Didnt Say, and What to Do About It Available at:

http://www.wescottdesign.com/articles/Sampling/sampling.pdf. [Accessed 01 June 15].

[12] Maxim Integrated. Understanding SAR ADCs: Their Architecture and Comparison with Other ADCs. Available at: http://pdfserv.maximintegrated.com/en/an/AN1080.pdf. [Accessed 05 June 15].

[13] Picture from wikipedia. Available at: http://en.wikipedia.org/wiki/Hall_effect#

/media/File:Hall_Effect_Measurement_Setup_for_Electrons.png. [Accessed 05 June 15].

[14] LabVIEW homepage. Available at: http://www.ni.com/labview/. [Accessed 08 June 15].

[15] National Instruments. OEM operating instructions and specifications, NI sbRIO-9605/9606 and NI sbRIO-9623/9626/9633/9636. 2012.

[16] National Instruments. User guide and specifications, NI 9683. 2013.

[17] TDK-Lambda. TDK-Lambda, HWS100A-24/A dataheet. 2015. Available at: http://www.

mouser.com/ds/2/400/hws-a-525007.pdf. [Accessed 23 June 15].

[18] Traco Power. Traco Power, TEN 60N Series dataheet. 2013. Available at: http://assets.

tracopower.com/TEN60N/documents/ten60n-datasheet.pdf. [Accessed 04 June 15].

[19] Texas Instruments. LM340-N, datasheet. 2013. Available at: http://www.ti.com/lit/ds/

symlink/lm340-n.pdf. [Accessed 04 June 15].

(38)

[20] ST Microelectronics. LD1117, datasheet. 2013. Available at: http://www.st.com/st-web-ui/

static/active/en/resource/technical/document/datasheet/CD00000544.pdf. [Accessed 04 June 15].

[21] Microchip. RN41XV, datasheet. 2012. Available at: http://ww1.microchip.com/downloads/

en/DeviceDoc/RN41XV-RN42XV-ds-v1.0r.pdf. [Accessed 09 June 15].

[22] Analog Devices. AD7490 datasheet. 2012. Available at: http://www.analog.com/media/en/

technical-documentation/data-sheets/AD7490.pdf. [Accessed 08 June 15].

[23] LEM. LA55-P, datasheet. 2009. Available at: http://www.lem.com/docs/products/la%

2055-p%20e.pdf. [Accessed 04 June 15].

[24] LEM. LV25-P, datasheet. 2012. Available at: http://www.lem.com/docs/products/lv%

2025-p.pdf. [Accessed 04 June 15].

[25] ChenYang. CYSJ166A, datasheet. Available at: http://www.hallsensors.de/CYSJ166A.

pdf. [Accessed 04 June 15].

[26] Texas Instruments. XTR117, datasheet. 2012. Available at: http://www.ti.com/lit/ds/

symlink/xtr117.pdf. [Accessed 04 June 15].

[27] Implementing SPI Communication Protocol in LabVIEW FPGA. Available at: http://www.

ni.com/example/9117/en/. [Accessed 08 June 15].

[28] S. Sj¨okvist and S. Eriksson. Experimental Verification of a Simulation Model for Partial De- magnetization of Permanent Magnets. IEEE Transactions on Magnetics, vol.50, no.12, pp.1,5, Dec. 2014.

(39)

Appendix A LabVIEW code

(40)

(41)

(42)

(43)

(44)

1 2 3 4

A

B

C

D

E

A

B

C

D

E

Title: Rotating RIO ADD-ON board

File: Add_on_Board_NI9683.sch Sheet: /

Uppsala University Fredrik Evestedt Bluetooth

Bluetooth.sch PSU

PSU.sch

Half_Bridge

Half_Bridge.sch Simultaneous_AI

Simultaneous_AI.sch

ADC1_2

ADC1_2.sch

Relay_CTRL

Relay_CTRL.sch Scanned_AI_AO

Scanned_AI_AO.sch

Sourcing_DI

Sourcing_DI.sch Global NETS +2.5V_REF +3.3V +3.3V_MEZZ +5V +15V +24V -15V COM

Ethernet

Ethernet.sch

H2 Hole

H3 Hole

H4 Hole

H5 Hole

Appendix B Schematics and Layouts

41

(45)

1 2 3 4

A

B

C

D

E

A

B

C

D

E

Uppsala University Fredrik Evestedt READY

1 2 C1+

3 V+

4 C1- 5 C2+

6 C2- 7 V-

8 RIN FORCEOFFFORCEONINVALIDROUTDOUTGNDVCCDIN101112131415169 U1 MAX3227E

C3 100n

C4 100n 1 1

2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 1010 X1 CON-10P

VDD_3V3 1 2 TXD 3 RXD

GPIO10 4

RESET_N 5

GPIO6 6

GPIO9 7

GPIO4 8

GPIO11 9 10GND

AIO1 20

GPIO8RTS1112 GPIO2NC1314 GPIO5CTS1516 GPIO317 GPIO7AIO0 1819

U2 RN41XV

R1100k R2100k R3100k R5100k R6100kR4100k12 D1LED C1

100n

C2 100n

C5 100n

C6 100n

C7 100n

C8 100n +3.3V

COM

+3.3V

H1 Hole

COM

Decoupling

(46)

1 2 3 4

A

B

C

D

E

A

B

C

D

E

Title: Rotating RIO PSU

File: PSU.sch Sheet: /PSU/

Uppsala University Fredrik Evestedt 1 +VIN

2 -VIN 3 RMT

+15V 4 COM 5 -15V 6 U3 TEN_30-2423WIN

GND1

VO 2

3 VI U4 LD1117S33TR C9

4.7u

C121n C10

1n C13

220nF

C15 100n

C16 10u 1

2

24V_INP1

1 F12

FUSE

1 2

BATT_INP2

1 F22

FUSE

D2

DIODE

D3

DIODE

+5V

+15V COM -15V

+3.3V +24V 1 2

P3 24V_RIO

24V_IN

BATT_IN

C14 100n 1 Vin

GND2

Vout 3

GND4 U5 LM7805_sot223

1 2

D16

R1191.5k LED -15V COM

V1 X9 +15V

V1X10 +5V

V1X11 +3.3V

V1X12 -15V

V1X13 COM V1X8

+24V

+24V +5V

+15V +3.3V -15V COM

Test points Power LED

(47)

1 2 3 4

A

B

C

D

E

A

B

C

D

E

Uppsala University Fredrik Evestedt

1 2 3 4 5 6 7 8 9 10

20 11

21 12

22 13

23 14

24 15

25 16

26 17 18 19

X2 HALF_BRIDGE_DO

HB_DO13 HB_Vext HB_DO12 GND HB_DO11 GND HB_DO10 GND HB_DO9 GND HB_D08 GND HB_DO7 HB_DO6 GND GND GND GND GND HB_DO5 HB_DO4 HB_DO3 HB_D02 HB_DO1 HB_DO0

1 2 3 4 5

6 7 8 9

J6

DB9

1 2 3 4 5

6 7 8 9

J7

DB9 1 2 3 4 5

6 7 8 9

J4

DB9

1 2 3 4 5

6 7 8 9

J2

DB9 1 2 3 4 5 6 7 8 9 10

20 11

21 12

22 13

23

14 24

15 25

16 17 18 19

J1

DB25

GND

+15V COM

+15V

COM C120

10uF C121

10uF C122

10uF C123

10uF C124

10uF C125

10uF

C127

10uF C128

10uF C129

10uF

Fault_ROTOR_HB

Fault_Buck 1 VCC

2 AOUT 3 AIN 4 BOUT 5 BIN 6 COUT 7 CIN

8 VSS DOUTEOUTFOUTVDDSELDINEINFIN 101112131415169 U25 CD4504

+15V COM +5V

Fault_Rectifier 1 A

2 B

3 D 4 E

5 F D+E+F 6

VSS 7 8 C A+B+C 9

G+H+I10 11IH 12G 13

14VDD

U26 CD4075

COM+15V

C130

100nF C131

100nF COM

+15V

C132

100nF +5V

COM

C126

10uF Power supply decoupling

COM

Wireless control and measurement system for a hydropower generator with brushless exciter

Examensarbete 30 hp Juni 2015

Wireless control and measurement system for a hydropower generator with brushless exciter

Fredrik Evestedt

Abstract

Wireless control and measurement system for a hydropower generator with brushless exciter

Contents

1 Introduction

2 Theory

3 Method

4 Results

5 Conclusions

References

Appendix A LabVIEW code

Appendix B Schematics and Layouts