Automation of Optical Test Equipment using LabVIEW
Contents
1. 44 PROGRAMMING Vertical offset Min Max Waveform a rss vscaleGraph vi Block Diagram rss vscaleGraph vi Vertical position Min Max Waveform Vertical offset Vertical range b rss vscaleGraph vi icon Figure 32 rss vscaleGraph vi Block Diagram generation of a waveform with maximum and minimum values of the display of the Oscilloscope Since the waveform is continuing changing during the configura tion and also in the measurements the autoscale propriety of the graph in certain situations causes a modification of the axis when the program is running This lead to movements up and down of the curve rss vscaleGraph oi was made to solve this problem Figure 32 It calculates the maximum and the minimum points that can be displayed from the Oscilloscope using the parameters for the vertical and horizontal scale These points are then inserted into an array of two elements The array containing the data and the one with the maximum and minimum points are concatenated together with the function Build Array The resulting array is the input of the wave form graph which displays the data and the two additional points However since they are only two these points do not compromise the correct visualisation of the data Build Array allows to plot more curves in the same graph useful e g to make figures that compares different results as the ones presented in Chapter 2 4 44 MEASUREMENT2. IB Zu en 0 5 Un T S uLED 28 8 uLED 579 1 a 1 5 da gt x S25 pae amt NT 2 35 AL E Lir E 4 5 ity 5 5 6 6 54 1 10 100 1000 Envelope dB Frequency MHz a Comparison of the uLEDs 5 V 10 100 1000 Frequency MHz b Comparison of the uLEDs 6 V Figure 9 Comparison Voltages and uLEDs 2 1 LED MEASURE OF THE ENVELOPE USING THE PULSE 19 waveform of the 8th uLED stays under the one of the 5th uLED until 80 MHz where it exceeds the other curve and stays over it Z uLED 5 uLED 8 uLED NU ss d M rd aero RD n 7 Chad u m at PELI Voltage V Figure 10 Bandwidth function of the voltage for all the uLEDs The hypotheses about the bandwidth made looking at the mea sures are confirmed by Figure 10 The bandwidth increases with the voltage With the same voltage fixed the bandwidth increases reduc ing the size of the pixels RR 1 E aS E a 4 9 ar A 1 10 100 1000 Frequency MHz Figure 11 Envelope vs frequency for different Duty Cycles 5 V 3rd 50um first row Finally an analysis of different duty cycles at 5 for the 3rd uLED
3. 2 2 5 lt nu voc Rm es mu B 35 ENE o n II sn E ow ya LA 5 IM 5 4 5 t 5 B 5 5 6 6 5 74 i i 1 10 100 1000 Frequency MHz a Calibrated and not calibrated results DF 2xf until f 500 MHz then DF 1 GHz 5th uLED 40um second row 40 um LED 7 2 Envelope dB 5 5 7 1 10 Frequency MHz 100 b Average of 5 measurements with calibrated data DF 2xf until fi 500 MHz then DF OFF 5th uLED 40um second row Figure 6 Calibration 5 V and 1 Duty Cycle 14 EXPERIMENTS not suspected an error using the digital LP filter Further analyses were done as a wider range of frequencies up to 1500 MHz The calibrated waveform stays steady even after 1000 MHz Figure 5a An analysis was then done without a digital LP filter in the oscillo scope Figure 5b The shape is more realistic decreasing with the frequencies and reaching the 3 dB point It is important to point out that these measures are normalised in respect of the largest envelope that is not the value at the lowest frequency However this analysis is only approximative and relates only the general shape of the data ob tained Furthermore the measurements were repeated with a digital LP filter equal to the frequency analysed The results show envelope measures independent from the frequencies with flat curves in th
4. and y is given from the data of the curve 2 1 2 Different configurations the insertion of attenuators As Figure 3 shows the first measure does not show any relationship between the frequency selected and the envelope For this reason the use of attenuators is necessary Different configurations were tried the best solution is the one shown in Figure 3c with the attenuator at the output of the NA Figures 3a and 3b may be explained by re flections that from the splitter where PG NA and LED are connected together go into the NA This kind of reflections are avoided with the attenuator between the splitter and the cable to the NA The presence of jumps especially for Figure 3a and 3b is due to the small number of frequencies analysed and to the sensitivity of the uLED Improve 10 EXPERIMENTS 40um pLED 0 0 54 14 1 54 24 5 3 2 3 5 5 2 4 4 5 5 5 5 6 6 5 1 10 100 1000 Frequency MHz a No attenuators 40um pLED 0 T 0 5 1 1 5 2 El 25 5 3 z t m O _35 ay 3 UNE e xps GFA 2 4 ore 5 a E SEEKS 4 5 u 2S i 78 i 54 WES rn 5 5 ri i 2 a 64 Y 6 54 a 10 100 1000 Frequency MHz b Attenuator at the input of the Oscilloscope 40um pLED 7 0 0
5. Configuration Pulse Generator Oscilloscope Analyzer Figure 20 Run vi front panel Once logged in it is possible to do the measurements Figure 20 shows the principal interface of the RUN part At the left of the window there is a tab control that allows to choose different group of controls and indicators The first tab includes all indicators associated with the configuration file loaded The list of objects divided by four fields Group name Display Filter Type and Time indicates some characteristics of the configuration chosen Username ConfigSpec and DefaultSpec are three indicators that show respectively the user name the path of the configuration file loaded and the path of the default configuration file The tabs for the PG and the Oscilloscope contain indicators that show the parameters chosen with the configuration file The only con trols are the VISA address the ID query and the Reset If active i e boolean true these two last controls perform the identification of the instrument and the reset The tab of the NA Figure 21 contains the list of frequencies that have to be analysed and an indicator which during the measure ments shows the actual frequency selected Before any measurement the VISA addresses for the instruments have to be set from the drop down list shown as control for each instrument They are usually GPIB 14 for the PG TCPIP 192 168 5 120 INST for t
6. cial LED supports higher voltages than the but it is limited by its internal proprieties as the capacitance and thus its bandwidth is much smaller compared to the same one of the uLED Furthermore for the uLED the bandwidth is bigger for smaller pixels with a sig nificant increase even with two consecutive pixels which differs only of 5 or 10 um in diameter The results obtained in pulsed conditions with a 100 duty cycle are consistent with the ones under continuous current even if the setup for the experiment is different and the first measurements are influenced by the noise at high frequencies 5 2 PROGRAMMING The programs work for different experiments using the Pulse Gener ator the Network Analyzer and the Oscilloscope They allow to mea sure and to save the envelope at different frequencies in spreadsheet files such as xls files with the frequencies in the first column and the envelope measures in the second column These kind of files can be used to plot the data in Envelope Frequency graphs which can be saved into different imaging formats The calibration of the instru ments is done simultaneously with the measures with the insertion of appropriate calibration factors inside the program More measurements for the same list of frequencies can be taken consecutively and automatically Each measurement is saved in a dif ferent file The user can make new configuration files with the con trols of the Pulse Gener
7. different from the one of the programs presented in 1 the other ideas and methods of programming are totally original 37 38 PROGRAMMING 4 2 CONFIGURATION 4 2 1 Pulse Generator Duty Cycle 5 X Voltage Amplitude V Delay Units Fixed Parameter 0166 Delay Value s Period s prn VISA Refnum out 63114 63114 AGA AG AGs error in no error I a error out Polarity 016 a PG vi Block Diagram PG vi Voltage Amplitude V Delay Units Polarity VISA Refnum in Delay Value s Enable Output Pulse Width s Duty Cycle 5 error in no error Period s Pulse Mode Fixed Parameter VISA Refnum out error out b PG vi icon Figure 26 Configuration for the Pulse Generator The configuration of the Pulse Generator uses simply the drivers for the instruments Figure 26 shows the connections between them This program is called PG vi the controls are the same shown in the front panel 4 2 2 Oscilloscope This is the main part of the configuration since the oscilloscope has to be configured to receive the data of the experiment to visualise them and to take some measures Figure 27 illustrates a sequence of the following instructions 4 2 CONFIGURATION 1 Fix the decimation mode to be one of the following options Sample Peak Detect High Resolution and RMS 2 Fix the method to build the waveform from more sequential ac quisitions Envelope Aver
8. of the first row 50 is presented This analysis is difficult due to strange shape that the curve assumes for duty cycles between 10 and 100 For this reason only the analysis for 1 5 and 100 DC is given in Figure 11 It is important to highlight that the results 20 EXPERIMENTS shown refer to fewer frequencies in comparison to the other results of this section The experiment was firstly done with a different list of frequencies and with a wrong calibration so the old results has been adapted with the list of the last calibration 5 V 1 DC Thus in this case the frequencies are 1 10 50 100 150 and 200 1000 sampled every 100 MHz Another issue is that the results are calibrated using the calibration for 1 DC and not for 5 and 100 DC This may lead to slightly different results However since the part analysed referred always to the top of the pulse in the same 0 8 section for 5 the width of the pulse is 5 us for 100 all the 100 us the instruments have a similar influence on the system in the three cases Therefore the error is considered negligible for the purpose of this section The bandwidths for each duty cycle are e 1 224 MHz e 5 188 MHz e 100 143 MHz Despite the few samples and the calibration the first bandwidth is consistent with the other results since the 3rd uLED is smaller than the 2nd uLED BW 120 MHz at 5 and is bigger than the 5th uLED BW 235 MHz at 5V Furthermo
9. arrays containing the measurements of the envelope in two distinct x s files one for the raw values and one for the calibrated results Figure 37 The name of the file is automati cally made and contains the state of the PG output the voltage the duty cycle the date in the form day month and the time in the form hours minutes seconds It accepts these inputs as controls The real process of saving the data is done by the subVI SaveSS vi Figure 38 It accepts as input an array and the path indicating the position for the new files The array is converted into a spreadsheet string with an automatic formatting and then saved into a file using the function of LabVIEW Write to text file The extension x s in the name of the file guarantees that it can be opened from Microsoft Excel Slightly different is the code for saving the graphs As Figure 39 shows the name of the file is done in the same way for the data store A propriety node is created from the waveform graph of LabVIEW and it is connected to the path built for the image file The proper ties selected are Get Image Image Depth BG Color and Image Data This last one is connected to the function Write JPEG File Also other formats as bmp and png can be used 4 6 DATA ELABORATION Figure 40 shows the block diagram for the data elaboration In se quence a normalisation a cubic spline fit and an interpolation are applied on the data The normalisation is done in the
10. current conditions 2 2 MEASURE OF THE ENVELOPE WITH DC CURRENT 25 2 S uLED 8 uLED 2 4LED DC 5 uLED DC 8 uLED DC Voltage V Figure 17 Comparison of the bandwidths for pulsed 1 duty cycle and direct current conditions REMOTE CONTROL This chapter concerns the concept of the Remote Control which in volves both the physical connections instruments laptop and the in terface used by the operator to control the instruments The section Front Panel presents the windows of LabVIEW that appear to the user when the program is running Further information about Lab VIEW are given in Appendix A 3 1 CONNECTIONS Two types of connection are used the General Purpose Interface Bus GPIB and the Local Area Network LAN A GPIB cable connects the NA and the PG through their GPIB port My laptop controls the two instruments together using a Controller GPIB per Hi Speed USB which allows the connection between the GPIB port of the PG and the USB port of the laptop An ethernet cable is used to connect the laptop to the oscilloscope LabVIEW recognises and fixes automatically the address for the instruments connected via A different situation arises for the LAN connection that has to be established manually using the TCP IP protocol as explained in 2 pp 450 452 3 2 FRONT PANEL Once LabVIEW detects correc
11. is much smaller com pared to the ones of the uLEDs This is the result of the limitation given by electrical proprieties of the Commercial LED 2 2 ULED MEASURE OF THE ENVELOPE WITH DC CURRENT frequency LED Digital Multimeter DC current supply Electrical 5 5 signal Network analyser Figure 15 Connections for the measure of the envelope with DC current The experiment under pulse conditions was repeated for the same pixels of the uLED under continuous current Figure 15 shows the new setup and connections used The DC current supply generates a continuous current It is connected to a Digital Multimeter that in 2 2 MEASURE OF THE ENVELOPE WITH DC CURRENT creases the precision of the current The calibration is done directly from the Network Analyzer using an apposite function After the de tector the data goes directly into the Network Analyzer where are saved The voltages analysed are now 4 5 6 V for the 2nd and the 5th uLEDs 4 5 V for the 8th uLED due to the limit of voltage that the uLEDs can carry with these new conditions The frequencies anal ysed are 0 1000 MHz with samples every 5 MHz The data saving and the elaboration is done using an apposite program in Matlab from MathWorks capable to read the file of the Network Analyzer Since this program was provided already made for other experiments it is not presented in this report The main important aim of this ex periment is to compare th
12. of the frequency at different voltages for pulsed 1 duty cycle and direct current conditions 24 Comparison of the bandwidths for pulsed 1 duty cycle and direct current conditions 25 MainProgram vi front panel 28 LogIn vi front panel 28 front panel 29 Run vi front panel tab NA 30 Results vi front panel 31 Configuration vi front panel 32 Configuration vi front panel tab with the con trols for the PG 33 Configuration vi front panel tab with the con trols for the oscilloscope 34 Configuration for the Pulse Generator 38 Configuration for the Oscilloscope 39 xiv List of Figures Figure 28 Figure 29 Figure 30 Figure 31 Figure 32 Figure 33 Figure 34 Figure 35 Figure 36 Figure 37 Figure 38 Figure 39 Figure 40 Figure 41 Figure 42 Figure 43 Figure 44 Configuration for the acquisition of the oscil loscope Configuration for the horizontal axis in the dis play of the oscilloscope 41 Configuration for the Network Analyzer 42 Part of the block diagram of Configuration vi controls and real time illustration of the con figuration used 43 rss vscaleGraph vi Block Diagram generation of a waveform with maximum and minimum val ues of the display of the Oscilloscope 44 Measure of the envelope and of the minimum voltage 45 Low function envelope in dB and data calibra tion 45 Calculation of the calibration factors 46 Part of the code of Run vi 47 Envelope saving 49 Sav
13. scale offset the trigger level and the digital filter change case by case These param eters are be selected to look at the top of the pulse since this ex periment focus on the envelope when the LED is ON An important advice is to fill the display of the oscilloscope with the curve in order to have the maximum resolution of the ADC as explain in 2 p 15 21 LED MEASURE OF THE ENVELOPE USING THE PULSE Run Single is the function used to acquire the data 100 acquisi tions are done for each single run This number was chosen to have a good quality of the data acquired in a small time Once the visualisation is correct automatic measurements are per formed to find the envelope The Peak to Peak function which calcu lates the difference between the maximum and the minimum values displayed is used In addition the Low function is used to find the minimum value displayed When the top of the pulse is not flat the y Low Envelope measure PtP Ay Figure 2 Solution used in case of no flat waves displayed by the oscilloscope Peak to Peak function does not measure the right envelope for this reason it is necessary to use the minimum value An example is Fig ure 2 which shows a possible situation displayed by the oscilloscope To find the correct measure of the envelope the calculation at the bot tom of the figure is made The blu quadrilateral represents the curve displayed PtP means Peak to Peak
14. this new set from the initial re sults frequency envelope given as input Finally the bandwidth is calculated using the interpolated data A loop processes all the envelope measures and stops when the enve lope is less than 3 dB with respect to the first value The frequency associated to the last envelope processed is the bandwidth searched 51 PROGRAMMING 52 Mg pue uone odaoqur 41 ejep Jo au 07 0035109233 341 20 5312416344 135 J3PIM E Jo uonejaua ydes AX pur aneg 4 7 STATE MACHINE CONCEPT AND MAINPROGRAM VI 4 7 STATE MACHINE CONCEPT AND MAINPROGRAM VI Start of the program Stop of the program Figure 41 State machine for the main program Figure 42 Block diagram of the main program The final part of this chapter gives an idea of how the main pro gram works using the state machine concept Figure 41 When the program starts it goes directly into the Log In After this step the program goes into an idle state It stays in this state until a button is pressed Every button except for Quit is associated with a VI that is opened just after the button is pressed The Front Panel of the pro gram is superimposed on th
15. 002 URL http cp literature agilent com litweb pdf 08753 90475 pdf Jeftrey Travis and Jim Kring LabVIEW for everyone graphical pro gramming made easy and fun London Prentice Hall 2007
16. 54 1 org 1 5 2 2 54 1 em 4 5 Envelope dB un 1 5 5 6 5 74 7 5 8 i 10 100 Frequency MHz c Attenuator at the output of the NA Figure 3 Envelope in dB versus Frequency in MHz pixel 40um 5 V 21 LED MEASURE OF THE ENVELOPE USING THE PULSE 11 ments of these measurements such as calibration of the instruments and average of more measures are presented in the next sections Finally different cables between the instruments do not cause sig nificant differences in the measurements 2 1 3 Calibration u Em j Pulse generator frequency Oscilloscope Network analyser Nie B LA Laptop Figure 4 Connections for Calibration The last step before starting any measurement on the LEDs is the calibration of the instruments For this reason I connected the three instruments together as in Figure 4 In this case the envelope is expected to be the same for all the frequencies no LEDs involved so calibrations factors are calculated in order to reach this condition 2 1 3 1 Calibration factors and Normalisation This paragraph describes the calculation made for the calibration and normalisation of the results y is the envelope measure in V at fre quency fi From Ymax max y1 where N is equal to the number of frequencies analysed the calibra tion fac
17. AUTOMATION OF OPTICAL TEST EQUIPMENT USING LABVIEW MATTEO RAMPADO Y 4 VERITAS Aprile 2014 Matteo Rampado Automation of Optical Test Equipment using LabVIEW Corso di Laurea Magistrale in Ingegneria dell Automazione Diparti mento di Ingegneria dell Informazione Padova O Aprile 2014 UNIVERSIT DEGLI STUDI PADOVA DIPARTIMENTO DI INGEGNERIA DELL INFORMAZIONE UNIVERSITY OF GLASGOW Conso DI LAUREA MAGISTRALE INGEGNERIA DELL AUTOMAZIONE TESI DI LAUREA AUTOMATION OF OPTICAL TEST EQUIPMENT USING LABVIEW CANDIDATO MATTEO RAMPADO RELATORE DI GLASGOW Dr ANTHONY KELLY RELATORE DI PADOVA Dott LUCA PALMIERI ANNO ACCADEMICO 2013 2014 ABSTRACT A blue array of GaN based uLEDs and a Golden Dragon Plus com mercial LED OSRAM were analysed under a pulsed current A LabVIEW program by National Instruments is used to automate the measurements of the envelope at the top of the pulse at different frequencies The program allows to save to plot and to elaborate the data measured The results characterised by noise at high frequen cies show an increase of the bandwidth with the voltage of the pulse applied In the array of uLEDs the bandwidth increases using smaller pixels However these pixels can carry a limited voltage On the other side the commercial LED carries high voltages but is limited in band width due to its internal properties Finally pulse conditions improve the perform
18. S Measurements vi calculates the peak to peak interval and the lowest value of the curve displayed by the oscilloscope Figure 33 the subVIs present are driver provided by Rohde amp Schwarz for the oscil loscope The output is an array of two strings with the envelope in the first position and the lowest value in the second position When the envelope and the lowest value are available the envelope can be added to the arrays containing all the measures ArrayRefresh vi contains the Low function explained with Figure 2 the calculation to convert the envelope in dB and the calibration Figure 34 The calibration factors have to be provided as an array control in this VI 44 MEASUREMENTS 45 1 y VISA Refnum T VISA Refnum out Results error in no error i 7 error out EH Amplitude And Time a Measurements vi Block Diagram Measurements vi VISA Refnum in VISA Refnum out Results error in no error error out b Measurements vi icon Figure 33 Measure of the envelope and of the minimum voltage Env not Cal Env not Cal Envelope meas Env Cal Waveform array ArrayRefresh vi Block Diagram ArraysRefresh vi Envelope meas Env Cal Env Cal Waveforms saved Env not Cal LOW b ArrayRefresh vi icon Figure 34 Low function envelope in dB and data calibration 46 PROGRAMMING If a new calibration is done this array has to be replaced
19. age and OFF which indicates that the waveform is recorded following the decimation mode selected 3 Set the acquisition options the digital filter the external source for the trigger and its horizontal position a fixed resolution a number of 100 acquisitions for each Run Single of the oscillo scope and fixed horizontal grid lines of the display 4 Set the offset and the scale of the horizontal axis in the display 5 Set the vertical position range offset and the resolution The acquisition options are included in the subVI OscilloscopeAcquisi tion vi shown in Figure 28 Some of them are set using direct com mands listed in 2 pp 484 884 In particular the command WAI is used to stop the execution of the program until all the previous processes are completed Also the horizontal options are included in a subVI OscilloscopeHorDisplay vi shown in Figure 29 Enable Filter Cut Off DF Hz Vertical Offset 0 0 ithmetii q arithmetic Resolution s Vertical Position V Channel VISA Refnum out 1 VISA Refnum in error in error Waveform a Oscilloscope vi Block Diagram out E S 5 Am 7 Horizontal scale s div Vertical rang Pery Horizontal offset s Werlica Oscilloscope vi Horizontal offset s Horizontal scale s div Vertical Range V Vertical Offset V VISA Refnum in Verti
20. ality of the results and the time used to obtain them For quality of the results the plots with the envelope function of the frequency are considered the best quality is given by curves that show a steady fall increasing the frequency without the presence of jumps The second solution is to average more data taken for the same experiment the following results refer to the average of 5 mea surements Even an average of more than 5 measures was tried with insignificant improvements in the quality of the results 2 1 LED MEASURE OF THE ENVELOPE USING THE PULSE Despite the use of these two expedients the results plots show still jumps especially at high frequencies In addition the curves stop to roll off quickly after a certain frequency and level out at high frequen cies These two problems happen at high frequencies where the noise is higher Since different options with the digital filter of the oscillo scope were tried the noise is present because of limits in the setup maybe in the oscilloscope or in the splitter where PG NA and LED are connected The results of the following sections are characterised by this limitation DATA ELABORATION The envelope measures in Volt are calibrated and normalised using equations 1 and 2 The data are fitted using a cubic spline fit and linearly interpolated increasing the number of data points The interpolation is necessary to find the bandwidth with a certain precision The details for
21. ance of the uLED In respect to a pulse DC current sent to the uLED leads to lower bandwidths The first part of the report presents the experiments and the results A manual for the LabVIEW programs used is given in the second and last part ACKNOWLEDGMENTS I would like to acknowledge the assistance that the following individ uals gave to me throughout the project To my supervisor Dr Anthony Kelly who assisted me well during the work of these months even in the most busy moments To Scott Watson who spent a lot of time with me in the lab and who gave me the best help possible with the equipment and the ex periments vii to my family and to Valentina CONTENTS II HIT 6 INDRODUCTION 1 INTRODUCTION 3 1 1 Abbreviations 4 1 2 Definitions 4 CHAPTERS 5 EXPERIMENTS 7 2 1 LED measure of the envelope using the pulse 7 2 11 Setup for the measurement 8 2 1 2 Different configurations the insertion of atten uators 9 2 1 3 Calibration 21 2 1 4 Error in the measurements and data elabora tion 14 2 1 5 Uu LED Results 15 2 1 6 Commercial LED Results 20 2 2 uLED measure of the envelope with DC current 22 REMOTE CONTROL 27 3 1 Connections 27 3 2 Front Panel 27 3 2 1 LOG IN 27 3 2 2 RUN 29 3 2 3 CONFIGURE 31 3 2 4 RESULTS 35 PROGRAMMING 37 4 1 General advices about the programs 37 4 2 Configuration 38 4 2 1 Pulse Generator 38 4 2 2 Oscilloscope 38 4 2 3 Network Analyzer 41 4 3 Visualisation 41 4 4 Measu
22. asure is displayed together with a led indicator that is ON when the program is taking measures The waveforms displayed by the oscilloscope during the 3 2 FRONT PANEL measurement are reported in the graph Display Oscilloscope Thus the user can control the display easily looking at the laptop Calibration a a io gt 10 100 2 Figure 22 Results vi front panel Results allows to look at the graph Envelope Frequency of the last measure taken Figure 22 Two curves are plotted respectively for the calibrated and not calibrated results It is possible to save the image of the graph with the name in the same format of the data files with the word IMG added at the beginning e g IMG 5V IDC Olimarl1h18m20s png As indicated in Chapter 6 Future Work an improvement of this option may be to calculate and to show the average of more results already made by one of the VI built to elaborate the results Pressing the button Initialise amp Configure the instruments are ini tialised and configured Finally with Quit is possible to close the actual window and to return to the main program 3 23 CONFIGURE This section of the program allows to make save and recall config uration files containing the parameters for the PG and the Oscillo scope In figure 25 on the left of the window there is a similar tab contr
23. ator and of the Oscilloscope Thus once the correct parameters for one experiment are correctly defined they can be saved and used more times Each configuration file is associated 57 58 CONCLUSIONS with a log in username password which is loaded automatically with the log in FUTURE WORK the project was characterised by suggestions and ideas for future work The following list contains some proposals for further experi ments The main issue of these experiments is the presence of the noise at high frequencies Additional equipment can be used to solve this problem such as isolators and or a BIAS TEE instead of the splitter In such sensitive measurements data resolution is fundamental for the quality of the results Here resolution means the interval of time between two data samples All the measurements were done with a resolution of 10 ps set in one of the first versions of the program where all the data regarding the waveforms dis played were saved This version was too slow to save many data for higher resolutions The last version of the program saves for each file only a number of envelope measures equal to the fre quencies analysed For this reason it is now possible to increase the resolution without having too much time per measurement and data save Other measurements may be done with a bigger resolution To improve the quality of the results another solution may be adopted such the increase of the
24. by the new one Path of the file containing the Envelope measures using only the instruments size s max value of the data in Volt 3 ci calibration factors Figure 35 Calculation of the calibration factors Figure 35 shows the program that calculates automatically the cali bration factors from an xls file at the input The indicator calibration factors has to be copied into ArrayRefresh vi where it can be changed to control for the new calibration 47 44 MEASUREMENTS IA uny JO jo Weg 9 48 PROGRAMMING An overview of how more measurements are made consecutively is given in Figure 36 The code seems to be complex but the key features are very simple There is a flat structure divided into three frames In the first one a frequency is selected from a list and the NA is configured to this frequency The frame in the middle simply stops the execution of the program for 1 second allowing the system to change under this new frequency Then the measures are taken the actual data are plotted into a graph and the envelope is processed with the programs explained in the previous sections this code is included into two For Loops The internal loop processes all the frequencies When the iterations are finished the data in the array are saved The external loop is needed to repeat all this task for the number of measures selected by the user 4 5 DATA SAVE SaveEnvelope vi saves the
25. ca Pulse generator frequency LED Electrical analyser p Oscilloscope Laptop Figure 1 Connections for the measure of the envelope using the pulse other hand the signal goes directly to the commercial LED The light of the LED is detected and converted into an electrical signal by the Photoreceiver This last signal goes to the oscilloscope where it is analysed 2 1 1 Setup for the measurement Firstly the PG is set for a pulse with a period of 100 us and a duty cycle of 1 1 us of pulse width Furthermore the is set to sweep frequency from an input list in the interval 1 MHz 1 GHz At this point the input signal to the LED is fixed the controls of the oscilloscope have to be set for a correct visualisation of the data For this reason the next controls are selected to constant values Resolution 10 ps the finest resolution that allows to save the data of the waveforms It is possible to improve it see Chapter 6 Future Work Arithmetic Envelope to detect the minimum and maximum values in a sample interval over 100 acquisitions Two curves are created one with the maximum and one with the minimum values Vertical Coupling DC 50 it is required from the specifica tions of the Detector 3 p 1 Trigger Source External the trigger output of the PG is con nected to the external trigger input of the Oscilloscope Other controls such us the horizontal and vertical
26. cal Position div Vertical Coupling Resolution s Trigger Level V error in no error Enable Filter Acquisition Type Channel Cut Off Hz Arithmetic VISA Refnum out error out b Oscilloscope vi icon Figure 27 Configuration for the Oscilloscope 39 PROGRAMMING 40 edooso rso ay Jo uonrsmboe ayy 10 ge 91314 uoot 10 uo ismboyado9so 11980 10448 OU 40443 A 1966111 suu 92410594 VSIA 40 in puuey 2721 ajqeuq ino 20449 no SIA 1 uonisinb yado2s0 11250 usu 19018 10 02801 1250 e AE 534 ano VSIA 001 NNOJDIV NOXiXDviddsia 130 WITSIJO TOL T puuey 32Jn0534 VSIA 1144 Buuey HO MI esi A 19437 1966111 4 3 VISUALISATION VISA out a OscilloscopeHorDisplay vi Block Diagram OscilloscopeHorDisplay vi Channel 1 VISA resource name VISA out horizontal scale s div horizontal offset s error in no error error out b OscilloscopeHorDisplay vi icon Figure 29 Configuration for the horizontal axis in the display of the oscillo scope 4 2 3 Network Analyzer NA vi changes the frequency in the form number unit which need to be analysed Figure 30 Firstly the program converts the frequency selected in Hz A case structure is driven by the unit string and for each case an appropriate ca
27. dB compared with its measure at the initial frequency The experiments are repeated changing some parameters of the pulse as the voltage and the duty cycle Finally only for the uLEDs the experiment is re peated under continuous current The second chapter describes the optical equipment and the experiments done showing results and their elaboration Automated systems are becoming more and more important in the technological development Computer programs facilitate the work of engineers and designers allowing to save time especially for repet itive measurements as in this case So the best way to do the mea surements for the calculation of the bandwidth is to control the in struments automatically LabVIEW presents two main environments strictly connected each other The Front Panel is the user s interface formed by controls and indicators which allow the user to control the instruments and to have access to the results It is presented in the third chapter together with the connections laptop instruments The Block Diagram is the code of the program in the form of graph ical programming It contains icons representing either functions or subprograms The code is investigated in the fourth chapter For the reader new to LabVIEW Appendix A offers a brief description of some features used in the program The Abbreviations and Definitions present throughout the report are given below 4 INTRODUCTION ABBREVIATIONS ADC Analogu
28. dB domain each envelope is sub tracted by the maximum one For the interpolation a new set of fre quencies is calculated such that the minimum f1 and the maximum fm frequencies do not change The difference between them is di 4 6 DATA ELABORATION Mac Users matteorampado Desktop Results Voltage Amplitude V Duty Cycle 5 da le PRO ES Envelope calibrated Envelope non calibrated peti a SaveEnvelope vi Block Diagram SaveEnvelope vi Duty Cycle 5 Envelope calibrated Envelope non calibrated Enable Filter True Voltage Amplitude V b SaveEnvelope vi icon Figure 37 Envelope saving 49 50 PROGRAMMING a SaveSS vi Block Diagram SaveSpreadsheet vi file path array b SaveSS vi icon Figure 38 Save into Spreadsheet file E True Voltage Amplitude V Enable Filter True Duty Cycle 5 Figure 39 Code for saving the graphs 4 6 DATA ELABORATION vided by a constant number e g 100 to find new df A for loop builds the new array of frequencies It starts from f and adds this first frequency to the array Then it finds f f f and it repeats the loop until fx fm is added to the array The number of itera tions for the loop is then M 1 x 100 with M the number of the initial set of frequencies The interpolation function of LabVIEW calculates the data corresponding to
29. e graphs envelope frequencies Again the LP filter was changed to be two times the frequency analysed until 500 MHz and a filter of 1 GHz after that threshold This because the maximum LP filter allowed is of 1 GHz A lot of jumps were observed after 500 MHz Figure 6a so the final decision was to put the same filter until 500 MHz and no filter for the remaining frequencies examined Figure 6b For the commercial LED the situation is different Previous and dif ferent measurements showed a bandwidth around 23 MHz Anoma lies arise using a LP filter of 250 MHz with the bandwidth that reaches a peak close to 50 MHz around 11 V before falling down around 20 30 MHz at higher voltages For this reason I chose a digi tal LP filter of 50 MHz which avoids high frequency noise and which is far enough from the bandwidth expected These last considerations are recalled in the paragraph Commercial LED 2 14 Error in the measurements and data elaboration Especially for the uLED this kind of measurements are very unstable and sensitive to vibrations To avoid errors two solutions were tried The first one is increasing the number of acquisitions per measure ment in the oscilloscope as given in 2 p 39 On the other hand it is not possible to put the maximum number of acquisitions allowed by the oscilloscope because this causes too much time per measure ment The number of 100 acquisitions was found to be the best ratio between the qu
30. e into Spreadsheet file 50 Code for saving the graphs 50 Block diagram for the elaboration of the data normalisation fit interpolation and BW calcu lation 52 State machine for the main program 53 Block diagram of the main program 53 Some structures of LabVIEW 64 Example of the use of the Shift Register 65 Part I INDRODUCTION INTRODUCTION Directly from the title Optical and Automation are the key words of this project Optical is related to the equipment used Automation is represented by the use of LabVIEW They include the two main aims to find the bandwidth of some devices and to design a program for the control of the equipment Two different LEDs are analysed a blue array of GaN based uLEDs made by the institute of photonics University of Strathclyde and a Golden Dragon Plus commercial LED OSRAM The array of uLEDs is a square of 16 x 16 pixels In each row there are from left to right and in the form number x diameter 2 x 5um 2 x 10um 2 x 15um 2 x 20um 2 x 30um 2 x 40um 2 x 50um 2 x 60um pixels The commercial LED has a diameter of Imm Firstly the LEDs are sub jected to pulse conditions The part of interest is the top of the pulse sent the interval of time when the LED is ON In this interval the envelope measure is taken at different frequencies to find the band width of the LED In this report the electrical bandwidth is consid ered such the frequency at which the envelope is reduced by 3
31. e main interface When the Exit button from the subVI opened is pressed the actual Front Panel window is closed and the main interface is displayed again Figure 42 shows the block diagram for MainProgram vi The ma chine is represented by the While Loop and the states programs are the cases of a string Case Structure The Idle case is displayed where the program waits until a button is pressed A specific subVI sends the command with the name of the button pressed to the suc 54 PROGRAMMING cessive iteration so that a new case is selected and a new program opened State machines are also present into and Configuration vi Part III CONCLUSIONS CONCLUSIONS 5 1 EXPERIMENTS The results obtained for the experiment using the pulse are affected by noise at high frequencies especially for the uLED that is investi gated to higher frequencies than the Commercial LED In addition the measurements with the are very sensitive to vibrations and have to be taken very carefully However the results give some impor tant indications The two experiments with a pulsed and a continuous current show an increasing of the bandwidth with the voltage applied This happens considering devices with different characteristics such are the uLED and the Commercial LED The uLED has a limit in the voltage that can be applied This limit can be extended decreasing the duty cycle of the pulse applied On the other hand the Commer
32. e results for the uLED under pulsed and continuous current Figures 16 and 17 show the comparison between the results ob tained by a pulsed and continuous current At the same voltage the pulsed curves stay over the ones obtained with continuous current As for pulsed conditions the bandwidth is bigger for pixels with smaller dimensions In addition for the same pixel the bandwidth increases with the voltage Finally the bandwidths at 5 V for the DC data are 103 and 195 MHz respectively for the 2nd 60 and the 5th 40 um uLED This is consistent with the bandwidth found for the 3rd uLED 50 um with a duty cycle of 100 which is 143 MHz as given from the analysis with different duty cycles 23 24 EXPERIMENTS 22 n 4 e Envelope dB 1 1 6 16 18 204 1 i 1 10 100 1000 Frequency MHz 2nd uLED av ev Fr pny DCS V a S 7 gt 12 T 1 1 10 100 1000 Frequency MHz b 5th uLED 07 4v sv DC4V 2 ncs v Envelope dB 104 114 12 1 10 100 1000 Frequency MHz c 8th uLED Figure 16 Envelope function of the frequency at different voltages for pulsed 1 duty cycle and direct
33. e shift register in Figure 44 does the sum of the first ten numbers The iteration number i goes from 0 to and the shift register is ini 2 TYPICAL FUNCTIONALITIES USED 65 shift register Figure 44 Example of the use of the Shift Register tialised to 1 The sum of the value contained by the shift register and the iteration number is saved into the shift register for the successive iteration Thus the first iteration gives 1 0 1 the second 1 1 2 the third 2 2 4 the fourth 4 3 7 and so on The shift register is widely used for the state machine programs e g it is present in figure 42 where it is necessary to pass data from a case to another or simply when the case is changed the new command is passed to the case structure in the next iteration through a shift register In figure 40 the shift register is instead used to make the new array of frequencies for the interpolation BIBLIOGRAPHY 1 66 2009 Data Science Automation Documented Source Code 2000 URL http www dsautomation com doctool Rohde amp Schwarz GmbH amp Co KG R amp S RTO Digital Oscillo scope User Manual 2013 URL http cdn rohde schwarz com dl downloads 41 _ common library dl manuals gb 1 _ UserManual en pdf FEMTO Messtechnik GmbH Ultra High Speed Photoreceiver with Si PIN Photodiode 2011 URL http www femto de datasheet DE HSA X S 1G4 SI R7 pdf Agilent Technologies Programmer s Guide 2
34. e to Digital Converter BW Bandwidth DC Direct Current or Duty Cycle it is evident from the con text DF Digital Filter GPIB General Purpose Interface Bus LAN Local Area Network LED Light Emitting Diode LP Low Pass referred to the filter NA Network Analyzer PG Pulse Generator USB Universal Serial Bus VI Virtual Instrument also present as vi or vi DEFINITIONS Envelope distance between the minimum and maximum val ues detected in a sample interval over a certain number of ac quisitions In this report this concept is referred to the data ac quired by the oscilloscope Pulse in the Signal and Systems language it is a Rectangular function here translated or changed in amplitude and width case by case Virtual Instrument program of LabVIEW Part II CHAPTERS EXPERIMENTS the project revolves around two kinds of Light Emitting Diode LED a blue array of GaN based uLEDs and a Golden Dragon Plus commercial LED As indicated in the Introduction the array of uLEDs is provided by the Institute of Photonics of the University of Strath clyde whereas the Golden Drangon Plus LED is given by OSRAM The analysis is divided into two main sections which differ from the current at the input of the LEDs This is respectively pulsed and con tinuous for the first and the second section Both the experiments focus on the calculation of the bandwidth of the devices under different condition
35. edly the instruments at every iteration of the while loop that contains them The driver rss Read Waveform gives as output an array containing all the data actu ally displayed by the oscilloscope The curve is made using the Bundle function of LabVIEW which generates the waveform from the initial time to the interval of time t and the data The initial time to is the horizontal offset that has been chosen On the other hand the incre ment which is the interval of time between two consecutive data is given from the resolution divided by 2 This because for each inter val of time between two samples which is the interval indicate by the resolution the envelope mode selects two points the maximum and minimum which are inserted into the waveform array of data 43 4 3 VISUALISATION pesn 8guoo ay JO uonezsng pue s o3juoo 71 jo ure18erp po q au Jo Weg 1 amg ad boe E 191 A 9 21 1265111 5 uonnjosay 5 19530 02 B 3 Dd qe na Ein paxi4 POW asind 5 ponad 56 AD Aang EEE ta 5 asing 3 qeu3 1 andino a qeu3 s Aejag orn suun EE 00250 11250 Aejdsiq psg mE A
36. ents The graph Display Oscilloscope shows the data displayed by the os cilloscope during the configuration Once the configuration is done Finish Configuration has to be pressed before saving the file With the control Path In for Recall Save is possible to select the destina tion of the new configuration file An important advice is that the file saved does not go automatically in Configuration File Specification Out from which is then loaded in Run To select this file for the output it is necessary to press the button Recall after the button Save If a configuration file for the experiment already exists it can be recalled simply selecting its path and pressing Recall Alternatively the Default file can be recalled pressing Default Finally to come back to the main program with the last configura tion file selected OK has to be pressed Otherwise Cancel allows to close the window without making any change in Configuration File Specification Out 3 2 3 1 Pulse Generator The controls of the PG can be divided into two groups as in Figure 24 The first group contains the quantities for the pulse the Delay Value and Unit the Period the Width the Duty Cycle and the Voltage Am 3 2 FRONT PANEL cpipo 14 iNsTR o seconds o Pomo 1 01 Figure 24 Configuration vi front panel tab with the controls for t
37. he Oscillo 29 30 REMOTE CONTROL Figure 21 front panel tab NA scope and GPIB 16 for the NA Then it is necessary to press the button VISA SET which initialises and according to the configura tion file selected sets the parameters for the instruments The list of frequencies of the NA can be changed before or after pressing VISA SET Now the instruments are ready for the measurement It is possible to select the number of measures to do through the apposite con trol under the diagram Display Oscilloscope After this step press ing the button Run amp Save the measurements are done automatically and the calibrated and not calibrated results are saved These are con tained in two distinct xls files per each measurement in the folder Results which has to be already present in the Desktop with the name in a specified format For example ON 4V 21marl5h27m57s CAL xls is the file associated to a PG with output ON amplitude 4 V and duty cycle 1 The file was saved in date 21st of March at the time 15 27 and 57 seconds The results are calibrated This format was cho sen to distinguish the measurements with different signals at the in put The date is important especially when the same kind of mea sure is repeated more times The time is necessary for not replacing more measures taken consequently under the same conditions Dur ing the measurement the number of the actual me
38. he PG plitude The second group indicates general parameters as the Pulse Mode Single Double the Fixed Parameter DC Pulse Width and the Polarity Positive Negative An external control allows to enable dis able the Output of the PG With DC as Fixed Parameter the width cannot be modified but changes with the DC The vice versa applies The functionalities for the other controls are easily understandable 3 2 3 2 Oscilloscope The controls of the oscilloscope are divided into four groups which relate the display the digital filter the acquisition and the Vertical Coupling Trigger level of the oscilloscope Figure 25 The vertical and horizontal offset define the start point of the respective axis in the display The horizontal axis has ten divisions each of a size fixed by horizontal scale There is not an equivalent function for the vertical axis here substituted with the Vertical Range which is the range be tween the maximum and minimum voltage displayed In addition to the offset there is a Vertical Position slide which moves the display up and down by a number of divisions each of amplitude Vertical Range 10 The continuous input signal of the oscilloscope is converted into digital form by an ADC A digital LP filter can be applied to reject high frequencies as explained in 2 p 51 The acquisition controls include the arithmetic mode which is the way how the waveform is made from more acquisitions of the input sig
39. lculation is made for the conversion Fi nally the number of the frequency is converted into string to form the GPIB instruction CWFREO lt freq gt required to change the frequency of the NA see 4 The outputs of the program are a string indicator with the frequency selected in the form number unit and a number indicator with the frequency selected in Hz 4 3 VISUALISATION Figure 31 shows an example for the data visualisation This is a part of Configuration vi and starts from the while loop on the left Once VISA SET has been pressed in the front panel it is possible to configure the instruments and the results of the configuration are automatically 41 42 PROGRAMMING Frequency VISA resource name Indicator Frequency Freq in Hz Control MHz Default 1009099 3 Unit MHz by default VISA resource name out a NA vi Block Diagram NA vi VISA resource name VISA resource name out Frequency Control Frequency Indicator Unit MHz by default Freq in Hz error in no error error out b NA vi icon Figure 3o Configuration for the Network Analyzer plotted in the Waveform Graph Display Oscilloscope The configura tion files for the PG and the Oscilloscope are inside case structures so that the instruments are configured only if either the button Config ure Pulse Generator or Configure Oscilloscope is pressed With this feature the program does not configure repeat
40. lope the data storage to spreadsheet files or to images Following this scheme I made many programs From very simple VIs which allow to find the envelope associated only to one fre quency to more complex VIs that process list of frequencies From manual to automatic save and visualisation of the data Then the divi sion of the configuration part from the measurement adding the pos sibility to take and save more than one measurement consecutively For the limits of space imposed only a part of the final programs are presented 4 1 GENERAL ADVICES ABOUT THE PROGRAMS The program used is LabVIEW 2011 v 11 0 for Mac OS but the VIs made work correctly also on other operative systems with only few changes The paths for example are different from the ones on Mi crosoft Windows The mark of division in the path on is and on Windows is In addition all the paths present in the program have to be verified with the actual paths of the pc used Part of the given VIs are instrument drivers available from the web sites of the manufacturers To distinguish my VIs from the drivers I put an orange triangle with a blue o letter in the bottom right part of the icon relative to each VI The state machine concept and the way to make and recall the configuration files cfg are taken from 1 They were completely done from the beginning adapted for Lab VIEW 2011 and changed following the scope of the measurement of this project
41. nal the resolution and the acquisition type which indicates the decimation mode selected This last option is automatically Peak de 33 34 REMOTE CONTROL 6 TCPIPO 192 168 0 2 a Figure 25 Configuration vi front panel tab with the controls for the oscillo scope 3 2 FRONT PANEL tect if the Envelope mode is active With Peak detect the minimum and maximum samples between more data in a sample interval are selected to form the waveforms displayed one with all the maximum and one with all the minimum samples 2 p 36 3 2 4 RESULTS This section does the same functionalities of the Results button in Run On the other hand every input data can be selected not only the last data measured The user has to select two file the first with calibrated data and the second with the correspondent raw values The program can be modified to make a comparison of any group of data 35 PROGRAMMING With this chapter I try to present a part of the code that is under the interface explained in Chapter 3 Appendix A can be useful for the reader new to LabVIEW who finds difficult reading the code I divided the control into three tasks the configuration which focuses on the parameters of the in struments that can be changed the visualisation which relates the waveforms plotted in graphs and possibly shows the same data displayed by the oscilloscope the measurement strictly connected with the enve
42. number of acquisitions by the oscilloscope The time per measurement increases but the data saved are more precise Only three pixels where analysed in this project This work can be extended for all the pixels of the uLED s array In addition the measurements can be repeated for different duty cycles with the calibration of the instruments for each duty cycle A Bit Error Rate Test on the uLEDs can be done using a Pattern Generator connected to the NA a source of BIAS the detector and the oscilloscope It is then possible to analyse the results displayed by the oscilloscope sending different sequence of bit This allows to understand if the device is error free looking at the effect of increasing the frequencies More errors are expected at high frequencies Furthermore the VI that makes the average of more measure ments can be inserted inside Run vi such that the average is 59 60 FUTURE WORK done immediately after the measurements and the plot of these results is given Finally the data elaboration may be added in the main program selecting the measurements available from a previous run Part IV APPENDIX LABVIEW TOOLS This appendix offers a small guide about LabVIEW looking in more detail to some functionalities explained in the previous chapter For limits of space it is not possible to explain all the programming fea tures and possibilities The reader unsatisfied by this section can find a com
43. ol of Run vi without the tab associated to the NA The first tab includes some indicators relative to the paths of the configuration files and a list of objects with four fields For this project only the field called Group name was used The other fields whose types are one boolean Display and two numbers respectively unsigned Filter and double Time can be useful for future uses of the pro gram Only when the first tab is selected the buttons SAVE RECALL DEFAULT are enabled 31 32 REMOTE CONTROL me 12Vconf Cw 1146 S Signal name pug Ena iat Enable All To M Disable iersmatteorampado Documents Uni presente Tesi Labview 1 experiment Configuration Files onf cfg Mac Users matteorampado Documents Uni_presente Tesi Labview 1 ex 1LED 11Vconf cfg Figure 23 Configuration vi front panel When the button CONFIGURE is pressed in the main program a dialog box appears asking if the user wants to make a new con figuration file In case of positive answer the VISA SET button has to be pressed after the selection of the VISA addresses for the PG and the Oscilloscope Now the parameters for the two instruments can be changed These controls are present in the specific tabs asso ciated to the instruments The buttons Configure Oscilloscope and Configure PG apply the new configurations on the instrum
44. p Flat Sequence structure case E selector stop condition Figure 43 Some structures of LabVIEW The structures are the equivalent instructions used in other com mon programming languages as the While loop the For loop and the Case structure if these structures and the Flat se quence structure are given in Figure 43 The While loop executes until the control connected to the stop condition becomes true it is possible to change the condition in order to keep executing the loop as long as the control is true The For loop executes for a number of times equal to N A numeric control can be wired to the symbol N to fix the number of executions For both the While and the For loops the symbol i contains the number of the actual iteration It can be wired as a control for different proposals The case structure may have different types of cases i e boolean number string etc The ac tual type is the same one of the control wired to the case selector This control select the case that the program has to execute The other not selected cases are ignored by the program Finally the flat structure executed the code for successive frames from left to right The code in a frame is executed only once it has received all the data coming from the previous frame The shift register is a functionality that allows to keep the value of a variable from one iteration to the successive in a loop For example th
45. plete reference about LabVIEW in 5 ABOUT LABVIEW A program in LabVIEW is called Virtual Instrument VI a name that forms the extension for the files built using this program The source code is made with a language called G a graphical program ming made of icons wires and structures The icons are functions or VIs called subVIs if we are referring to the VI that includes them They are connected together by wires All revolves around the con cept of dataflow each part of the program is executed only if it has received all the necessary data from the input For example consider the case of a program A with an output O connected as input to the program B This last program is executed only after A since it has to wait for its input O provided by A As introduced briefly in the programming there are two main envi ronments strictly related together the Front Panel and the Block Di agram The former recalls the buttons and the indicators of the real instruments the latter is the source code that made possible the con trol As for the real instruments there are controls and indicators here in form of variables These are present in both the front panel and the block diagram A control variable provide a certain type of informa tion e g number string or boolean In the front panel the user can modify this kind of variable Its equivalent in the block diagram can be wired as input to one program or function An indica
46. re the bandwidth falls down with the increase of the duty cycle This result is recalled with the experiment under continuous current equivalent to the case of 100 duty cycle 2 1 6 Commercial LED Results The measures were done using the same configuration of the uLED except for the list of frequencies and the digital filter of the oscillo scope Since the bandwidth is expected to be around 23 MHz from other experiments previously made the analysis is done from 1 to 30 MHz at every 1 MHz sample then is more sparse and it stops at 100 MHz Figure 12 shows the bandwidth function of the voltage using a dig ital LP filter of 250 MHz Nevertheless these results are not calibrated it is evident that the peak of the bandwidth around 11 V is not realis tic However a LP filter of 50 MHz is not restrictive it is far enough from 23 MHz and such anomalies are avoided The analysis is done for voltages in the interval 4 14 V using the Low function for 12 14V due to the situation explained with Figure 2 The results given in Figure 15 show a similar behaviour investigated with the uLED Looking at two different curves the one associated to the bigger voltage stays steady for a wider range of frequencies and falls down at higher frequencies This is evident until 10 11 V then the curves 21 LED MEASURE OF THE ENVELOPE USING THE PULSE 21 Comm LED filter 1 un o gt vi 10 7 V g
47. rements 44 4 5 Data Save 48 4 6 Data elaboration 48 4 7 State machine concept and MainProgram vi 53 CONCLUSIONS 55 CONCLUSIONS 57 5 1 Experiments 57 5 2 Programming 57 FUTURE WORK 59 CONTENTS IV APPENDIX 61 A LABVIEW TOOLS 63 About LabVIEW 63 2 Typical functionalities used 64 Bibliography 66 LIST OF FIGURES Figure 1 Figure 2 Figure 5 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24 Figure 25 Figure 26 Figure 27 Connections for the measure of the envelope using the pulse 8 Solution used in case of no flat waves displayed by the oscilloscope 9 Envelope in dB versus Frequency in MHz pixel 40um 5 V 10 Connections for Calibration 11 Calibration 5 V 1 Duty Cycle 12 Calibration 5 V and 1 Duty Cycle 13 Comparison Voltages and uLEDs 16 Comparison Voltages and uLEDs 17 Comparison Voltages and uLEDs 18 Bandwidth function of the voltage for all the uLEDs 19 Envelope vs frequency for different Duty Cy cles 5 V 3rd uLED 50um first row 19 BW versus Voltage for the Commercial LED DF 250 MHz raw values 21 Comparison Voltages for the Commercial LED DF 50 MHz 21 BW versus Voltage for the Commercial LED DF 50 MHz 22 Connections for the measure of the envelope with DC current 22 Envelope function
48. s This is done through the measure of the envelope on the output signal of the LEDs at different frequencies Furthermore the envelope is expected to be steady at low frequencies and to roll off quickly at high frequencies The instruments used for the experiments and their functions by which they are used for are listed below 1 Pulse Generator HW 8114 to analyse the LEDs under pulse conditions 2 Network Analyzer HP 8753ES to analyse the LEDs at different frequencies 3 Ultra High Speed Photoreceiver HSA X S 1G4 SI made by the company FEMTO Messtechnik GmbH called also Detector in this report to detect the optical signal generated by the LED and to convert this signal into its electrical form 4 Oscilloscope Rohde amp Schawarz RTO1022 to show the data generated by the Photoreceiver to make measurements and save the results 5 DC current supply Thurlby Thandar PL 3100MD to analyse the LEDs under continuous current 6 Digital Multimeter Keithley 195A to improve the source of continuous current to the order of mA 2 1 LED MEASURE OF THE ENVELOPE USING THE PULSE For this experiment the instruments 1 4 are used connected as in Figure 1 A pulse is generated from the PG and goes inside a T splitter where the NA and the LED are connected In case of the uLED the signal goes through a probe connected to the pixel analysed On the EXPERIMENTS ma Detector ES mu O pulse p 1 Opti
49. t Pick 2 a MH N u 1 Bandwidth 3 35 4 45 5 55 6 65 7 75 8 85 9 95 10 105 11 115 12 12 5 13 Voltage V Figure 12 BW versus Voltage for the Commercial LED DF 250 MHz raw values 4 sv x x LI 2 5 s p x lov 3 MA LA x n 3 3 5 E S 12 v LL S QA amp SR e 4 5 TE UT Av j g 5 a E te nin 4 5 5 x tm tte ui x nei Er Su 707 n ILE TET Bid 6 5 x 7 PS IES 7 5 wu w i dii d LT 8 5 1 10 100 Frequency MHz Figure 13 Comparison Voltages for the Commercial LED DF 50 MHz 22 EXPERIMENTS Comm LED filter 50 MHz 7 Bandwidth MHz M o 9 10 11 12 13 14 8 Voltage V Figure 14 BW versus Voltage for the Commercial LED DF 50 MHz assume a similar shape and position in the graph The bandwidths at these voltages are respectively 23 24 MHz then there is a peak of 26 MHz at 12 V On the other hand using the fact that the curves stop increasing after 10 11V the bandwidth can be fixed at 23 24 MHz To sum up the analysis at different voltages shows the same be haviour for the uLED However the bandwidth
50. t for a wider range of frequencies before falling down at high frequencies Figure 7 8a This suggest larger bandwidths for higher voltages A comparison for different uLEDs is given in Figures 8b 9a and 9b respectively at 4 5 and 6 V In this case comparing two uLEDs the curve associated to the bigger remains under the other curve The only exception is Figure 8b where the 15 16 EXPERIMENTS 4V 0 5 d SV 1 Bm t 6v 1 5 E E 7V _2 m 2 9 65 2 5 Pa a 2 LN xA 2 3 54 y Y 4 x gt a LU 1 10 100 1000 Frequency MHz a Comparison of the Voltages 2nd uLED Envelope dB ES 1 E 4 5 5 5 5 1 10 100 1000 Frequency MHz b Comparison of the Voltages 5th uLED The 7 V curve does not reach the 3dB point Figure 7 Comparison Voltages and uLEDs 21 LED MEASURE OF THE ENVELOPE USING THE PULSE 17 0 25 0 5 0 75 1 25 n T 71 754 0 v 2 25 2 5 2 75 Envelope dB 3 25 3 5 3 75 44 1 10 100 Frequency MHz a Comparison of the Voltages 8th uLED The 6 V curve does not reach the 3dB point 04 Envelope dB 74 i i 10 Frequency MHz b Comparison of the uLEDs 4 V Figure 8 Comparison Voltages and uLEDs 18 EXPERIMENTS
51. the fitting and the interpolation are given in Chapter 4 because all these operations are done using LabVIEW In some situations the envelope measured at the lowest frequency is not the biggest as expected Thus the first envelope is not at 0 dB To compare different data this kind of curves are translated to have the first point at 0 dB 2 1 5 Results The first and the second row of pixels were investigated The results for the 2nd diameter 60 um 5th 40 um and 8th 30um pixels of the second row are presented in this section Due to the different amount of voltage that the LEDs can support the 2nd LED was anal ysed for 4 5 6 7 V the 5th LED for 4 5 6 7 V and the 8th LED only for 4 5 6 V Considering by convention the top of the pulse from 0 to 1 us in the time scale the envelope is calculated looking the interval 0 1 0 9 us to avoid the oscillations at the beginning and at the end of the top of the pulse The frequencies analysed are 1 10 MHz with a sample every 1 MHz then 25 500 MHz sampled every 25 MHz and 500 1000 MHz sampled every 50 MHz Figures 7 8 9 show the envelope in function of the frequency for the uLEDs analysed The noise is higher at high frequencies and the curves do not roll off quickly especially at high voltages In addition to this some curves do not reach the 3 dB point However some important indications can be extrapolated from the graphs Increasing the voltage the curve stays fla
52. tly the two instruments it is possible to make and run the program for the experiments Figure 18 shows the front panel of MainProgram vi There are 5 but tons LOG IN RUN CONFIGURE RESULTS QUIT To start the program ctrl R Windows cmd R Mac have to be pressed QUIT allows to stop the execution of the program The first time that the program is run the log in window is automatically opened which is the front panel of the VI LogIn vi 3 21 LOGIN A common log in has to be done to use the functionalities of the pro gram Figure 19 The buttons Run and Configure are disabled until the user has logged in correctly Each username is associated to a LED or to a pixel of the uLED array The configurations parameters previously defined for the LED associated are loaded automatically with the log in If it is the first time for a log in with a LED a default file of configuration parameters is loaded This is very useful consid 27 28 REMOTE CONTROL Figure 18 MainProgram vi front panel Figure 19 LogIn vi front panel 3 2 FRONT PANEL ering that e g the vertical offset control of the Oscilloscope which has to be changed for having the maximum vertical resolution is dif ferent for each LED Thus using the log in the instruments have to be configured only one time per experiment Further measurements for the same experiment use the configuration file already defined 3 2 2 RUN
53. tor variable is the program output shown in the front panel Every VI has an icon associated and can be included in another VI in the form of this icon The controls and the indicators in the front panel of the VI can be made as the external inputs and outputs of the VI which appear automatically as inputs and outputs of the icon and which can be wired to other icons and functions For example SaveSS vi Figure 38 has two external inputs path and array which are connected to the respective controls in SaveEnvelope vi that includes SaveSS vi 63 64 LABVIEW TOOLS A 2 TYPICAL FUNCTIONALITIES USED VISA address and error in are control variables present in all the pro grams involving instruments The VISA address can be a control or an indicator and contains the address of the instruments It indicates to LabVIEW the way to communicate with a specific instrument The error is another variable common in the programs It is an essential feature of the programming because it allows to have information about the kind of error that occurs with the stop of the execution of the program the programs that relate instruments should have an the error variable as input and output of the VI in order to give the possibility to wire the error from an icon to another one of the same instruments Following this way the error propagates from one program to another and from the output it is possible to understand where is the error While Loo
54. tors are ES Ui where ij 1 i is the index of the frequency analysed The envelope is then calibrated using this expression Yicat 20108 ci 1 12 EXPERIMENTS Finally yi ca is normalised in this way Uimorm Uical max Y1 cal D UN call 2 2 1 3 2 Calibration Problems and Solutions Calibrated values Raw values Vai 1 Envelope dB y 1 ba 1 ll 11 u NEL TI n 25 12 rm cm 13 mi 144 L 1 10 100 1000 10000 Frequency MHz a Calibrated and not calibrated results DF 250 MHz frequencies up to 1 5 GHz 4th 50um first row Calibrated values Raw values ca 9 gt 97 i 1 1 10 100 1000 10000 Frequency MHz b Calibrated and not calibrated results DF OFF 4th uLED second row Figure 5 Calibration 5 V 1 Duty Cycle The first calibration was done using a digital LP filter in the oscil loscope of 250 MHz The results show a wrong calibration because the envelope measures increase at high frequencies Initially it was 21 LED MEASURE OF THE ENVELOPE USING THE PULSE 13 Calibrated values Raw values a m lini 0 5 izle wa Ba LITT gi 1 p A n heiss VITTI Je Eu E ung
Download Pdf Manuals
Related Search