Home

ECG Demo Board Final Report

Contents

1. Attachments ica 815 ISIS a 12 213 Techni IX Append Bill of Materials 000 xo s s it 313 eit ls SIS DU 9e a ores 61 0 Loan be KSE 1esosans EE SE Kon to moun mou ro moun Moun moun 810 LR RG moun kai SCT SIZO veo moun DU abe 3 S G34 dWONNY dIHD agi Od DILSW1d AINN A6 Leg 3JATIOH OWS 9021 265 MEIL WHO Xz 8 SJA OWS FOZIL SC wi WHO 3095 SJA OWS 9021 S MEIL WHO XLS SIN AWS 90ZI SC M7 1 WHO XOOE Sa OWS 9021 S AFIL WHO ZZ eg OWS 9021 25 MP WHO WO L 529 AWS FOZIL S weil WHO al SJA SDE L ASA AOL JNO AID dY AWS 9021 265 MEI WHO 201 Sad OWS 90ZI 5 M 1 WHO X001 SF OWS 9021 25 MEIL WHO 251 SIJ 9071 XLX BOL ASZ ANTZZ O HID dVD 9021 ALZX BOZ ADO 4N L O YI dY 9021 BZX AOD JNEEO O AFO dY 9071 XLX ACEN 4d00001 YID dY SIS AOSZ JNS 100 X39 dY OWS 90ZL 25 M7 l WHO 001 Sg Fen pen n LL WR Se Fen RE en ko fot 62dJeADizsunouo peoupapy L OSIZOWSH OO ONTLSZEZ9IS CT amawspay Oo PAT Opel LI TT H AZZOFAJOETAI ONE Rd Z sam i AV9STAJOBTAF GN 1Day0 Sd Z em ACISTAIOBTA Loomsd ot a Avoerazoerea ON 403X00 d CT a Acezrazoerea Loutecd CT oa ASOUAJOBTAF NOG Ld ot Aveirazoerea ONDaxozId z aasa i O SPO ZIOLWIASAVIZED GN l e6Si Spy z BOZI ACOUFAJDETAT ON 4O3M0ld ICT AVOUASOETAI ON 1O3XO0ld gt ES 75 6999 L Aesirazoerea NOR dT E PZA IZA LA MPOZALAZ
2. 35 36 The first step in testing of the Stellaris EVB consisted of becoming familiar with the technical documentation and user manual After this was achieved the board s display options and menu were understood The team then generated a 100 Hz 1 Vpp sine wave to test the accuracy of the analog to digital ADC and LCD display used for the board An Agilent 333250A function generator setup in high z high impedance mode was the source of the test signal This reference signal allowed the group to confirm the operation and calibration of the display The appropriate next step in the testing for the Stellaris board was to connect it to display the ECG signals from the Phase 1 2 and 3 analog front end PCB s For each case the analog front end circuitry was able to correctly connect with the EVB and the ECG signals were correctly displayed on the small LCD screen as shown in Figure 29 Figure 29 Stellaris EVB oscilloscope displaying the conditioned ECG signals Testing System Gain amp Bandwidth Two main system metrics were measured and tested once the PCB s were fabricated and populated The system bandwidth and the pass band gain were both measured by the group to confirm the theoretical design with the actual boards produced The system bandwidth was an important measurable feature that directly impacted the ECG signal integrity present at the output of the analog front end circuitry The ECG analog filtering used in the circ
3. efficiency Although the efficiency for the integrated switcher was much higher the noise levels and larger PCB layout additional components were decided to be unnecessary for the specific portable ECG application the team was developing for Tl Using the LDO provided a simplified layout and smaller board footprint and the battery s current draw of 74QuA still provided an adequate 783 hours 32 days of continuous board operation The power efficiency and noise measurements taken allowed the design team to make a well informed decision for the final power circuit used in the analog front end demonstration board Testing the Input Finger Sensors Pads The team was given the initial challenge to interface the CardioSim Il ECG simulator with the analog circuitry needed to output a clean signal to the Stellaris EBM oscilloscope A way that the team went above and beyond the initial scope of the design was by implementing input sensors to allow a user s ECG signals to be measured from the fingertips Two methods were experimented with and tested Figure 33 shows the breadboard circuit using the Plessey Semiconductor PS25253 ultra high impedance active sensor Figure 33 Plessey Semiconductor sensors tested on breadboard The PS25253 is an active sensor with a built in voltage gain of 10 V V This required adjustment of the gain of the INA333 circuit as well as the output filter gain to eliminate saturation of the outputs of the IC s An examp
4. 200 mv div and the time base to 200ms div Setting the default offset for channel one was not as easy solve Instead of being set through a default defined variable it is set through a data structure that is used by many functions in the code This data structure is then interpreted by a function and passed to the correct places with correct values This allows a negative value for an offset to be represented with a negative sign in front of the number Figure 21 shows the data structure and where the 2600 was placed to achieve a 2 6V offset tRendererParams 7 sRender true bDrawGraticule true bDrawTrigLevel true bDrawTrigPos true bShowCaptions false bShowMeasurements true bDrawGround DEFAULT SCALE MV DEFAULT SCALE MV ulmVPerDivision DEFAULT TIMEBASE US uluSPerDivision 2600 0 IVerticalOffsetmV 0 lHorizontalOffset DEFAULT TRIGGER LEVEL MV ITriggerLevelmV Figure 21 Data structure that holds the channel one offset Overall the group was able to successfully implement appropriate changes to the function and operation of the Stellaris EVB oscilloscope program The final display board functions more efficiently because of the changes made to the board s source code Display Stand A display stand was constructed not only for aesthetics but also for the protection of the board The demonstration board alone is somewhat prone to damage when bein
5. Example TINA TI simulation circuit of input RC filter network 11 Figure Sa TINA TI simulation for DC servo OOP CIFCUIF eee 12 Figure 5b Results of DC servo loop removing low frequency adrift of input12 Figure 5c Right leg drive circuit simulation removing 60 Hz noise 13 Figure 6 Input RC filter network schemoftc eee eee 14 Figure 7 INA333 IC diagram showing Rg left as the gain setting resistor 15 Figure 8 Schematic for system high pass servo loop integratoFr 15 Figure 9 MATLAB script written to graph the entire system bode plot 16 Figure 10 Bode plot analyzed in MATLAB from system transfer function 17 Figure 11 Phase 1 circuit initial testing on breadboard JdvOUT 18 Figure 12 Lab testing showing breadboard AFE and the cardio simulator19 Figure 13a Phase 1 schematic of AFE and TPS62120 ICU eee 20 Figure 13b Phase TPC BAQV OUT E 21 Figure 14 Phase 2 schematic developed and fabricated in PCB Artist 22 Figure 15 Phase 2 PCB layout including copper finger pads 23 Figure 16 Phase 3 schematic with switches developed in PCB Artist 24 Figure 17 Phase 3 with switches PCB IQYOUt eee 25 Figure 18 Time base data SIFUC TON e sa avi ko aa si kaa a pk n de ka pe poet e kk Rd ate a a n 26 Figure 19 Code to update screen more rOpidlhy 27 Figure 20 Default Ted 28 Figure 21 Data structure that holds the channel one offset sese ee 29 Figure 22 SolidWorks 3D model
6. after being set to 5V The circuit was tested and measured to confirm the lab measurements with the results and performance seen in the TINA TI simulations Figure 11 shows the breadboard test circuit used prior to phase 1 of the AFE development The sub circuits placed on the breadboard included the input RC network the INA333 instrumentation amplifier the servo loop and the active low pass output filter Figure 11 Phase 1 circuit initial testing on breadboard layout The pass band gain of the analog circuitry as well as the bandwidth was measured and the signal integrity of the output was observed A bandwidth from 0 7 Hz to at least 20 Hz was necessary to retain a clear and recognizable ECG output signal This bandwidth confirmed the theoretical analysis and research of the required bandwidth for the ECG AFE circuit The gain was also adjusted using the adjustable RG resistor on the INA333 circuit and by adjusting the pass band gain of the active output LP filter The ECG signal was biased halfway between the 5V power rail and ground to allow for maximum differential swing of the ECG signal To accomplish this on the breadboard a voltage divider made up of two 560k resistors and two 0 01UF capacitors was used to divide the supply rail to a 2 5 V reference voltage that was used appropriately throughout the circuit to bias the ECG signal correctly After a clear output signal was found the output signal integrity was then observed w
7. and offset Slight movements in the electrode finger connection were researched and found to create artifacts in the ECG signal The inputs and outputs were also grouped together 100 mil spacing to allow Molex connectors and custom cables to connect the CardioSim Il and Stellaris board with the AFE PCB Figure 17 shows the phase 3 board with switches PCB layout Figure 17 Phase 3 with switches PCB layout Stellaris EVB Code Modification For displaying the conditioned ECG signal a Stellaris LM3S3748 Evaluation Board was used while running a 2 channel oscilloscope program Although the default oscilloscope program is robust and reliable it did not meet all the needs of the specific ECG application Modifications were made in the source code to better suit the needs of the project In order to make the modifications the source code that make up the oscilloscope application were studied and better understood After the basic operation of the programming and structures were laid out the first objective was to extend the time base of the oscilloscope program The time base is the amount of time per division s div on the LCD screen The maximum setting available with the default oscilloscope program was 50ms After testing this was found to only allow one period a typical ECG waveform To extend the time base and allow several periods of ECG signal to be displayed better visual results the menu controls c file had 25 26 to be
8. at what angle After the calculations were completed Nate along with Mike and Chaoli machined the display This proved to be an interesting experience in that it did not associate with electrical engineering Chaoli Ang During the project Chaoli had been working on different technical processes During phase 1 and 2 Chaoli simulated the filter which was later implemented in power management circuitry to reduce the high frequency noise from the DC power supply During the design process several attempts had been made to tune the cutoff frequency to get as close as possible to the desired value of 36Hz while demanding a relatively high gain By trying to have an accurate high frequency performance Chaoli and Justin picked the OPA333 amplifier as a major component of the circuit By using the simulation software Filter Lab the basic schematic of the filter circuitry was generated The simulation by TINA Tl gave out an unexpected result with large variance Changing the value of capacitive and resistive components of the circuit did not worked out well for the goal of design Chaoli and Mike buit up the simulation circuit to examine its AC characteristics of it Problem was narrowed down to the functional frequency bandwidth of the operational amplifier Chaoli and Mike replaced the OPA333 with LM741 to solve the problem and finalize the design To test the circuit Chaoli built up simulation circuits with the parts provided by Texas Inst
9. but represents the cost required to produce the demonstration board from scratch Some of the parts listed below were provided to the group for the design project The itemized cost for the entire solution is listed in Table 3 below A majority of the cost the team paid was in the fabrication of the PCB and acrylic display Table 3 Cost summary for the ECG demonstration board Item Cost Acrylic Display 30 Battery Pack for Stellaris EVB 30 Stellaris Evaluation Kit 120 Analog Front End PCB 33 PCB Passive Components 25 PCB Integrated Circuits 5 Accessory Components 15 Total 258 The team was given a budget of 500 to develop the project throughout the semester The team was able to develop the analog circuitry four PCB s and the final solution without exceeding this budget The ECE shop and technical engineering support at MSU helped with providing some of the smaller components used during the testing and development Project Timeline and Schedule Throughout the semester the team was able to research design fabricate and test a functioning ECG demonstration board The following Figure 38 details the project timeline that the team followed Completed Tasks as of Wed 4 24 13 Team3Project Task Mode Task Name Auto Schedul Meet with Pete Semig Auto Schedul Meet with Faculty Advisor Auto Schedul Consolidate Schedules an googie Auto Schedul Research SMD sizes Auto Schedul
10. rapidly Another implementation the group made with the oscilloscope application code was by changing the default settings For the ECG demonstration an offset of 2 5 volts was needed along with turning off channel 2 and turning off channel 1 voltage metrics that cluttered the screen The default settings were changed because the manual selection of the settings became a hassle with the small joystick on the board It was also performed to eliminate the necessary process of adjusting the display settings every time the board turned off or reset These issues were fixed by setting the default values in the code to the optimal settings for the ECG demonstration First in Figure 20 to turn of channel 2 the second value 27 28 was changed to false indicating the second channel should be turned off tBoolean og pbActiveChannels 2 Crue Ey false Figure 20 Default active channels The default scale for the oscilloscope was 1V division with 10 divisions which is too large for the ECG signal with a magnitude of around 2 5 Vpp The default time base was also set to 100 us another setting that would have to be changed on startup Both of these values were set with default variables defined in a header file and therefore to change the default values all that was needed was to change the defined value The variables DEFAULT SCALE MV and DEFAULT_TIMEBASE_US were set to 200 and 200000 respectively This sets the scale to
11. to assist in fabricating acrylic display 30 Figure 23 Final solution mounted securely on the custom acrylic display 31 Figure 24 CardioSim Il ECG simulator provided by Texas Instruments 32 Figure 25 Figure 26 Figure 27 Figure 28 Instrumentation amplifier circuit used to test SIMUIQTOF 33 Breadboard testing setup used to measure the CardioSim lI 33 ECG signal measured at test circuit output 16 Hz bandwidth 34 Stellaris EVB oscilloscope displays ECG signals for system Figure 29 Stellaris EVB oscilloscope displaying the filtered ECG signals 36 Figure 30 Bode plot of phase 1 system bandwidth 0 7 15 Hz Bandwidth 38 Figure 31 Schematic used to measure the input and output current 41 Figure 32 Jumpers shunt current measurements using Fluke 8840A 41 Figure 33 Plessey Semiconductor sensors tested on breadboard 43 Figure 34 ECG signal measured at the fingertips using Plessey sensors 43 Figure 35 Successful results of live ECG measurements for Phase 2 44 Figure 36 Phase 3 demo board displaying live ECG measurements 45 Figure 37 Stellaris displaying the ECG signal using the cardiac simulator 46 FiQUf 38 Project Timeline eee zona s ae ao pe a es emcee code a DR 47 Figure 39 Final ECG demonstration board developed by Team 3 48 Figure 40 Phase Schematic eee ik ka po a a a l Ban ak n na 58 Eiere 41 Phase 2 SONE MONO e
12. well as a DPDT switch to allow appropriate selection between the two circuits Other features included in the second board were test points and shunted jumpers to allow for input and output current measurement This made for easy measurements in the power efficiency for both circuits The testing schematic used to measure the input and output current is shown in Figure 31 Figure 32 shows the jumpers used to shunt the currents through a multi meter connected in series A Fluke 8840A digital multi meter was connected in series across the jumpers for each measurement of input and output current The input voltage output voltage input current and output current measurements were taken for the TPS7A4201 LDO and the TPS62120 integrated switch converter The results of these measurements are shown in Table 2 Power Circuit Under Test Figure 32 Jumpers allow shunted current measurements with Fluke 8840A 41 42 Table 2 Results from measurements of power efficiency for both solutions TPS7A4201 TPS62120 Input Voltage V 9 5 9 5 Input Current mA 0 74 0 43 Input Power mW 7 03 4 08 Output Voltage V 4 93 5 12 Output Current 0 67 071 mA Output Power W 3 30 3 63 Efficiency 47 00 88 98 As shown in Table 2 the results of the measurements showed that the TPS62120 operates at 88 98 efficiency matches with datasheet specifications for range of operation and the TPS7A4201 operates at 47
13. Electrocardiogram Demonstration Board Sponsored By Texas Instruments Precision Analog lA TEXAS INSTRUMENTS ECE 480 Senior Capstone Design Team 3 Spring 2013 Michigan State University Faculty Adviser Dr Rama Mukkamala Electrical amp Computer Engineering Texas Instruments Contact Peter Semig Applications Engineer Design Team Members Mike Mock Justin Bohr Nate Kesto Chaoli Ang Xie He Yuan Mei Abstract The team s design challenge for the semester was to research design fabricate and test the analog circuitry needed to develop an electrocardiogram ECG demonstration board for Texas Instruments TI The precision analog group at Tl soonsored the development of the project at Michigan State University The group sponsored the project due to their need for another tool to showcase several precision analog components in a biomedical application Many TI technologies and components are featured using demonstration boards and are put on display at technical trade shows and other professional events The defined objective for the design team was to develop the battery powered analog circuitry needed to interface a Stellaris microcontroller based oscilloscope with an ECG simulator CardioSim Il The scope of the project work included the designing of the analog front end circuitry needed to condition an ECG signal produced by a cardio simulator The semester s work also included choosing the approp
14. Go ower documents from Pete Auto Schedul BK amr wouw ni modify code for stellaris to have refire filter values according tot Fabricate board Solder parts Test board both power solution correct issues from last layout di Toe 3 19 13 san san san sann saco saco 0 00 san sao sans san san san san saco san san 0 00 san san sav san san sann san saoo 0 00 san sann san saco san san BURRRERE Page 1 Completed Tasks as of Wed 4 24 13 Team3Project Task Mode Task Name Duration Start decide on power solution idy Fa 3 22 13 integrate sensors into design 1 day Pn 3 22 13 layout final prototype in pcd ett 4 days Mon 3 25 13 order parts 1 day manye fabricate phase 3 board 13 wis Pm 3 29 13 Pn 3 29 13 Mon 4 8 13 Tue 4 9 13 The 4 14 13 Tue 4 16 13 Figure 38 Project Timeline Project Summary Throughout the semester the group was challenged to design and fabricate a portable ECG demonstration board for the team s sponsor Texas Instruments The team was given 500 to develop the project and was able to finish the project on time without going over budget The team was able to successfully develop a working ECG demonstration board that meets and exceeds the specified project requirements The defined project requirements were to develop the battery powered analog front end circuitry needed to interface a Stellaris microcontroller based oscilloscope with an ECG simulator CardioSim Il The ov
15. XAIZED ON 1 910r Str CT n Weg LWPOLYZAZX9 IZED EE 61D 81DZIDZID IZI ZZIFIDSLIFIMLI ZID II WWSLIMESEWZAZXSIZED GN 1 9Srr Spy Z eech VWSLLNEOLFZHYZXSIZED GN l l zz Spy e 29970 WWSLLJESLIZAZXIIZED AN L PLOPSP 1 d y O ALOLFABO Tea ON 19300ld tL a Sat Papers soi HUN Gate TOqUINN IRR Jeng meet n EE EECHER ERR REES 57 58 Figure 40 Phase 1 Schematic 3989 pce a a dan tas 43 an A ep DH ast ann A8 T z men P Bt none een del DI e F SEksazevan ae F sg K eva indig Jl om d T s ta L pa DEIR ast a GI 1 i id n a I n Ke 7 EH Mossagavagiredl THAAD aco e O LO dE ROU LN GA Figure 41 Phase 2 Schematic 59 Ligen L L lan Jk ZI t Fhe as SCH G 6ZI Z D 25 k t JH D een ta ik en KO amp toads Gf na n ps ans ano T z T Auen aaen BZE BLZEYaD aaen Ta te astvdd 1074 zta z N si ano g ase Zl n oug 12 AST aadal Tr L 13 HIG ASt 119 Be i SE e CH GC Hi S 381 L A 192 znd ang Jstw dOBWTE EYNI YNI anp z e z ci j lane E T DH 1 i asy JJ AS ouc AS k z aaen BZE BLZEYJQ DEL Di e H e ON Lem DE Fa snooty asa au CH E T ang 0 8 G su T gt BEZ fe nu TI an R HN lad m o 8 Di Ge vn mzz nv a a EI EIH HEI BOY ABHDDL INNGI za za WEZ DSS on S Z Q EI D 9
16. a IC Phase 3 Schemati 42 Figure 60 Project Gantt Chart Meet with Pete Semig Meet with Faculty Advisor Go over documents from Pete Consolidate Schedules on google calendar Discuss questions with Pete about Project Plan goals Get Schematics from Pete Research SMD sizes Learn PCB Artist Phase 1 Decipher analog designs Layout 1 board Fabricate the boards Test boards using o scope and func gen Order parts Solder parts Preproposal Critical Critical Split III Critical Progress m LEAR Split DIHITT Task Progress Manual Task Start only Task Name Wed 1 16 13 Tue 1 22 13 Wed 1 16 13 Wed 1 16 13 Wed 1 16 13 Wed 1 16 13 Thu 1 17 13 Fri 1 18 13 Wed 1 16 13 Wed 1 16 13 Mon 1 21 13 Mon 1 21 13 Tue 1 22 13 Tue 1 22 13 Fri 1 18 13 Fri 1 18 13 Wed 1 16 13 Wed 1 16 13 Fri 1 18 13 Tue 1 22 13 Mon 1 21 13 Tue 2 12 13 Mon 1 21 13 Tue 1 22 13 Wed 1 23 13 Fri 2 1 13 Mon 2 4 13 Fri 2 8 13 Tue 2 12 13 Tue 2 12 13 Thu 1 31 13 Fri 2 1 13 Mon 2 11 13 Mon 2 11 13 Tue 1 29 13 Thu 1 31 13 Fri 2 15 13 Mon 3 18 13 Baseline Split noman Baseline Milestone Milestone Summary Progress soss Summary F Manual Summary G Project Summary F External Tasks External Milestone Z Inactive Task Lem Inactive Milestone Inactive Summary rv Deadline Jan 20 13 w LS Ir Correct issues from last layout design Design ECG analog front end with as many circuits as pos
17. a Glelarcs a O O a a O a OF 0 zap O O O A A a a eee o ofe e o o on MN nn n eja a o eje e Gluten o a o o o e o Dik o a o a o e a ll a O A O O a o a Strong Relationship Moderate Relationship Weak Relationship Strong Positive Correlation Positive Correlation Negative Correlation Strong Negative Correlation Objective ls To Minimize Objective ls To Maximize x K aal LT K O o Objective ls To Hit Target Figure 2 House of Quality matrix used to analyze customer requirements Project FAST Diagram Another tool utilized by the group to analyze the most important functions of the overall ECG demonstration board is called a FAST Diagram It was developed to understand the functional uses that the demonstration board would provide TI or any user of the board Figure 3 shows the results of this elementary way of focusing in on the products main functionality The group analyzed the main function of the demonstration board to be analyzing an ECG signal Sub functions of the demonstration board included amplifying analog signals and converting analog to digital and other supporting operations CS Amplify Analog Signal Hi Convert Analog to Digital laa gt Bias Analog Signal Filter Analog Signal d Store Digital Signal Data Display Signal Waveform Analyze ECG Signal gt Display Signal Metrics implement Algorit
18. altered to include new menu selections The file has a data structure called g psTimebaseChoices which holds all the options available for calculating the time base This is used by the other functions such as the renderer function that actually updates the waveforms to the screen This data structure contains entries with names that would show in the menu on the left and the value of the menu item in microseconds for each division of the screen on the right Five entries were added for 100ms 200ms 500ms and 1s This allowed the oscilloscope to be set in a time base to show multiple QRS complex waveforms throughout one screen update Figure 18 shows the data structure used tControlChoice g psTimebaseChoices 20s 2 Sus 5 LOUSY L Fy Oueren SE E e S50uS 50 E 100uS 100 kinnt 250 ki e 5008S S00 Ims 1000 L 2 5mS 2500 5ms 5000 lOms 10000 C 25mS 25000 50mS 50000 100mS 100000 200mS 200000 500mS 500000 1S 1000000 ki Figure 18 Time base data structure When selecting the time base the oscilloscope application uses that value for the basis of most of its operations including how long it takes to update the screen It was analyzed that the larger the time base the longer it takes for the screen to update When setting the time base to 200ms since there are 12
19. ble for a user to read the oscilloscope After the display was modeled in SolidWorks a few modifications were needed It was not feasible to fit the angled acrylic into standoffs Instead countersunk holes were drilled and 10 32 and 8 32 screws were used to hold the acrylic in place The final display board features the Stellaris EVB secured using Velcro as well as a battery pack to power the Stellaris The final display can be seen in Figure 23 and provides an aesthetically pleasing yet safe method of demonstrating the ECG demonstration board Figure 23 Final solution mounted securely on the custom acrylic display Chapter 4 Project Testing During the design planning and development of the ECG demonstration board for Texas Instruments TI three major design iterations were performed The following section will detail the work and steps taken to test and verify each design before re work and improvements were made The major learning from each phase of the project designs will also be covered to clearly communicate the group s process of learning from each printed circuit board s PCB failures and successes At the beginning of the design process measurements were taken from the CardioSim II ECG simulator to understand the signal amplitude and noise levels present at the differential output from the simulator Once the group began to design and print the circuit boards metrics were taken and observed for each phase of the design
20. cessary components in PCB Artist using the data sheets of the circuit components to be used including the INA333 OPA378 TPS7A2401 and TPS62120 After the Phase 1 AFE was populated Justin assisted in testing and troubleshooting the boards functionality A wiring error was identified and then fixed for the Phase 2 layout He then collaborated with Mike to add features to the schematic such as two power circuits that could be switched between to test efficiency thumb pads to test the feasibility of using these pads to acquire a live ECG signal a battery connector and a switch to turn the battery on or off A switch was also added to switch the RLD on or off Justin then made the changes in PCB Artist to the schematic and layout creating phase 2 of the AFE After the phase 2 AFE board was populated Justin aided in testing and troubleshooting the circuit He then removed all testing components and organized the phase 3 PCB While developing the Phase 3 board Justin followed Peter Semig s suggestion to spread out the circuit and showcase each part of the circuit for ease in demonstration Finally Justin became familiar with the Stellaris Oscilloscope application code to make modifications so that the oscilloscope would update more rapidly and start with the optimal settings for viewing ECG signals 51 52 Nate Kesto Throughout the semester Nate assisted in many areas of the various phases of the project including practice with PCB Arti
21. correct issues from last layout design decide on power solution test plessy sensors filters gain integrate sensors into design refine filter values according to further filter tests Fri 2 15 13 Mon 2 18 13 Mon 2 18 13 Thu 2 21 13 Thu 2 21 13 Tue 2 26 13 Tue 2 26 13 Tue 3 5 13 Thu 2 21 13 Mon 2 25 13 Tue 3 5 13 Wed 3 13 13 Thu 3 14 13 Thu 3 14 13 Fri 3 15 13 Mon 3 18 13 Fri 2 1 13 Fri 2 8 13 Mon 2 25 13 Thu 4 18 13 Tue 3 19 13 Thu 3 21 13 Fri 3 22 13 Fri 3 22 13 Mon 2 25 13 Fri 3 1 13 Fri 3 22 13 Fri 3 22 13 Fri 3 1 13 Mon 3 4 13 Critical Critical Split rnn pnp pn Critical Progress mH TH HTH H H Mar 3 13 Finish only Duration only Baseline Baseline Split nn manman Baseline Milestone Milestone Summary Progress sxxx Manual Summary Q Project Summary reem External Tasks n External Milestone Inactive Task oo Inactive Milestone Inactive Summary OU Deadline Mar 31 13 F T Critical Critical Split vente Critical Progress Task Split TO Finish only Duration only Baseline Baseline Split tH TRD anan anan Baseline Milestone Milestone L Summary Progress Summary vy Manual Summary Qee Project Summary Oy External Tasks D External Milestone Inactive Task Inactive Milestone Inactive Summary GQ 9 Deadline Mar 31 13 h Manual Summary Weg mmmn Duration erh Project Summary ey Baseline qu
22. d Cl 0 01UF as shown above in Figure 6 This provides an input low pass filter with a corner frequency at 32 3 Hz 1st order LP Filter Transfer Function 2 SRC 1 Where the time constant RC is equal to 4 92ms The INA333 provides a differential gain that can be thought of as constant vs frequency A resistor placed between pins 1 and 8 for the IC provides the adjustable gain for the INA333 The equation for the gain of the amplifier is found in Equation 3 The circuit block for the INA333 is shown in Figure 7 and was taken from the datasheet developed by Tl INA333 gain 1 3 Rg RFI Filtered Inputs RFI Filtered Inputs 50kQ A V W 50kQ AAA VVV RFI Filtered Inputs RFI Filtered Inputs Figure 7 INA333 IC diagram showing Rg left as the gain setting resistor The servo loop schematic can be seen in Figure 8 It effectively creates a high pass filter in the system by inverting and feeding back a low pass signal into the INA333 s reference pin 5 This effect is summarized by the transfer function for a first order high pass filter shown in Equation 4 Servo Loop Feedback OPA378 lt 5 cy RI INA Ref NAN r1NA333 Out 0 22u 1Meg Figure 8 Schematic for system high pass servo loop integrator SRC SRC 1 1st Order High Pass Filter Transfer Function 4 The output filtering is also can be analyzed using a Ist order low pass
23. divisions on the screen the application will take 2 4 seconds between each update and when it is updated the entire screen would refresh at once When viewing an ECG signal it is best to have the signal displaying in real time instead of a static screen that updates on large intervals of time every 2 4 sec Changing this required calling the UpdateWaveform function more often than the original application was defaulted to call the function The UpdateWaveform is a function that refreshes the LCD screen with the new data digitally sampled from the analog inputs In the defaulted programming this function was being called only when the entire 12 divisions worth of data was collected It was then programmed to wait the entire 2 4 seconds time base at 200ms To increase the rate at which UpdateWaveform was called an IF statement was created in the infinite while loop the application used to run continuously This IF statement is conditional on whether or not a variable called g_ulSysTickCounter is at a multiple of ten The variable g_ulSysTickCounter is similar to a timer that continuously counts up while the application runs and updating the screen it every time g_ulSysTickCounter changes would be unnecessary The resulting code to update the screen and display the signal in real time is shown in Figure 19 below if g_ulSysTickCounter 10 0 UpdateWaveform g bMenuShown g bShowingHelp true Figure 19 Code to update screen more
24. e TPS7A4201 linear dropout regulator LDO as well as the OPA378 operational amplifiers Using the OPA378 op amps in conjunction with best practice PCB layout techniques and appropriate bandwidth selection the circuit s output contained very little noise when subsequent testing was performed Other design considerations for cost optimization were also included while making design decisions for the AFE boards The main weight of the project cost included the 33 00 charge for PCB fabrication Chapter 2 House of Quality Matrix To further help the team breakdown and analyze TI s demands relating to the project solution a Six Sigma based tool called a house of quality matrix was constructed and populated Shown in Figure 2 the tool shows the correlation of each detailed customer Tl demand as well as the capability of the potential design solutions for this project In this diagram items listed in the rows are the customer s requirements while the columns are populated with the design function requirements The team then defined the correlation between them using three levels strong moderate and weak The importance and weight of each customer requirement was then addressed accordingly and prioritized 7 f SOU SE Fee DDUH inimization of PCB Size lid Assembly C Power Supply ow Noise d Precise Filtering ow Power Consumption Ease of Sensor s Application g 8 0 2 o H 2 C a a a a E d D D Ku 5 5
25. e team built two power circuits buck Converter and low dropout regulator LDO and compared their performance in efficiency battery life and noise The TPS7A4201 LDO was chosen as the final power management IC used in the design The comparison as well as the rationale used in the decision will be documented in the body of this report Overall the design team was able to develop a solution that not only interfaced the analog front end AFE circuit with the cardio simulator but also allows for a user to measure live ECG signals present at a user s fingertips This publication has been written for the purpose in documenting the detailed steps taken throughout the design of the ECG demonstration board for the precision analog group at Texas Instruments Project Solution Throughout the semester the senior design team met with Peter Semig to discuss the direction and progress of the project Initial meetings consisted of the team asking many questions to become familiar with the desired deliverables for the demonstration board as well as the final performance desired from TI It was discussed that because TI will use the demonstration board to showcase their integrated circuit performance it was crucial that the demonstration board was reliable and measured ECG signals accurately and clearly In order for the overall system to work well the signal integrity needed to be very high as well which required using low noise Components such as th
26. eee 8 Project FAST BDIGON E 10 E lele sea ve nda stance it ya essay E ca eens eases 11 lee BI SIMIC ORS tege 11 Transfer Function YANN EE 13 Breadboarding the ee GE 17 Phase TP GB ei EE 19 Phase an WOM ninine a a e aa 22 Phase PEB VY aln naa ot det aa ton a Cede ee esl ha ddr pa a 23 Stellaris EVB Code MOUNE NOU Ee 25 Display Stand EE 29 E lae T EE 31 OEE E 31 Testing the CardiosSim HE 32 Testing the Stellaris Oscilloscope EVB 35 Testing System Gain amp Bandwidth see 36 Power Management Solution Testing sse sees ee eee eee 40 Testing the Input Finger SENSOrs PAUS eee eee 42 Be al i EE 46 FIRGI PROJSCT COST hicieron vekse ve ke vi oi anko a a See e e n e meet A ei 46 Project Timeline and Schedule sss 47 PROS Eloi wee E 48 Project COME SYON E pt vi esp pate padon pk pass popi k esse A 49 Suggested Future Developments sese eee 49 Acknowledgements sss in l tait ak eee 49 Awards Recognition EE 49 Appendix ENEE EE 50 Team Members E 50 Appendix Ee 56 SOWO EE 56 ANNO GE 56 Powerpoint Presentations see ee 56 Appendix Ill Technical ATTOcChnments see eee eee 57 UE EE 57 PROJECT SE ML anlel 58 List of Figures Figure 1 Superposition of action potentials that produce ECG signal 6 Figure 2 House of Quality matrix used to analyze customer requirements 9 Figure 3 Project FAST Diagram used to break system into basic functions 10 Figure 4 Example reference schematic provided by Matthew Hann at TI11 Figure 5
27. egene Eege eege 59 FIQUF 425 Phase 3 3CHEM ONG niesen ec e a a an km a ob k ai aa 60 Figure 43 Semester Project GANTT chart see 63 Figure 44 Final ECG Demonstration BONN E 64 RIT ee Te TTT 65 List of Tables Table 1 Recorded data for first PCB gain vs ffeQUENCY sese 39 Table 2 Measurements of power efficiency for both solution 42 Table 3 Cost summary for the ECG demonstration Doqgrd see 46 Chapter 1 Project Background The goal for the design project was to develop a demonstration board for Texas Instruments TI precision analog group The team s project sponsor Pete Semig works in TI s precision analog group in Dallas Texas as an applications engineer The precision analog group sponsored the MSU senior design team s development of an electrocardiogram ECG demonstration board An electrocardiogram ECG is a piece of electronic medical equipment that measures and displays the electrical activity associated with the heart The measured cardiac signals are commonly used for diagnosis and understanding of patient conditions in the medical and research fields One of the challenges in developing ECG systems lies in the fact that bio potentials measured at the surface of the skin have low amplitude and are mainly low in frequency fundamental below 3 Hz with spectral content up to 200 Hz This requires precise filtering and low noise amplification Another challenge lies in the fact that the ECG bio potentials measured f
28. emonstration board Mike contributed to the project by driving results in simulating designing and verifying proper operation of the entire system for each phase of the design He also provided a majority of the recorded results documented throughout the semester Mike selected and analyzed the two power solutions TPS7A4201 vs TPS62120 to compare efficiency and noise measurements Mike also aided in the fabrication of the acrylic display stand for the demo board Throughout the semester Mike Communicated with Pete Semig sponsor and reported the results successes and failures In conclusion Mike provided technical support throughout the project in all areas of the development of the ECG demonstration board Justin Bohr Justin assisted technically in several areas of the project including PCB design code modifications breadboard testing and simulations During the initial project specification and objectives outlining he heloed understand and simulate circuit diagrams given to the team to base the project The circuits included the servo loop and the Right Leg Drive He also helped with determining their function in the overall circuit and how they should be modified to suite the specific ECG application Justin helped determine part values when testing the circuit on the breadboard Justin then transferred the completed and tested circuit from the breadboard into a schematic using PCB Artist To accomplish this Justin made the ne
29. ency Input Output Gain Hz Vpp Vpp dB 0 1 0 002 0 232 41 29 0 2 0 002 0 6 49 54 0 3 0 002 1 0 54 32 0 4 0 002 1 4 56 65 0 5 0 002 1 6 58 28 0 6 0 002 1 9 59 46 0 7 0 002 2 1 60 34 0 8 0 002 2 2 60 98 0 9 0 002 2 4 61 44 1 0 0 002 2 5 61 87 1 1 0 002 2 6 62 14 1 5 0 002 2 8 62 92 2 0 0 002 2 9 63 29 2 4 0 002 3 0 63 46 3 0 0 002 3 0 63 58 3 5 0 002 3 0 63 58 4 0 0 002 3 0 63 52 5 0 0 002 3 0 63 46 6 0 0 002 2 9 63 17 7 0 0 002 2 8 62 98 8 0 0 002 2 7 62 73 9 0 0 002 2 7 62 48 10 0 0 002 2 6 62 14 15 0 0 002 2 1 60 59 20 0 0 002 1 8 58 89 25 0 0 002 1 5 57 21 30 0 0 002 1 2 55 78 35 0 0 002 1 1 54 49 40 0 0 002 0 9 53 18 45 0 0 002 0 8 52 04 50 0 0 002 0 7 50 83 55 0 0 002 0 6 49 88 60 0 0 002 0 6 48 88 65 0 0 002 0 5 47 96 70 0 0 002 0 4 47 00 100 0 0 002 0 3 42 28 39 40 The pass band gain of the first PCB measured can also be obtained from the data in Table 1 as 63 58 dB 1510 V V The group theoretical gain of the first PCB designed was set by the gain of the INA instrumentation amplifier set by an external resistor and by the output filter gain The gain setting resistor for the INA333 on the first PCB was 6 6kQ and the pass band gain of the output filter was set to be 100 V V This yielded a total theoretical gain of 1615 V V The gain was set to this value to set the amplitude of the output signal The voltage bias 2 5 V set the ECG signal at the correct half supply reference to conservat
30. erall scope of the project included precise amplification and filtering of low amplitude and low frequency bio potentials The actualized in the design layout and fabrication of the analog circuitry needed to do this The group researched the ECG application and went on to successfully design fabricate and test four iterations of the analog front end PCB s The major results found during this iterative design process helped the team improve the design throughout the semester and ultimately helped the project to obtain quality ECG measurements Due to the success of the project the precision analog group at TI plans to use the team s demonstration board to showcase the instrumentation amplifier INA333 and op amps OPA378 to customers at technical trade shows The team was able to take the specifications and list of requested deliverables provided by TI development from a theoretical concept to a reliable working product The major success in the group s results was in implementing the circuitry and hardware needed for the board to take live ECG measurements from a user s fingertips Figure 39 shows the final solution the group developed for the TI precision analog group Figure 39 Final ECG demonstration board developed by Team 3 Project Conclusion The design team was able to successfully develop an ECG demonstration board for Texas Instruments The requested functionality of the board was to interface the CardioSim Il simulato
31. es External Tasks n Baseline Split Vu mmm mum External Milestone Z mmm Baseline Milestone bi Inactive Task SS Milestone Inactive Milestone Summary Progress ess Inactive Summary U Deadline Mar 3 13 Apr 14 13 F IT l T Critical Finish only Manual Summary Qm Critical Split monona Duration oh m i Project Summary F Critical Progress Baseline en External Tasks n Baseline Split vm mmm manm External Milestone Z Split TI HHT Baseline Milestone Inactive Task Keng Task Progress Milestone Inactive Milestone Manual Task Summary Progress ess Inactive Summary rv Start only Summary ny Deadline Figure 43 Semester Project GANTT chart 63 Figure 44 Final ECG Demonstration Board 64 ui Elechrocoardiogroph 48 Texas 1 Demonstration Board INSTRUMENTS Phase 2 System Block Diagram Eb En Summary of Resulta The final PCB combined with the acrylic display KOR and exceeds the project requirements The final demonstration board successfully Miters and amplifies an input ECG signal from the CardicSim H simulator and can display the signal needed to capture a five ECG signal from a user s fingers tips Texas instruments plans to use the beard to showcase the perfcemance of their precision analog components Output from AFE Clreuk Live Output trom AFE Circalt Simulator Faculty Advisor Project Sponsor Or Ramakrishna Mukkamata Peter Semig Mike Mock Xie He Yuan Mei Texns instru
32. etween the two inputs from being amplified by the INA333 instrumentation amplifier The signal was placed at 2 5 V to allow maximum signal swing between the 5V power supply rail and the ground reference The output amplitude peak to peak was measured using an Agilent 54833A digital storage oscilloscope and the gain dB was calculated and recorded The input signal frequency was swept from 100 mHz to 100 Hz and the gain was recorded at intervals along the sweep Figure 30 shows the bode plot logarithmic x axis of the gain dB vs the frequency swept and Table 1 shows the data recorded during the experiment and As shown in Figure 30 the 3dB corner frequencies were measured to be 0 7 Hz and 15 Hz This was a success in confirming the theoretical system design with the actual results of the first board measured 37 38 65 00 7 Bode Plot of System Bandwidth 60 00 55 00 Gain dB 50 00 45 00 40 00 0 01 0 1 1 Frequency Hz 10 100 1000 Figure 30 Bode plot of phase 1 system bandwidth 0 7 15 Hz Bandwidth Table 1 Recorded data for first PCB gain vs frequency Frequ
33. filter s transfer function It has the same form as the transfer function in the inout RC network s analysis The difference between the input filtering RC network and the output filtering is the fact that one is passive and the other is active The output filtering has a pass band gain greater than unity instead of passive circuit topology used in the input section Using this analysis and process of piecing together the transfer functions of each sub circuit the total system transfer function was analyzed using algebraic manipulation and control theory The resulting total symbolic system transfer function theoretical is shown below in Equation 5 System Transfer Function 16 15 C3 R3 R4 s s 3 C1 C2 C3 RI R2 R3 R4 s 2 C1 C3 R1 R2 R4 C2 C3 R2 R3 R4 C1 C2 R1 R2 R3 s C3 R2 R4 C1 R1 R2 R2 R3 C2 R2 5 The resulting transfer function was analyzed using MATLAB and theoretical component values were plugged into the script The script used to process the transfer function symbolically is shown in Figure 9 The theoretical bandwidth expected for the MATLAB analysis was from 0 15 Hz to 27 Hz cle clear Cl 0 05E 6 RI 56E3 INAgain 16 15 C3 1E 6 R4 1E6 R2 1E3 R3 100E3 C2 0 047E 6 num 1 INAgain C3 R3 R4 0 denom C1 C2 C3 R1 R2 R3 R4 C1 C3 R1I R2 R4 C2 C3 R2 R3 R4 C1 C2 RI R2 R3B C3 R2 R4 C1 R1 R2 C2 R2 R3 R2 sys tf n
34. g moved around To prevent damage to the board an acrylic stand was designed and fabricated to keep the board stationary while also keeping it comfortable for the user to use and see their results The final design consisted of two pieces of acrylic along with eight standoffs allowing for a comfortable yet safe design for the ECG demonstration board The team chose to use standoffs with lengths of 1 25 and 3 This allowed for a comfortable viewing angle of the top acrylic sheet to be at approximately 45 degrees for the Stellaris microcontroller The display was modeled in SolidWorks and the dimensions were also calculated by hand before fabrication Figure 22 shows the SolidWorks drawing developed to mock up the team s design 29 30 Figure 22 SolidWorks 3D model to assist in fabricating acrylic display The first four standoffs are positioned to fit the PCB through holes and are located 3 75 from the bottom of an 8 x10 acrylic piece All four are 1 25 in length to ensure that the PCB is level and is appropriately spaced far enough to allow the user the ability to rest their hands comfortably on the display The next four standoffs were positioned towards the rear of the board The spacing distance was chosen to create a 45 degree angle for better front side viewing of the Stellaris screen The 3 standoffs were position approximately 1 88 from the 1 25 standoffs to ensure the 45 degree angle thus making it comforta
35. ght to alert the user when the circuit was being powered Adding the LED reduced the battery life of board however with the LED drawing only 1 5mA the AFE final board still has a sufficient battery life of 211 hours of continuous operation Figure 16 below shows the schematic for the phase 3 board including the additional circuitry and features Figure 16 Phase 3 schematic with switches developed in PCB Artist Keeping the same sub layout of each individual circuit element servo loop LDO output filter etc the layout of phase 3 was spread out to allow for labeling The precision analog group requested this for the purpose of allowing customers an ability to easily recognize the circuitry used in the board as well as to highlight the components featured in the circuits The thumb pads were separated to either side of the board to make it easier to grasp when the board is mounted on a display This was done to make it easier for the user to grip the AFE board and move as little as possible reduce DC drift
36. hile tuning the gain of the system to approximately 64dB which allowed the signal to swing fully positive and negative between the rails without saturating the op amps throughout the signal path Figure 12 shows the breadboard test setup using the CardioSim Il simulator to connect to the AFE The ECG signal quality is shown on the Agilent 54833A digital storage oscilloscope and the group then began designing and laying out the phase 1 PCB to include the analog circuitry tested on the breadboard Figure 12 Lab testing showing breadboard AFE and the cardio simulator Phase 1 PCB Work After the AFE circuit operation was confirmed through simulation and breadboard testing a schematic was created in Advanced Circuit s PCB Artist software for the Phase 1 PCB The purpose of creating a detailed schematic in PCB Artist allowed for the team to organize and order a professional two layer PCB Creating the schematic required the use of components from built in libraries that come with PCB Artist For some of the components custom footprints and schematic profiles needed to be developed Custom components needed a schematic symbol PCB symbol and an overall component that would link the schematic symbol pins and PCB symbol pins together This process was required for most components used that did not fit a 1206 surface mount pad layout This included the INA333 instrumentation amplifier OPA378 op amps test 20 points and TPS62120 switch conver
37. hm i Identify Signal Patterns Obtain Signal Metrics Process Digital Signals Figure 3 Project FAST Diagram used to break system into basic functions After researching the project deliverables needed to successfully complete the demonstration board the group analyzed several PowerPoint presentations provided by TI which included several reference schematics as well as some general information about ECG signals and their composition The presentations detailed some theoretical circuit topologies and reference designs for an ECG system developed and analyzed by Matthew Hann a linear applications engineer at TI Following the suggestions of Peter Semig the team spent a few weeks simulating the reference circuits using TINA TI Spice Software to verify the theoretical operation The team also built some of the reference circuits and verified live measurements with the simulation results An example of the reference schematic given to the group is shown in Figure 4 These reference designs helped the group get started in understanding theoretical ECG systems during the development of the phase 1 PCB however subsequent PCB s were improved upon and re designed after appropriate measurements and decisions were made 7 CM 1 Y E ar de all Average VCM is Inverted and Fed Tapping off center of split gain resistor feeds the following Back to RL voltage to the RL Drive Circuit Cancels 50 60Hz Ven ECG V ECG H 2 noi
38. ind The signal integrity for the output signal needed to be good to have a clear enough ECG signal to show up on the Stellaris built in 1 5 LCD screen Project Development The group went through three phases of the design process The first phase included the research and simulation of the circuitry needed to perform the ECG signal conditioning The second phase included the layout of two printed circuit boards PCB s to confirm practical circuit performance and function The third phase in the development of the board included the design and fabrication of the final PCB and demonstration board During each phase the group learned many things and necessary changes were made fo steer the strategy and direction of the project As requested by the team s sponsor the group designed the ECG demonstration board using the INA333 instrumentation amplifier and the OPA378 operational amplifiers The INA333 was chosen by the team to provide high common mode rejection of noise present in ECG applications The OPA378 was researched and chosen to fit the necessary SOT23 5 footprint and was desirable due to its low noise and low voltage offset operation Another feature of the OPA378 is its low power operation which includes quiescent currents under 150UA To provide a low power voltage supply for the analog circuit another Tl power management solution was researched and chosen to regulate the 9V battery to a 5V supply rail During the semester th
39. ional features and the design for measurement the layout was larger and took a longer amount of time to complete The extra time taken was filled with creating the Custom components for the switches pads battery and through holes for mounting the board on stand offs The ground planes remained separated due to the fact that the TPS62120 integrated switch converter would still introduce noise from the power circuit Jumpers were also left in the power traces so that power efficiency could be measured for each power management solution The results of these power IC comparisons will be discussed in further sections of the report Figure 15 shows the PCB layout of phase 2 Figure 15 Phase 2 PCB layout including copper finger pads Phase 3 PCB Work The layout for the Phase 3 PCB included the final design for the project All testing components test points jumpers were removed and the TPS7A4201 LDO was chosen as the final power management solution It s small board layout and lower output noise were the driving factors that led the team to selectit for use in the final design Several switches were added so that different circuits for the SLD and filtering could be demonstrated to customers for Tl to allow for additional interaction with the board A second output filter circuit was added to allow the selection 23 24 of two system bandwidths 50 Hz and 100 Hz Another feature added to the final Phase 3 PCB was an LED indicator li
40. ively allow the appropriate signal swing The measured input signal amplitude 1 5 mVpp was used to choose the gain to provide the output amplitude a value of 2 42 Vop This allowed a cushion of approximately 1 28 V between the maximum and minimum expected peaks the output signal amplitude The actual vs theoretical produced an error of 6 5 but yielded good results at the output Power Management Solution Testing The team was given the design goal of making the final demonstration board portable and ultimately battery powered The first revision and board layout was populated with a buck regulator circuit using the Texas Instrument TPS62120 integrated switching converter This device was chosen and designed in the first board because of its flexible adjustable design and high efficiency The circuit was designed to output a 5V reference that would power the analog circuitry needed to condition the output ECG signal After populating the first PCB the reference design for the TPS62120 worked very well and provided the desired 5V DC signal It was then decided to compare the TPS462120 integrated switcher with another power management solution to make a better decision for the power circuit used on the final demonstration board The two parts chosen to be measured against each other were the TPS62120 integrated switcher and the TPS7A4201 linear dropout regulator LDO The second PCB included the power circuitry for both integrated circuits as
41. le of the ECG signals at the output of the proto board while using the Plessey Semiconductor Epic sensors is shown in Figure 34 Figure 34 ECG signal measured at the fingertips using Plessey sensors Another solution designed was to simply have copper squares designed and poured on to the surface of the board The thought process behind the design was that it would allow the user to comfortably grab and measure directly from the fingertips Instead of using a high input impedance active Plessey sensor the group decided to test the simple copper pad The copper area would provide a dry electrode on the board for the user to touch The major benefit towards using the copper poured area vs the Plessey sensors were cost and the simplicity of the 43 44 board layout Each Plessey sensor cost 5 66 each and the copper patterns placed on the surface could easily be included into any PCB order free The group ordered the second PCB and included the copper pads in the layout One of the goals for the second PCB was to test and verify the proof of concept in using the simple copper pads Figure 35 shows the copper pads and an example of a user placing their fingers for measurement as well as an example of the output signal integrity Small movements as well as a higher filter bandwidth at 30 Hz caused the noise seen with the ECG signals Figure 35 Successful results of live ECG measurements for Phase 2 The copper pads worked very we
42. ll during the phase II board testing Because of the success the copper surface pads were placed in the final Phase 3 design Pads for the two thumbs the right leg drive and a ground reference were included on the final boards The finger sensor implementation makes for an excellent feature for the ECG demonstration board because it allows for live measurements to be made from a user which improves the experience of the demo The tests and measurements performed allowed the team to effectively design layout and fabricate the final PCB used for the ECG demonstration board Figure 36 shows the final demonstration board working as intended while displaying a user s live ECG signal The testing and verification of the final PCB s performance was found to be satisfactory in meeting and exceeding the minimum deliverables for the project The signal integrity of the output signal when interfacing with the cardio simulator is shown in greater detail in Figure 37 As shown the final PCB is capable of handling both live signals from a user s fingertips as well as interfacing with the Stellaris microcontroller based oscilloscope Figure 36 Phase 3 demo board displaying live ECG measurements 45 46 Figure 37 Stellaris displaying the ECG signal using the cardiac simulator Chapter 5 Project Cost The final cost for the final demonstration board was determined to be 258 This was not the cost that the team paid to develop the board
43. lty facilitator Dr Rama Mukkamala for meeting weekly with the group and supporting the team throughout the semester as well as providing weekly feedback on the group s progress and results Awards Recognition MSU ECE 480 Senior Design Competition Spring 2013 2n9 Place Award 49 50 Appendix I Technical Contributions Team Members Mike Mock Mike assisted the technical development of the project by simulating the analog front end circuits used in the entire AFE system Taking the results of these simulations Mike developed the control theory analysis of the system and documented the results of the MATLAB analysis of the system transfer function Mike developed a technical understanding of the entire system Mike also contributed by building the breadboard analog front end circuit as well as implementing the finger sensor pads for testing on the breadboard After his work with simulations and the breadboard testing he confidently chose the component values for each PCB layout in phases 1 and 2 as well as for the final phase 3 PCB Mike was also responsible for ordering the components used for each phase of PCB design Although rare throughout the semester Mike contributed at times by aiding in the process of laying out the PCB s For PCB layouts Mike more actively participated in the decision making for layout arrangements as well as verifying the boards accuracy before ordering them For the development of the final d
44. ments Precision Ansiog Nathan Kesto CheoliAng Justin Bohr Figure 45 Final Poster 65
45. nd return currents When using the power circuit a jumper was placed over the 2 pin headers which connected the power ground plane and the rest of the circuit ground This ultimately reduced the interference from the power circuit to the rest of the circuit Figure 13 shows the PCB layout fabricated for Phase 1 including the test points separated ground planes and jumper pins Figure 13b Phase 1 PCB layout After initial tests were performed on the phase 1 PCB the output was saturated at the OV ground reference This was unexpected and after further troubleshooting the group discovered that a mistake had been made in the schematic layout The non inverting and inverting pin connections to the servo loop feedback op amp were switched around which resulting in unwanted positive feedback saturation The group was able to temporarily fix this issue by bending the input pins on the op amp up off the board and small gauge wire was soldered in to reverse the input pin wiring After rectifying the positive feedback problem the board operated properly and further testing was performed to confirm correct operation gain bandwidth etc The group learned from the wiring 21 mistake in phase 1 and the servo loop wiring error was corrected during the development of the phase 2 PCB Phase 2 PCB Work Phase 2 was created with further circuit testing in mind Switches were added to turn on off appropriate circuitry and a space for a 9V batte
46. of other teammates and assisted in testing them The populating process on PCBs is good experience for practicing soldering surface mount components After the project had been done for the enhancement of final product for design day Xie assisted in testing thumb pads the sensors used for detecting life signal and worked on digital signal processing In detail Xie was assigned to digital filtering programming and beat per minute BPM detector programming For BPM detector programming Xie looked over Fast Fourier transform FFT code on the Internet and designed o signal filtering in MATLAB which were two essential stages for successfully detecting the impulse of R waves later in envelope detecting process Lar i ma 55 56 Appendix Il References Software PCB Artist Software http www 4pcb com free pcb layout software index html TINA TI Software http www ti com tool tina ti PCB Artist Tutorial http www 4pcb com media PCBArtistTutorial pdf Datasheets OPA 378 Operational Amplifier http www ti com lit ds symlink opa378 pdf INA333 Instrumentation Amplifier http www ti com lit ds symlink ina333 pdf TPS62120 Switch Converter http www ti com lit ds symlink tps621 20 podf TPS7A4201 Linear Dropout Regulator http www ti com lit ds symlink tps5410 pdf PowerPoint Presentations Analog Fundamentals of the ECG Signal Chain Matthew Hann TI PCB Artist Quickstart Guide Peter Semig TI
47. process The major tests and verification work done following the population of each PCB was to 31 32 measure system bandwidth gain and signal integrity Specific measurements for power efficiency and noise were also taken during the second iteration of design to guide the team to choose an optimal power solution for the specific ECG application The measurements taken and observed allowed the group to effectively shape the performance and success of the final solution Testing the CardioSim II After the design project was initially given to the team the major goal of the project was to design and develop the analog circuitry required to interface a cardio simulator with a Stellaris evaluation board EVB portable oscilloscope Figure 24 shows the simulator that was provided to the group to use for the project The lack of documentation and the ambiguous black box operation of the cardiac simulator CardioSim Il inspired the group to immediately test and measure the signals present at the output of the CardioSim II simulator Due to the differential nature of ECG signals the group decided to build a 3 op amp instrumentation amplifier Figure 24 CardioSim H ECG simulator provided by Texas Instruments Using the electronics parts readily available to the group in a nearby lab an instrumentation amplifier circuit was constructed using a small breadboard three LM741 operational amplifiers and several other passive component
48. r with the portable Stellaris EVB oscilloscope The group met this requirement and was able to condition and display the simulator waveforms on the portable display The group exceeded the requested functionality by implementing a solution to allow live ECG measurements to be taken from a user s fingertips The group went through several design iterations throughout the semester and as shown was able to successfully simulate design test and fabricate the final demonstration board Suggested Future Developments Future work that could be performed To improve the functionality of the demo board includes the following e Implementing an FFT based beats minute calculation of the signal e Implementing digital filtering using the Stellaris microcontroller e Designing the analog system using higher order filters e Integrating the Stellaris display board and AFE board into one PCB Acknowledgements Special thank you to Peter Semig and Collin Wells from Texas Instruments for supporting the group and offering your expertise and accountability It was a pleasure working with Pete and Collin over the course of the semester Thank you for sponsoring the project Special thank you to Gregg Mulder for assisting the group with encouragement as well as some of the soldering for the surface mount IC s on the PCB s the group developed The team enjoyed Gregg s personality and professionalism throughout the semester Special thank you the group s facu
49. ration of the CardioSim Il is still heavily undocumented and slightly ambiguous the signal amplitude and noise levels were very close to representing the ECG signals that could easily be measured at the surface of a patient s skin This discovery helped the group design and define the analog front end circuitry with the intent of preventing the operational amplifiers from operating in their non linear regions where the output signals would experience rail to rail saturations The knowledge of the battery life of the simulator helped the team make decide on a final power management solution to use for the last two demonstration boards that were fabricated Testing the Stellaris Oscilloscope EVB The Stellaris evaluation board was provided to the design team To utilize as the output display of the ECG signal The design customer precision analog group shipped the board to the group during the first four weeks of the project development It s small platform and miniature LCD screen fits perfectly with the project goal to be small and portable The Stellaris EVB used for the project is shown in Figure 28 When paired with the cardiac simulator and the analog front end interface it allows the entire system to easily travel to trade shows and customer locations to demonstrate the capability of the integrated circuits IC s used in the application LUMINARYMICRO Te Figure 28 Stellaris EVB oscilloscope displays ECG signals for system
50. riate Tl components to fit the project needs as well as the layout and fabrication of four printed circuit boards PCB s This work was performed to reduce the inherent noise present at the output of the cardio simulator The simulator generated differential ECG signals with relatively large amounts of noise which required appropriate signal conditioning to maintain the quality of the output signal at the oscilloscope To properly condition the displayed signal the team went through three major iterations in the design of the analog front end circuitry Throughout these iterations the group simulated tested and verified the expected results of the circuit operations The design team exceeded the initial project deliverables by implementing circuitry and hardware needed to handle live ECG measurements from a user s fingertips The team also implemented the code in the oscilloscope application to display the ECG signal while scrolling in real time This was an improvement from the default oscilloscope application and is better suited for the specific application of the ECG demonstration board Table of Contents EE tee EE 2 List TRIE Saw ka ke ta l Ra l a ala e abo ab DI n a ok bi lta dala s be 4 PISTON TOU ES tris kounya kal pate k ad ann pan pen ao eki pe A a ka a nama 5 S tele G E 6 Sales e Ieai TTT 6 Project SS CIC ANONS EE 7 Project Development esse sees eee 7 nela een EE 8 UT A EE 8 House of Quality MOTS sss sese e ee eee
51. rom the heart are differential in nature as they are measured on opposite polarities of the cardiac muscles A typical QRS complex ECG waveform is a superposition of many physical actions in the body s cardiac muscles See Figure 1 The composition of these activities occurring in time relate to a variety of valuable ECG frequency spectrum across the lower frequency bands The usable spectral content in a typical ECG signal falls between 0 3 Hz and 100 Hz Because of this most commercial ECG systems are designed with that bandwidth Figure 1 Superposition of action potentials that produce ECG signal Texas Instruments asked the team to design the ECG demonstration board to replace their current ECG system solution In recent years customer interest and design of ECG products has seen noticeable growth however TI is in need of a portable solution to showcase a mixed signal ECG system application In order to provide TI with a functioning and interactive demonstration board the team was tasked with researching designing testing and fabricating several iterations of board designs Project Specifications The desired specifications and main objective given to the group was to create a reliable analog front end circuit to interface a CardioSim II cardiac signal generator simulator with a Stellaris EKS LM3S3748 evaluation board The board was specified to run from a portable battery as well designing with battery life in m
52. ruments Other than designing and simulation on filters Chaoli dedicated in testing printed circuit board to achieve the optimization of the output of the Stellaris demonstration board After the PCB board is ready Chaoli and Mike populated the board by surface mount and hand soldering When testing analog front end PCB the signal output are so noisy that no clear ECG pulse is displayed Chaoli and Mike tested sub parts of the board and found a potential problem with the right leg drive By looking back into the biomedical theory and analyzing the signal flow he correct the value of resistive load in the circuit which increases the cut off frequency of the filtering process thus eliminated the noises 53 54 Yuan Mei Throughout the semesters Yuan assisted in many aspects for this project Soonsor Pete Semig from TI provided the front to end schematic Justin Mike and Yuan draw the schematic on the TI Tina and run the transient analysis Then because the input signal was pretty small Mike and Yuan designed and built a very basic instrumentation amplifier using three 741 operational amplifiers Although it had limit due to the power connection it s a dual power supply 9 volt and 9 volt respectively the amplification yet do help team detect the small ECG signal generated by the CardioSim with gain 1000 In addition Mike and Yuan made a small sub team to build the prototype of the front to end schematic on the breadboard Af
53. ry connector was created To test the proof of concept in measuring an ECG signal from a person s fingertips large copper pads electrodes were added on the surface of the board A double pole double throw DPDT switch was added to allow the selection of two power management circuits The noise and efficiency metrics were taken and two devices were compared The two power management solutions tested were both packages in TI s power management portfolio Circuits for the TPS62120 integrated switch converter and the TPS7A4201 linear dropout regulator were developed on the phase 2 board for comparison A second switch allowed the user to select between a 2 5V reference and the RLD output to the body common mode cancelation A third SPST switch was added in the power circuit to turn on off the connection to the battery Figure 14 shows the schematic for Phase 2 Figure 14 Phase 2 schematic developed and fabricated in PCB Artist The PCB layout of the phase 2 PCB became more complicated then the phase 1 PCB due to the addition of the switches and additional components Because of the addit
54. s resistors capacitors The instrumentation amplifier topology was chosen to measure the simulator s differential output signals The exact circuit schematic that was designed is shown in Figure 25 and an image of the test setup is showcased in Figure 26 The additional 1uF capacitors were added in parallel with the 10kQ resistors to add a pole in the circuit s transfer function which would roll off high frequency spectral content above 16 Hz estimated bandwidth needed The addition of the capacitors poles dramatically improved the signal to noise S N ratio at the output of the test circuit by attenuating high frequency noise 50k lp l i R4 R7 741 4 RA_Input ES NN 10k SS 741 Output 1k R2 RS SAN 50k 1k at jer 10k T 741 H 7 LA Input Figure 25 Instrumentation amplifier circuit used to test simulator Figure 26 Breadboard testing setup used to measure the CardioSim II Analyzing the schematic topology and component values in Figure 25 results in a pass band gain defined by the analysis below The resulting pass band gain is shown in Equation 6 below p Rena Vout T 2 R1 R3 A AA x ik cena E T Rgain RP 33 34 Where R1 50kQ R3 10kA R2 1k0 and Rgain 1kA Pass band Gain Vout RA LA 1010 6 Knowing the pass band gain of the test circuit the output signal amplitude Vout needed to be found in order to calcula
55. s of the DC servo loop is shown below in Figure 5a and 5b The team was able to input a differential ECG signal using some of TINA TI s signal source defining tools Ie d Cl Agnd lt N ba Ln TAI v Feedback U1 INA333 6 THz artifact remains on output wout DC Servo Loo 7 r Figure 5b Results of DC servo loop removing low frequency drift of input Other simulations were performed for the right leg drive circuit Figure 5c below shows the right leg drive functioning to remove common mode signals 60 Hz noise from the input to the analog front end Figure 5c Right leg drive circuit simulation removing 60 Hz noise The power management IC s TPS7A4201 and TPS62120 were also simulated to confirm the proper voltage regulation and output voltages that would be expected This helped the team order appropriate parts and confidently layout the power management circuits on the PCB s The ideal regulated output voltage to run the AFE was 5V Both the buck converter and LDO were able to regulate a 5V rail from a 9V battery The layouts for the LDO and buck converter were designed using the reference schematics given in the datasheet s application notes for each device Transfer Function Analysis To understand the entire system s behavior the sub circuits used to piece together the ECG demonstration board s AFE were separated and studied This was done by the group to piece together their transfer f
56. se Vem ECG ECG 2 Figure 4 Example reference schematic provided by Matthew Hann at TI Chapter 3 Circuit Simulations After analysis of the critical customer requirements following the team s conversations with Peter Semig circuit solutions began to develop and the first step in the design process was to simulate the system sub circuits to understand their function and impact towards the total system behavior The sub circuits simulated towards the beginning of the project were the input low pass RC filter network the INA333 gain and transient performance the DC servo loop and the output filtering The integrated circuits used in each sub circuit were the low noise OPA378 as well as the INA333 instrumentation amplifier Figure 5 shows TINA TI simulation circuit for the input RC input low pass filter network used to bandwidth limit the input signals to the INA333 The function of the circuit was found to include both a common mode low pass filter as well as a differential mode low pass filter Figure 5 Example TINA TI simulation circuit of input RC filfer network Another reference circuit explored by the team was the right leg drive RLD amplifier This circuit works together with the input filter and a DC servo loop to cancel common mode noise as well as set the DC reference for the input signal to be between the power supply 5V and ground rails The simulation schematic and result
57. sible in the design Simulate circuits Layout in PCB Artist Order parts Fabricate board Solder parts Test board both power solutions and metal pads using o scope and func gen Oral Proposal presentation Phase 3 correct issues from last layout design decide on power solution test plessy sensors filters gain integrate sensors into design refine filter values according to further filter tests Fri 2 15 13 Mon 2 18 13 Mon 2 18 13 Thu 2 21 13 Thu 2 21 13 Tue 2 26 13 Tue 2 26 13 Tue 3 5 13 Thu 2 21 13 Mon 2 25 13 Tue 3 5 13 Wed 3 13 13 Thu 3 14 13 Thu 3 14 13 Fri 3 15 13 Mon 3 18 13 Fri 2 1 13 Fri 2 8 13 Mon 2 25 13 Thu 4 18 13 Tue 3 19 13 Thu 3 21 13 Fri 3 22 13 Fri 3 22 13 Mon 2 25 13 Fri 3 1 13 Fri 3 22 13 Fri 3 22 13 Fri 3 1 13 Mon 3 4 13 Critical Critical Split mmmn Critical Progress Finish only Duration only Baseline Ce Baseline Split Te Baseline Milestone Milestone KL Summary Progress Summary vy Manual Summary Weg Project Summary ee External Tasks as External Milestone Inactive Task Inactive Milestone Inactive Summary QO 9 Deadline 61 62 Correct issues from last layout design Design ECG analog front end with as many circuits as possible in the design Simulate circuits Layout in PCB Artist Order parts Fabricate board Solder parts Test board both power solutions and metal pads using o scope and func gen Oral Proposal presentation Phase 3
58. st Tina Tl and filter pro However Nate focused more so on various aspects of each phase Specifically he performed the filter calculations for the common and differential mode filtering associated in the AFE circuit using prior knowledge of RC networks from previous classes After the calculations were completed Nate simulated the filters Using Tina TI by using variations of the filter models obtained from Mr Pete Semig to assure a high integrity signal Nate also assisted in the population of each of the PCBs at the different phases of the project This proved to be a tedious process due to Nate s inexperience with surface mount components however it was a great opportunity for him to practice and become familiar with PCB design and population before going into industry This summed up his work in phase 1 After the PCBs were fabricated and populated his role in phase 2 was to assist in various testing of the different systems on the boards Specifically he and Mike determined the efficiency of the two proposed power chips using the available lab equipment and he helped decide the best option Along with the power testing for the chip Nate also contributed to the calculation of the battery life of the ECG demonstration board itself with the help of Mike and various datasheets Finally for phase 3 Nate s major contribution was the design and machining of the display He assisted with the calculations on where to place the standoffs and
59. te the differential signals present at the output of the simulator RA LA The output of the test circuit was measured using a Philips PM 3365 100 MHz analog storage oscilloscope The ECG simulated waveform in Figure 27 was measured at the output of the test circuit and the amplitude of the signal was found to be 1 48Vpp peak to peak Using this information the differential signals present at the output of the simulator were conservatively estimated to be between 1 1 5mVpp This amplitude seemed realistic and matched the amplitude ranges that were researched during the first weeks of the project The correct range for typical ECG signals measured from the skin was researched to be 0 05 3 mVpp As shown in Figure 27 and was later verified the lower bandwidth limited the spectral content needed for higher definition of certain components of the QRS complex signal PM 3365 100MHz 100MS s Figure 27 ECG signal measured at test circuit output 16 Hz bandwidth The current draw was also tested and measured for the CardioSim Il A Fluke 8840A digital multi meter was connected in series with a power supply set to 9V DC and the DC current was measured The current draw for the simulator was measured to be approximately 21 8 mA A 9V battery portably powers the simulator and each battery typically is rated for 580 mAh The simulator s estimated battery life is therefore approximately 26 continuous hours of operation Although the large scale ope
60. ter successfully build the prototype they measured and tested the output data which verify the analog design and ensure the PCB design Besides the hardware building and testing Yuan downloaded the code compiler installed the code library and worked on the add on features to Stellaris microcontroller The initial thought was to change the time base to increase the resolution of the output signal implement the Fast Fourier Transform and calculated the heart beats per minute Yuan changed the code and enable the time base large enough for use Also Yuan has a excellent skills on soldering he responsible for the 30 of the soldering part through phase 1 phase 2 and phase 3 Xie He During the designing process of the project Xie contributed in multiple aspects in various phases In phase 1 Xie co working with Nate was focusing on analyzing the input RC filtering circuit including common and differential mode filtering associated in the AEF circuit Specifically the analysis includes the derivation of transfer functions for various filters the calculation of time constants and corner frequencies and the stimulation of filtering circuits In the stimulation of filtering circuits he stimulated and compared each sub circuit contained in the overall filtering circuit in purpose of understanding how each part works separately and assuring an output signal with high resolution In phase 2 Xie populated 55 of each PCB with the assistance
61. ter IC PCB Artist has a simple wizard for creating schematic and PCB symbols for op amps and almost any IC Using the wizard shortened the time taken to create the custom components Components that could not be created with the wizards had to be hand drawn in the editor which required more precision and patience to complete Using the datasheets for each of the components and IC s accurate dimensional measurements allowed these PCB symbols to be drawn by hand and the accuracy was verified using measurement tools in PCB Artist Since the phase 1 board was developed mainly for testing purposes several test points and jumpers were placed at appropriate places in the schematic Figure 13 below shows the Phase 1 schematic used to design and layout the PCB Figure 13a Phase 1 schematic of AFE and TPS62120 circuit When laying out the PCB design for Phase 1 a high priority for the team was in minimizing interference through the power circuit s ground plane and sensitive nodes of the circuit The power circuit used was a TI integrated switcher buck converter TPA62120 which bucked a 9V battery supply to the required 5V chosen to power the board The ground plane for the power circuit was separated due to the noise generated by the internal MOSFET switching in the power IC To protect the noise from coupling into the high impedance input pins on the op amps the ground plane was also cut out from under to reduce EMI from fast switching grou
62. uitry created the total system effect of a band pass filter The three main sub components that created the poles and zeros required to roll off the frequency content outside the desired pass band were the input RC network filtering the servo loop and the output filtering The desired bandwidth of the ECG demonstration board was researched and defined to be from 0 3 100 Hz After experimenting with the bandwidth on the breadboard circuit by altering the corner frequencies of the three circuits acceptable signal integrity was found when using a bandwidth of 0 7Hz 15 Hz If was later decided that increasing the bandwidth above 50 Hz increased noise levels present on the signal but also included meaningful spectral content present in the faster occurring events such as the very recognizable R wave ECG spike After determining the desired system bandwidth the first ordered PCB was populated with the parts chosen and taken to the lab for testing The system s bandwidth was measured to compare the actual board measurements with the theory used to select the circuit components The technique for measuring the bandwidth was to use an Agilent 33250A function generator to generate a 2 mVpp input sine wave Because of the differential nature of the inputs to the board one lead RA was connected with the signal generator while the other LA was set to a DC level that matched the level of the RA signal 2 5VDC This eliminated the common mode offset b
63. um denom h bodeplot sys setoptions h FrequUnits Hz Figure 9 MATLAB script written to graph the entire system bode plot The resulting bode plot was generated from the MATLAB script s symbolic transfer function using theoretical component values The resulting plot is shown below in Figure 10 As shown the total system behavior can be summarized as a band pass filter with a bandwidth from 0 16 Hz to 27 04 Hz The pass band gain of the filter is 64 2 dB 1621 81 V V As shown the theoretical results matched very well with the total symbolic analysis of the system s transfer function Bode Diagram a D 8 2 5 kw Frequency Hz Figure 10 Bode plot analyzed in MATLAB from system transfer function Breadboarding the Circuit Prior to developing the first PCB s for the initial phase of the design project The group DUIT and tested as much of the analog front end circuitry AFE as possible using a breadboard in the labs at Michigan State University The INA333 was placed on a DIP to MSOP8 adapter to allow the surface mount IC to be placed in circuit and tested using through hole passive components In the circuit on the breadboard the OPA378 op amps were replaced with readily available LM741 op amps The goal in putting the circuit together was to test and confirm the functionality of the reference circuits for the analog front end system The circuit was powered using an HP 6216C power supply
64. unctions into a total system wide transfer function As stated previously the sub circuits were simulated using TINA TI spice software and the team was able to confirm each circuit s theoretical transfer function and its corresponding transient behavior The ultimate goal of this work was to understand each sub circuits impact on the total system and to work towards designing the final schematic The entire circuit can be broken into four main parts that contribute to the differential amplification and filtering of the system The four main sub circuits are the input RC filtering INA333 instrumentation amplifier the servo loop and the output filter stage Figure 6 shows the schematic for the input RC filtering network It is comprised of a differential mode low pass filter as well as a common mode low pass filter The noticeable contribution to the overall system is from the circuit s differential filtering so this will be included in the following analysis pou Figure 6 Input RC filter network schematic The differential corner frequency for the RC input network low pass filter can be summarized by Equation 1 below For the sake of making the analysis easier to understand the result can be summarized again into a single ended first order low pass filter transfer function as detailed in Equation 2 Input RC Network Corner Frequency an ea 1 For the later designs RI and R2 12k C3 0 1UF and C2 an

Download Pdf Manuals

Related Search

Related Contents

Starlite Garden PT-JD Instructions / Assembly  Novell Linux Desktop  1－1 仕様  Fujitsu Prestige Pro Case Midi 14  L2352 取扱説明書  diagramas - Lincoln Electric  L-INX 10X User`s Manual  Gestor de su Calendario & Reservas  

Copyright © All rights reserved. Failed to retrieve file