Development of an MRI-compatible ECG...
-
Upload
nguyenminh -
Category
Documents
-
view
242 -
download
0
Transcript of Development of an MRI-compatible ECG...
D I P L O M A T H E S I S
D E V E L O P M E N T O F A N M R I - C O M PAT I B L E E C GS Y S T E M
csaba hajdu
supervisors
Dr Dávid LégrádyBME
Béla MihalikMediso
Budapest University of Technology and EconomicsFaculty of Natural Sciences
Institute of Nuclear TechniquesMay 2012
colophon
This thesis project was executed within the framework of the MedicalPhysics specialization of the Physics MSc course at the Institute ofNuclear Techniques, Faculty of Natural Sciences, Budapest Universityof Technology and Economics.
This document was typeset using the typographical look-and-feelclassicthesis developed by André Miede. The style was inspiredby Robert Bringhurst’s seminal book on typography “The Elements ofTypographic Style”. classicthesis is available for both LATEX and LYX:
http://code.google.com/p/classicthesis/
Ö N Á L L Ó S Á G I N Y I L AT K O Z AT ( D E C L A R AT I O N O FI N D E P E N D E N T W O R K )
Kijelentem, hogy jelen szakdolgozat saját munkám eredménye, aztönállóan, nem megengedett segédeszköz igénybevétele nélkül készí-tettem el. A szakdolgozatban csak a megadott forrásokat használtamfel. Minden olyan részt, amelyet szó szerint, vagy azonos értelem-ben, de átfogalmazva más forrásból vettem, egyértelmuen, a forrásmegadásával jelöltem.
Budapest, 2012. május
Hajdu Csaba
A B S T R A C T
Cardiac gating is an important technique used in many differentmodalities of medical imaging to eliminate artifacts resulting fromthe motion of the heart. It is often implemented with the use of anelectrocardiograph (ECG).
This thesis project was executed at Mediso Medical Imaging Sys-tems. The main objective was the development of an ECG systemaimed at cardiac gating, to be used in the pre-clinical (small animal)and clinical (human) imaging devices of the company.
The ECG system is based on a microcontroller and an ECG-specificanalog frontend. Prototype hardware was created using developmentkits and the appropriate firmware was developed. Since the tests car-ried out with the prototype system confirmed the suitability of theconcept, the final hardware was designed. The process is explainedin detail.
Cardiac gating also finds use in Magnetic Resonance Imaging (MRI),but this causes challenges in ECG acquisition. Measurements werecarried out to assess the usability of the prototype system in MRI.These results are also reported.
K I V O N AT
A szívkapuzás fontos technika, amelyet számos különbözo orvosiképalkotó modalitás esetében használnak a szív mozgásából eredomutermékek kiküszöbölésére. A szívkapuzást gyakran elektrokardio-gráf (EKG) használatával valósítják meg.
Ez a diplomamunka a Mediso Orvosi Berendezés Fejleszto és Szer-viz Kft.-nél készült. A fo célja egy szívkapuzásra szolgáló rendszerkifejlesztése volt, a vállalat preklinikai (kisállat) és klinikai (humán)képalkotó készülékeivel való használatra.
Az EKG rendszer egy mikrokontrolleren és egy EKG-specifikus ana-lóg frontend áramkörön alapul. Fejlesztokészletek felhasználásávalprototípus hardvert alkottam, és kifejlesztettem az azon futó szoftvert.A prototípus rendszerrel lefolytatott tesztek alátámasztották, hogy azelképzelés megfelel a célnak, így megterveztük a végleges hardvert.Ezt a folyamatot a dolgozat részletesen leírja.
Szívkapuzást mágneses rezonancia képalkotó berendezésben (MRI-ben) is alkalmaznak, de ebben az esetben kihívást jelent az EKG mé-rése. A prototípus rendszer MRI-ben való használhatóságának vizs-gálatára méréseket hajtottam végre, amelyek leírása szintén szerepela dolgozatban.
A C K N O W L E D G M E N T S
I would like to express my gratitude to all those people who havecontributed to the completion of this project with their assistance andencouragement.
I am extremely grateful to my supervisors, Dr Dávid Légrády andBéla Mihalik for their invaluable support from the first steps of theproject to the very end.
Many thanks to Miklós Czeller for his assistance in writing this the-sis and to Attila Lantos for some truly great ideas that helped me outof tight spots. It is a pleasure to have had a chance to cooperate withSándor Török in the design of the hardware. I really appreciate thehelp of Magor Babos, Sándor Hóbor, Péter Koncz, Domokos Máthéand Zoltán Nyitrai in the organization and execution of small animalECG measurements.
I am grateful to all the colleagues in my office for an inspiringworking environment. I am also very thankful to everyone else atMediso who lended a hand during my work. Special thanks to themanagement of the company for letting me tackle this demandingproject.
Last but not least, this work can only be dedicated to my familyfor the love, support and inspiration they’ve given me all the waythrough.
v
C O N T E N T S
1 introduction 1
1.1 Electrocardiography (ECG) . . . . . . . . . . . . . . . . 1
1.1.1 Physiological background . . . . . . . . . . . . . 1
1.1.2 Recording an ECG . . . . . . . . . . . . . . . . . . 2
1.2 Cardiac gating . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 ECG in MRI . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 specifications 5
2.1 Specifications of the ECG system . . . . . . . . . . . . . 6
3 prototyping 8
3.1 Choice of architecture . . . . . . . . . . . . . . . . . . . 8
3.1.1 The ADS1298R IC . . . . . . . . . . . . . . . . . . . 9
3.2 The ECG prototype . . . . . . . . . . . . . . . . . . . . . 9
3.2.1 The ADS1258 MDK . . . . . . . . . . . . . . . . . . 9
3.2.2 The ADS1298R ECG FE-PDK . . . . . . . . . . . . . 10
3.2.3 The ECG prototype system . . . . . . . . . . . . . 11
3.3 Firmware development . . . . . . . . . . . . . . . . . . . 12
3.3.1 The firmware of the ADS1258 MDK . . . . . . . . 12
3.3.2 The firmware of the ECG prototype system . . . 14
4 hardware design 20
4.1 Choice of microprocessor . . . . . . . . . . . . . . . . . 20
4.2 Hardware design considerations . . . . . . . . . . . . . 21
4.2.1 Patient protection . . . . . . . . . . . . . . . . . . 22
4.2.2 Power supplies . . . . . . . . . . . . . . . . . . . 22
4.2.3 ECG hardware . . . . . . . . . . . . . . . . . . . . 24
4.2.4 Small animal respiratory measurement . . . . . 25
4.2.5 Microcontroller unit . . . . . . . . . . . . . . . . 26
4.2.6 Communication interfaces . . . . . . . . . . . . . 26
4.3 The hardware . . . . . . . . . . . . . . . . . . . . . . . . 27
5 measurements 28
5.1 ECG signal acquisition . . . . . . . . . . . . . . . . . . . 28
5.1.1 ECG simulator . . . . . . . . . . . . . . . . . . . . 28
5.1.2 Human ECG . . . . . . . . . . . . . . . . . . . . . 30
5.1.3 Small animal ECG . . . . . . . . . . . . . . . . . . 30
5.2 Performance in MRI . . . . . . . . . . . . . . . . . . . . 33
5.2.1 Initial measurements . . . . . . . . . . . . . . . . 34
5.2.2 Small animal measurements . . . . . . . . . . . 34
5.2.3 Discussion . . . . . . . . . . . . . . . . . . . . . . 37
6 summary 39
6.1 Accomplishments . . . . . . . . . . . . . . . . . . . . . . 39
6.2 Possibilities of improvement . . . . . . . . . . . . . . . . 39
a appendix 40
vi
L I S T O F F I G U R E S
Figure 1.1 Schematic representation of normal ECG. . . . 2
Figure 2.1 Block diagram of the ECG system. . . . . . . . . 5
Figure 3.1 Block diagram of the ADS1298R. . . . . . . . . . 9
Figure 3.2 Photo of the ADS1298R ECG FE-PDK. . . . . . . . 10
Figure 3.3 Photo of the ECG prototype system. . . . . . . . 11
Figure 3.4 IIR filter frequency response of the ADS1258 MDK. 13
Figure 3.5 FIR filter frequency response of the ADS1258 MDK. 13
Figure 3.6 Flowchart of the ECG prototype system firmware. 15
Figure 3.7 Screenshot of the PC application for ECG display. 16
Figure 3.8 Frequency response of the DC removal filter. . 17
Figure 3.9 Response of the human band-pass filter. . . . . 18
Figure 3.10 Response of the animal band-pass filter. . . . . 18
Figure 3.11 Impulse response of the human band-pass filter. 18
Figure 3.12 Frequency response of the notch filter. . . . . . 19
Figure 4.1 The architecture of the Concerto MCU family. . 21
Figure 4.2 Frequency response of the antialiasing filter. . 25
Figure 4.3 Rendered visualization of the ECG board. . . . 27
Figure 5.1 ECG simulator output. . . . . . . . . . . . . . . 29
Figure 5.2 ECG simulator output, corrupted by 50Hz noise. 29
Figure 5.3 Human ECG recording. . . . . . . . . . . . . . . 30
Figure 5.4 Human stress ECG recording. . . . . . . . . . . 31
Figure 5.5 The anesthetized rat fitted with ECG electrodes. 32
Figure 5.6 Rat ECG recording. . . . . . . . . . . . . . . . . 32
Figure 5.7 Hamster ECG recording. . . . . . . . . . . . . . 33
Figure 5.8 Background spectrum measured inside the MRI. 35
Figure 5.9 Spectrum of a hamster ECG within MRI. . . . . 36
Figure 5.10 Spectrum of a hamster ECG within MRI, zoomed. 36
Figure 5.11 Spectrum of a hamster ECG during MRI sequence. 37
L I S T O F TA B L E S
Table 2.1 Design parameters of the ECG system. . . . . . 7
vii
L I S T I N G S
Listing A.1 Configuration of the ADS1298R. . . . . . . . . . . 40
A C R O N Y M S
AD Analog-to-Digital
ADC Analog-to-Digital Converter
AFE Analog Frontend
CT Computed Tomography
CAN Controller Area Network
CMRR Common-Mode Rejection Ratio
DSP Digital Signal Processor
ECG electrocardiograph
EVM Evaluation Module
FIR Finite Impulse Response
GPIO General Purpose Input/Output
IC Integrated Circuit
IIR Infinite Impulse Response
LCD Liquid Crystal Display
MCU Microcontroller Unit
MDK Medical Development Kit
MRI Magnetic Resonance Imaging
PDK Performance Demonstration Kit
PET Positron Emission Tomography
PGA Programmable Gain Amplifier
PSRR Power Supply Rejection Ratio
viii
acronyms ix
RLD Right Leg Drive
SAR Successive Approximation Register
SPI Serial Peripheral Interface
TI Texas Instruments
UART Universal Asynchronous Receiver/Transmitter
USB Universal Serial Bus
WCT Wilson Central Terminal
1I N T R O D U C T I O N
This thesis describes the development of an electrocardiograph (ECG)system designed for cardiac gating. This chapter is aimed at high-lighting the significance of cardiac gating in pre-clinical (animal) andclinical (human) medical imaging.
Section 1.1 discusses electrocardiography based on [1]. Assuming abasic knowledge of the anatomy of the heart, the physiological back-ground of human ECG is explored, without delving into the clinicalsignificance of ECG traces. Next, the practical aspects of ECG recordingrelevant to this project are reviewed.
Section 1.2 gives an overview of cardiac gating, and Section 1.3deals with the challenges of ECG measurement in Magnetic ResonanceImaging (MRI) machines.
1.1 electrocardiography (ecg)
1.1.1 Physiological background
As the heart undergoes depolarization and repolarization during thecardiac cycle, the electrical currents that are generated spread notonly within the heart, but also throughout the body. This electricalactivity generated by the heart can be measured by an array of elec-trodes placed on the body surface. The recorded tracing is called anelectrocardiogram (ECG, or EKG). Figure 1.1 shows the schematic rep-resentation of a "typical" human ECG tracing. The different waves thatcomprise the ECG represent the sequence of depolarization and repo-larization of the atria and ventricles.
The P wave represents the wave of depolarization that spreads fromthe sinoatrial (SA) node throughout the atria, and is usually 80-100 msin duration. The brief isoelectric period after the P wave representsthe time in which the impulse is traveling within the atrioventricular(AV) node and the bundle of His.
The period of time from the onset of the P wave to the beginning ofthe QRS complex is termed the P-R interval, which normally rangesfrom 0.12 to 0.20 seconds in duration. This interval represents thetime between the onset of atrial depolarization and the onset of ven-tricular depolarization.
The QRS complex represents ventricular depolarization. Heart ratecan be calculated by determining the time interval between QRS com-plexes. The duration of the QRS complex is normally 0.06 to 0.1 sec-
1
1.1 electrocardiography (ecg) 2
Figure 1.1: Schematic representation of normal ECG. Courtesy of Wikipedia.
onds. This relatively short duration indicates that ventricular depolar-ization normally occurs very rapidly.
The isoelectric period (ST segment) following the QRS is the time atwhich the entire ventricle is depolarized.
The T wave represents ventricular repolarization and is longer induration than depolarization, since conduction of the repolarizationwave is slower than the wave of depolarization.
The Q-T interval represents the time for both ventricular depolar-ization and repolarization to occur. This interval can range from 0.2to 0.4 seconds depending upon heart rate.
There is no distinctly visible wave representing atrial repolariza-tion in the ECG because it occurs during ventricular depolarization.Because the wave of atrial repolarization is relatively small in am-plitude, it is masked by the much larger ventricular-generated QRS
complex.
1.1.2 Recording an ECG
By convention, in a 5-electrode ECG configuration, electrodes are posi-tioned on each arm and leg, and one electrode is placed on the chest.The limb electrodes are referred to as Left Arm (LA), Right Arm (RA),Left Leg (LL) and Right Leg (RL), while the chest electrode is referredto as V1. LA, RA, LL and V1 are used as input electrodes in monitor-ing the electrical activity of the heart, whereas RL is electrically drivenby the Right Leg Drive (RLD) circuitry to reduce the common-mode
1.2 cardiac gating 3
voltage on the other electrodes. In general, the RL signal is obtainedas the average of the other limb electrodes.
The conventional ECG leads are formed from the electrode signals.The ECG leads are subdivided into the following categories:
• Standard limb leads: I, II, III
• Augmented limb leads: aVL, aVR, aVF
• Chest lead: V1
The leads are formed according to the following rules:
I = LA − RA
II = LL − RA
II I = LL − LA
aVL = LA − 12· (RA + LL)
aVR = RA − 12· (LA + LL)
aVF = LL − 12· (LA + RA)
V1 = V1 −13· (LA + RA + LL)
The standard limb leads are bipolar because the lead voltage is mea-sured between two electrodes, while the other leads are unipolar be-cause they are measured with respect to the Wilson Central Termi-nal (WCT) voltage, which is defined as VW = 1
3 · (LA + RA + LL).In practice, ECG signal acquisition may be implemented with dif-
ferential amplifiers. In this case, the standard limb leads may be ob-tained directly by connecting the appropriate electrodes to the inputsof the differential amplifiers. The chest lead may be obtained by gener-ating the WCT voltage and measuring the chest electrode with respectto it. The augmented limb leads may be calculated digitally from theother leads.
In the case of small animal measurements, three electrodes are gen-erally used. The electrode configuration may be LA, RA and LL, oralternatively, LA, RA and RL to provide feedback for common-modesignal reduction.
1.2 cardiac gating
Cardiac gating is a technique used to reduce artifacts due to heartmotion in medical imaging.
The heart is in motion during the cardiac cycle. It completes thecontraction required for its pump function during the systolic phase,
1.3 ecg in mri 4
then relaxes in the diastolic phase. Regardless of the imaging modal-ity in use, it is desirable to create images of the heart while it is sta-tionary in order to minimize motion artifacts. The diastolic periodsof lesser cardiac motion are ideal candidates for artifact-free imaging,however, the temporal resolution of current imaging techniques is notsufficient for the creation of images in a single stationary period. Thisproblem can be addressed with the use of ECG-synchronized tech-niques to record and combine data from multiple stationary periodsfor motion-free imaging.
The occurrence of the QRS complex in the ECG corresponds to theperiod of greatest heart motion. Therefore, by recording QRS triggersalong with the image data, data from the periods of lesser motioncan subsequently be extracted for the reconstruction of motion-freeimages (retrospective ECG gating). Moreover, the occurrence of QRS
complexes can be predicted on the fly from the previous ECG data,which can be used in Computed Tomography (CT) imaging for re-ducing the output power of the X-ray tube during systole in order toreduce patient dose (prospective ECG triggering) [2].
Cardiac gating has been proven to be efficient in cardiac CT andcoronary angiography [2], cardiac Positron Emission Tomography(PET) [3] and even in MRI diffusion tensor imaging of the brain [4].
1.3 ecg in mri
As discussed above, ECG acquisition in MRI is clinically relevant. How-ever, it is a challenging task.
In addition to the inherent distortion of the ECG waveform in MRI
systems operating at high field strengths, the static, gradient and RF
electromagnetic fields of the MRI system can cause additional arti-facts in the ECG signal [5]. Moreover, the ECG electrode leads may actas antennas, guiding radio-frequency interference into the MRI bore,thereby generating imaging artifacts or even causing image acquisi-tion to fail.
An existing MRI-compatible ECG system for human applications em-ploys long MRI-compatible cables and filtering in order to leave theelectronics outside of the MRI chamber [6], but solutions allowing themonitor to be placed near the patient are also available [7, 8].
Solutions also exist for small animal applications [9, 10]. The doc-umentation available on-line reveals very limited detail about thesesystems, in particular with respect to the MRI architectures and fieldstrengths they are compatible with. However, both systems appear torely on MRI-compatible ECG electrodes and lead wires, and the elimi-nation of artifacts in the ECG signal by digital filtering.
2S P E C I F I C AT I O N S
The thesis project was executed at Mediso Medical Imaging Systems,Hungary. The company’s specifications for the ECG system project re-quire the development of a modular ECG system. The main moduleshould implement the acquisition of the ECG signals with QRS com-plex detection, trigger generation and transmission for cardiac gatingin pre-clinical and clinical imaging machines, along with pre-clinicalrespiratory measurement and triggering. The slave module shouldimplement a temperature control loop for small animal systems, mea-suring and regulating the temperature of the animal bed. Each deviceshould implement two channels, i.e. it should be able to operate withtwo patients (animals) simultaneously, with a shared microprocessor(see Figure 2.1). Although the possibility to visualize signal wave-forms is required, the project is not aimed at the development of adiagnostic quality ECG system.
The thesis deals with the development of the main module, andonly describes details pertaining to the ECG functionality.
Figure 2.1: Block diagram of the ECG system.
5
2.1 specifications of the ecg system 6
2.1 specifications of the ecg system
An excerpt of the original specifications by Mediso follows.
1. Input/output
• Digitization of signals from 4 ECG electrodes in parallel for7-lead ECG
• RLD bias electrode
• ECG cables with standard connectors should be usable
• Communication interfaces: IEEE 1588 Ethernet, UniversalSerial Bus (USB), Controller Area Network (CAN)
2. Electronics, signal processing
• Digitization of human and rat/mouse ECG signals
• Automatic QRS complex detection, trigger generation andtransmission
• Configurable ECG preamplifier gain
• Power line interference filter
• Switchable high- and low-pass filters
• Automatic detection and reporting of ECG electrode faults
• Protection against overvoltages from defibrillators
3. Safety
• Patient isolation: at least 4 kV from communication inter-faces and power network
• Defibrillator protection: at least 5 kV at inputs
• Compliance with the IEC 60601-2-25 standard for medicalelectrical equipment
4. Design parameters (see Table 2.1)
2.1 specifications of the ecg system 7
Table 2.1: Design parameters of the ECG system.
Min Typ Max
Supply voltage (DC) 9 V 24 V 36 V
Number of ECG electrodes 4 + 1
Analog preamplification 1x 6x 12x
AD converter resolution 12 bits
Input bandwidth 0.05 Hz 250 Hz
Low pass filter frequency (human) 35 Hz
Low pass filter frequency (animal) 75 Hz
High pass filter frequency 0.05 Hz
Power line filter notch frequency 50 Hz /60 Hz
Input sensitivity 10 µV
CMRR 100 dB
Input voltage range ±300 mV
Input impedance 20 MΩ
Input-referred noise 15 µV
Heart rate (human) 400 bpm
Heart rate (animal) 800 bpm
Trigger latency 100 ms
Trigger jitter 10 ms
3P R O T O T Y P I N G
3.1 choice of architecture
Many digital ECG designs are based on discrete electronic compo-nents. Depending on the level of complexity used, the entire ana-log signal processing stage is implemented using numerous separateIntegrated Circuits (ICs) up to the point of Analog-to-Digital (AD) con-version. Publicly available examples include simple, 2-electrode ECGmonitor circuit designs by Analog Devices [11] and by Texas Instru-ments (TI) [12], and the ADS1258 Medical Development Kit (MDK) byTI [13].
However, at present several Analog Frontend (AFE) ICs are avail-able that implement a significant portion of the analog functionalityrequired for ECG in one piece of silicon [14]. This might include leadsignal preamplification, an RLD DC bias circuit, generation of the WCT
voltage used as reference voltage, ECG lead-off detection, respirationmeasurement over the ECG electrodes and AD conversion.
In order to facilitate hardware development and accelerate the de-sign, I decided to execute the project based on an AFE IC. Two optionswere commercially available for consideration:
• the ADS119x/129x1 family by TI,
• the ADAS10002 by Analog Devices.
The features put forward by both candidates are very similar. In con-trast to TI’s high-end ICs, the ADAS1000 only has 5 input channels, butthis would be sufficient to implement the 4+1 electrode ECG requiredby the specifications (see Chapter 2). However, the ADAS1000 is still inthe pre-release phase at the time of writing this document, thereforeTI’s offer was the reasonable choice.
The TI ADS129x family offers better performance than the ADS119x,notably in terms of resolution, noise and Common-Mode RejectionRatio (CMRR). Optionally, it also integrates the possibility to imple-ment respiration measurement over ECG electrodes [15]. Moreover, itis relatively inexpensive, thus it was chosen for development.
The prototyping phase was carried out using an ADS1298R IC, butthe final product does not require all the input channels it offers and,depending on the configuration, the respiration capability might also
1 http://www.ti.com/ww/en/analog/ads1298/index.shtml?DCMP=analog_
signalchain_mr&HQS=ads1291-pr
2 http://www.analog.com/en/analog-to-digital-converters/ad-converters/
adas1000/products/product.html
8
3.2 the ecg prototype 9
Figure 3.1: Block diagram of the ADS1298R [15].
be unnecessary, so one of the less expensive pin-compatible variantsmay be used in the end-product.
3.1.1 The ADS1298R IC
The ADS1298R3 is an 8-channel AFE meant for biopotential measure-ments (see Figure 3.1). It features eight low-noise Programmable GainAmplifiers (PGAs) and eight high-resolution Analog-to-Digital Con-verters (ADCs) for fully parallel data acquisition at a wide range ofconfigurable data rates. It offers low power consumption and smallinput-referred noise alongside a high CMRR. It provides built-in RLD
circuitry, lead-off detection and WCT circuitry for ECG. It also inte-grates support for respiration impedance measurement over ECG elec-trodes.
The ADS1298R provides a convenient Serial Peripheral Interface (SPI)for configuration and measurement data readout, thus it requiresonly a limited number of additional digital control signals [15].
3.2 the ecg prototype
3.2.1 The ADS1258 MDK
At the beginning of development, an ECG MDK based on TI’s ADS12584
IC was already available at Mediso. This development kit consistsof an Evaluation Module (EVM)5 by Spectrum Digital based on TI’s
3 http://www.ti.com/product/ads1298r
4 http://www.ti.com/product/ads1258
5 http://support.spectrumdigital.com/boards/evm5505/revd/
3.2 the ecg prototype 10
Figure 3.2: Photo of the ADS1298R ECG FE-PDK.
TMS320VC55056 Digital Signal Processor (DSP), and a piggy-back mod-
ule hosting the ADS1258 and accompanying circuitry. The PerformanceDemonstration Kit (PDK) also includes the firmware running on theDSP implementing ECG signal acquisition and processing, and a PC
application which receives the ECG waveforms over a serial link andplots them.
Since the ADS1258 is merely a 16-channel ADC and all ECG-relatedfunctionality is implemented using external components, this IC wasnot considered for design. However, the TMS320VC5505 EVM has provedto be essential in later development (see Section 3.2.3), and I alsoreused the PC application supplied with the MDK (see Section 3.3).
3.2.2 The ADS1298R ECG FE-PDK
To start experimenting with the AFE I’ve chosen, I ordered an ADS1298R
PDK. The kit contains the MMB0 Modular EVM motherboard that hostsa TI TMS320VC5509
7 DSP, and the ADS1298R ECG AFE on a piggy-backboard. The assembled system is shown in Figure 3.2.
Once out of the box and wired up to a PC, the demo kit acquiredECG signals perfectly, yet it wasn’t usable for further developmentbecause of the lack of documentation and development tools for the
6 http://www.ti.com/product/tms320vc5505
7 http://www.ti.com/product/tms320vc5509
3.2 the ecg prototype 11
Figure 3.3: Photo of the ECG prototype system connected to an ECG simulator.The acquired ECG waveforms are visible on the LCD display.
MMB0 board. According to TI, “The MMB0 is not intended to be a fullembedded development tool, it is an evaluation platform.”8
3.2.3 The ECG prototype system
Fortunately, the two PDKs share the same piggy-back interface, hencemounting the ADS1298R piggy-back module on the TMS320VC5505 EVM
is possible. The piggy-back interface provides all the necessary sig-nals for the SPI interface used in the communication with the ADS1298R
IC and the digital control signals. Since the EVM offers excellent devel-opment and debug support through TI’s Eclipse-based Code Com-poser Studio suite, it is an ideal platform for a detailed explorationof the features of the ADS1298R AFE. The EVM also features a colorAMOLED9 Liquid Crystal Display (LCD) and a Universal AsynchronousReceiver/Transmitter (UART) port which were useful for visualizingthe ECG waveforms both locally and through data transmitted to a PC
(see Section 3.3).Figure 3.3 shows the ECG prototype system connected to an ECG
simulator (see Section 5.1.1). The digitized output of the simulator isdisplayed on the LCD screen.
8 Official comment on http://e2e.ti.com/support/data_converters/precision_
data_converters/f/73/t/151713.aspx, retrieved on 09/05/2012.9 Active-matrix organic light-emitting diode.
3.3 firmware development 12
3.3 firmware development
Since the ECG prototype system is implemented using the DSP moduleof the ADS1258 MDK, several features of its firmware are based on thefirmware supplied with the MDK by TI.
3.3.1 The firmware of the ADS1258 MDK
Following are the features of the firmware supplied with the ADS1258
MDK, as listed in TI’s application note [13]. I also appended brief expla-nations, mainly focusing on how they relate to the features requiredin the firmware of the ECG prototype system (see Section 3.3.2).
Data acquisition through ADC
The firmware establishes communication with the ADS1258 IC over theSPI bus for configuration and data retrieval. The ADS1298R also has anSPI interface, however, the protocol is different. In addition to adapt-ing the protocol, I rewrote the entire SPI implementation in the ECG
prototype system firmware for clarity.
Lead-off detection
ECG lead-off detection is implemented using external circuitry on theADS1258 MDK board. The firmware takes care of the communicationwith these. Since the ADS1298R implements lead-off detection inter-nally and the status information can be read out over SPI, this featureis not required in the ECG prototype system.
DC signal removal
The firmware implements a first-order digital Infinite Impulse Re-sponse (IIR) filter for baseline restoration of the ECG signal. The fre-quency response of the filter is shown in Figure 3.4. The filter is im-plemented in platform-independent C code and I reused it in theprototype firmware almost without modifications.
Multi band-pass filtering
The firmware implements a digital Finite Impulse Response (FIR) fil-ter. The coefficients of the filter realize a low-pass filter for the elim-ination of high-frequency noise with a notch for the suppression ofpower line interference. Figure 3.5 shows the frequency response ofthe filter with coefficients realizing a notch at 50 Hz.
The FIR filter is implemented in architecture-specific assembly code.I reused the implementation in the ECG prototype system firmwarewith different filter coefficients yielding different filter characteristics.
3.3 firmware development 13
Figure 3.4: Frequency response of the IIR filter of the ADS1258 MDK [13]. Thefilter suppresses the DC component of the signal for baselinerestoration.
Figure 3.5: Frequency response of the FIR filter of the ADS1258 MDK [13]. Thefilter has a notch at 50 Hz to suppress power line interferenceand has a stopband with a very high attenuation above 150 Hzto filter out high frequency noise.
3.3 firmware development 14
ECG leads formation
The firmware carries out ECG lead formation from the single endedinput signals with simple arithmetics according to the rules describedin Section 1.1.2. Since the ADS1298R has differential inputs, the arith-metic operations to perform are different.
QRS (heart rate) detection
The firmware implements the detection of the QRS complex (see Sec-tion 1.1.1) in the ECG signal for heart rate measurement. QRS detectionis a crucial feature required for cardiac gating, but this requires real-time detection with limited latency. However, this algorithm is onlyaimed at heart rate measurement, it has unacceptably high latencyand is unsuitable for the purposes of the ECG system.
Display of ECG data
The acquired ECG waveform is plotted on the AMOLED LCD screenwith a possibility to select the ECG lead to plot using the keys of theEVM5505. This feature was useful for demonstration purposes and hasbeen reused in the ECG prototype system with negligible modifica-tions.
UART communication
The acquired ECG data is transmitted over UART to the PC applicationfor display. This feature was also useful for demonstration and hasbeen reused in the ECG prototype system firmware.
3.3.2 The firmware of the ECG prototype system
The main purpose of the ECG system is to issue triggers for gatedimaging at the occurrence of QRS complexes with deterministic delay,i.e. limited and unvarying latency, and small jitter (see Chapter 2).The firmware has been designed with this requirement in mind. Mostof the project is implemented in platform-independent C code, withthe exception of the low-level CPU peripheral drivers and the FIR filter.
The flowchart of the ECG prototype system firmware is shown inFigure 3.6. A detailed description follows.
The main function
The main function performs the initialization of the entire ECG sys-tem.
It initializes the peripherals of the DSP: SPI for communication withthe ADS1298R, UART for data transfer to the PC, a Successive Approx-imation Register (SAR) ADC that reads out keypresses for LCD con-figuration, the LCD for ECG waveform display, and General Purpose
3.3 firmware development 15
Figure 3.6: Flowchart of the ECG prototype system firmware.
Input/Outputs (GPIOs) for controlling the ADS1298R and debuggingfeatures.
It then initializes and configures the ADS1298R AFE. The register set-tings can be found in Listing A.1 (for detailed explanations, see [15]).The following set of main configuration parameters has proved to beusable in all cases tested so far (see Section 5.1):
• Sample rate set to 1000 SPS by default.
• The gain of all PGAs is set to 6.
• Lead-off detection is enabled for all ECG electrodes.
• The RLD and WCT signals are generated as the average of theLA, RA and LL signals (see Section 1.1.2).
Finally, it enters an infinite loop where, whenever data is available, itplots ECG waveforms to the LCD and transmits ECG data over UART tothe PC for display. It also monitors keypresses and updates the LCD
display configuration accordingly.Figure 3.7 shows a screenshot of the PC application. The ECG wave-
forms and actual measured heart rate of a rat are displayed as ac-quired by the ECG prototype system. The measurement configurationonly included the electrodes necessary for the calculation of Lead I(see Section 5.1.3), thus the waveform displayed for Lead II is calcu-lated and displayed incorrectly and should be considered an arith-metic artifact.
3.3 firmware development 16
Figure 3.7: Screenshot of the PC application displaying the ECG waveformsand actual measured heart rate of a rat, acquired by the ECG
prototype system.
The DRDY interrupt subroutine
The ADS1298R is configured for continuous data acquisition, thus itprovides new data at the sampling rate specified in the configuration,signaling its availability through its Data Ready (DRDY) output signal.The prompt response of the ECG system to the availability of newdata is ensured by an interrupt mapped to the DRDY signal. Whenevernew data is available, the firmware reads it out over the SPI bus, thenexecutes the following steps of processing:
• Data preprocessing. Lead-off information is extracted from thestatus word received from the ADS1298R.The 24-bit output of the ADC is converted to 16-bit resolutionby truncation of a configurable number of bits. To allow for aflexible input configuration, the ADC channel values are sortedinto an internal buffer according to a changeable map. The inputvalues that contain signal from any disconnected electrode arezeroed out. Based on the lead-off status, one healthy lead isselected for QRS detection.
• Filtering.
– Baseline restoration. A first-order IIR filter adapted from theADS1258 MDK (see Section 3.3.1) is used for suppressing theDC component of the ECG signal. The frequency responseof the filter is shown in Figure 3.8.
3.3 firmware development 17
Figure 3.8: Frequency response of the DC removal IIR filter. The filter restoresthe baseline of the ECG signal.
– Band-pass filtering. I reused the FIR filter implementationin the ADS1258 MDK with a set of custom coefficients I de-signed in Matlab for a 128-tap FIR filter. The lower stop-band of the band-pass filter reduces undesirable signal con-tributions from muscle movement, weakens the P and Twaves and softens the QRS complex, but since the project isnot aimed at the creation of a diagnostic quality ECG, thisis acceptable. The higher stopband is designed to filter outhigh frequency noise. Figure 3.9 and Figure 3.10 show thefrequency response of the filters used in human and smallanimal applications, respectively (see Chapter 2). Both fil-ters have linear phase in the passband, which is a desirablecharacteristic since it implies equal delay at all frequencies.FIR filters have a characteristic delay which is determinedby the location of the peak of the filter impulse response.Since both filters have this peak at half the number of taps,64 samples (see Figure 3.11), the delay introduced is 64 msat 1000 SPS.Due to the integer arithmetics used in the FIR filter im-plementation, the filter coefficients were converted fromdouble-precision floating point to integers. As a result, pre-serving the relatively flat response in the passband, thefilters have significant amplification. However, with suf-ficient care taken to avoid arithmetic overflows, this wasvery easily compensated.
– Notch filtering. Since the stopband of the band-pass filter forpreclinical applications lies above the frequency of powerline interference, in these applications, a separate filter isrequired for the suppression of these undesirable contribu-
3.3 firmware development 18
Figure 3.9: Frequency response of the human band-pass FIR filter. Smallerheart rates allow for a lower cutoff frequency.
Figure 3.10: Frequency response of the small animal band-pass FIR filter.Higher heart rates require a wider passband.
Figure 3.11: Impulse response of the human band-pass FIR filter.
3.3 firmware development 19
Figure 3.12: Frequency response of the IIR notch filter for power line inter-ference suppression. Notch frequency is 50 Hz. The filter has arelatively blunt notch to reduce transients.
tions. For this purpose, I designed an IIR filter in Matlab.Based on the coefficients obtained, I implemented this fil-ter as a cascade of two second-order IIR stages in platform-independent C code. Figure 3.12 depicts the frequency re-sponse of the filter with a 50 Hz notch.
• QRS detection. As discussed in Section 3.3.1, the QRS detection al-gorithm implemented in the ADS1258 MDK is unsuitable for thepurposes of the ECG system. Therefore, I implemented a detec-tion algorithm based on the Pan-Tompkins algorithm [16].The algorithm relies on the previously filtered signal. As men-tioned above, the band-pass filter distorts the P and T waves inthe ECG signal. It also softens the peak of the QRS complex, how-ever, by subsequent application of a five-point digital derivativefilter and squaring its output, the output has proved to be anexcellent metric for QRS detection in all cases tested (see Sec-tion 5.1). Pan and Tompkins applied a moving window averagefilter to this signal, which yields a series of pulses and allows forQRS detection through rising edge detection. I simply performthreshold crossing detection on the squared derivative with anadaptive threshold established based on the last five detectionamplitudes. Each detection inhibits further detections for a con-figurable amount of time, thereby avoiding multiple detectionsof the same QRS complex.
• UART and LCD flag generation. Not all samples are sent to theUART and LCD. The ratio of decimation is configurable and thecorresponding data ready flags, processed by the main loop, aregenerated accordingly.
4H A R D WA R E D E S I G N
I carried out the hardware design in cooperation with my colleagueSándor Török. Many design choices originate from his professionalexperience.
4.1 choice of microprocessor
The specification of the project requires the implementation of a vari-ety of communication interfaces (see Chapter 2). Among those speci-fied, the TI TMS320C5000 Ultra Low-Power DSP family supports USB, butnot IEEE 1588 Ethernet or CAN. These interfaces could have been real-ized using external ICs, but in order to keep the number of electroniccomponents to a minimum, I decided to search for a microprocessorthat has integrated support for all required interfaces. This means achange of architecture, but since a significant part of the ECG systemfirmware is platform-independent high-level C code, porting the codeis possible with reasonable effort.
In order to provide ample processing power for filtering the ECG
signal, I started looking for another DSP. Among the manufacturers Iconsidered - TI, Analog Devices and Freescale -, TI appeared to havethe most detailed documentation, and they offer excellent on-linetools for microprocessor selection1. In addition, I had already famil-iarized myself with their design tools, therefore I decided to choose amicroprocessor by TI.
In TI’s high performance DSP & ARM CPU line, I was left with twocandidate families. The high-end DaVinci DSP family2 has severalmembers that offer all required communication interfaces but thecapabilities of these processors far exceed the requirements of theproject and they are also rather expensive. The selection tool also rec-ommended the Sitara family3. Even though they are not DSPs, theseARM-based CPUs are aimed at performance applications and seemedto offer sufficient processing power for the ECG application at an af-fordable price. However, before I could order a development kit, acolleague brought TI’s Concerto Microcontroller Unit (MCU) family4
to my attention. These microcontrollers feature all communication in-terfaces required by the specification, and development boards werereadily available at the office.
1 http://focus.ti.com/en/multimedia/flash/selection_tools/dsp/dsp.html
2 http://www.ti.com/lsds/ti/dsp/platform/davinci/device.page
3 http://www.ti.com/lsds/ti/dsp/platform/sitara/device.page
4 http://www.ti.com/mcu/docs/mcuproductcontentnp.tsp?sectionId=
95&familyId=2049&tabId=2743
20
4.2 hardware design considerations 21
Figure 4.1: The architecture of the Concerto MCU family [17].
The Concerto family is based on a very interesting architecture,which is shown in Figure 4.1. These MCUs combine an ARM Cortex-M3 core aimed at communication with a C2000 DSP core that offerssimilar performance to the C5000 DSP I had used.
I carried out preliminary tests with the development board. I portedthe most computationally intensive part of the firmware, the FIR fil-ter (see Section 3.3) to the C2000 DSP core, and I also experimentedwith the communication between the two cores. These tests suggestedthat the high-end Concerto F28M35H52C1
5 would be very suitable forthe ECG system project. Since it is quite reasonably priced and alsooffers an integrated recyclic ADC useful for small animal respiratorymeasurement (see Section 4.2.4), it was chosen for hardware develop-ment.
4.2 hardware design considerations
Since the specification of the ECG system requires the implementa-tion of two channels on a single board, the entire ECG hardware (seeSection 4.2.3) and small animal respiratory measurement stage (seeSection 4.2.4) is doubled.
Nearly all digital signal lines in the design feature series resistorsfor limiting transient currents.
The design features several LEDs driven by the MCU to provide adirect means of conveying information to the user.
5 http://www.ti.com/product/f28m35h52c
4.2 hardware design considerations 22
4.2.1 Patient protection
The IEC 60601-2-25 standard for ECGs requires that in human applica-tions, the patient be electrically isolated from the power network andcommunication interfaces of the device. The breakdown voltage mustbe at least 4 kV.
The alternatives considered for the implementation of the isolationare the following:
• Isolation on the analog side, before the inputs of the ADS1298R.
• Isolation of the signal paths between the ECG AFE and the MCU,keeping the AFE in a separate power domain.
• Isolation of the external communication interfaces and power,keeping the MCU and the AFE in the same isolated power do-main.
Creating separate, isolated power domains for each ECG AFE and iso-lating the digital signal paths to the MCU offers several advantagesover the two other methods. Excellent methods are readily availablefor the isolation of on-board digital signal paths, making this alterna-tive less cumbersome than the medical quality isolation of the com-munication interfaces and more robust than the isolation of analogsignals. Moreover, this solution intrinsically fulfills the requirementsfor isolation between the two ECG stages for two patients, thus it ap-peared to be the best choice.
In order to provide the required isolation strengths, the portionsof the board containing the ECG AFEs have been protected againstleakage currents by incisions in the board layout (see Figure 4.3).
Additionally, the main power supply (see Section 4.2.2) and theCAN interfaces (see Section 4.2.6) are also electrically isolated, eventhough no medical quality isolation is provided in these cases.
4.2.2 Power supplies
The external components required for each regulator have been se-lected following the guidelines set forth in the respective datasheets.A number of ferrite bead chokes have been used throughout the de-sign for the suppression of high frequency interference on power sup-ply lines.
Main power supply
Since the highest supply voltage required in the ECG system is 5 V,this has been chosen as the output voltage of the main power supply.In order to comply with the requirement for a wide input voltagerange and provide electrical isolation between the power network and
4.2 hardware design considerations 23
the board, the TEN5-2411WI DC-DC converter by TracoPower has beenselected. This power supply provides a regulated output voltage of5 V with a maximal output current of 1 A, and offers an efficiency oftypically 78% [18]. Its output current is sufficient for the application,and since the sensitive components require 3.3 V power, its ripple willbe suppressed by additional linear regulators.
ECG power supplies
The ECG circuitry resides in a separate power domain, which must beisolated from the rest of the board to comply with the requirementsof the standard (see Section 4.2.1). In order to achieve this, the RV-0505S
DC-DC converter by Recom was used. This device possesses the medi-cal approval required by the application. With an input voltage of 5 V,it supplies a maximum output current of 400 mA at an unregulatedoutput voltage of 5 V, with standard efficiencies of 70-75% [19].
Since the ADS1298R requires 3.3 V power and the output of the iso-lated power supply is unregulated, additional regulation is necessary.In order to minimize the ripple of the output voltage, linear regu-lation is desirable. Moreover, to suppress interference from the digi-tal side of the IC to the analog side and thereby reduce the noise ofthe ADC, the best solution is having two separate regulators providepower to the analog and digital domains of the AFE. The TPS79201 byTI was chosen for its ultralow noise and high Power Supply RejectionRatio (PSRR) [20]. Its current sourcing capabilities are largely sufficientfor the requirements of the ECG stage.
CAN power supplies
The two CAN interfaces used (see Section 4.2.6) are also isolated. Theirisolated power domains are supplied by TME 0505S isolated DC-DC con-verters from TracoPower, which provide a maximal output current of200 mA at 5 V with an input voltage of 5 V, with a typical efficiencyof 70% [21].
MCU power supply
The F28M35H52C1 MCU can operate with a single 3.3 V power supplysince it has internal regulators supplying all other voltages it requires.TI’s TPS79601 was selected for its excellent performance similar to theregulators used for the ADS1298R and its higher current sourcing capa-bilities up to 1 A [22]. Since in this case, the suppression of noise to theanalog subsystem is not a critical issue, the same regulator suppliesboth digital and analog power, with a ferrite bead choke separatingthe two power domains to ensure high frequency noise reduction.
4.2 hardware design considerations 24
4.2.3 ECG hardware
The external components required by the ADS1298R AFE have beenselected following the guidelines of the datasheet [15].
Patient protection
As discussed in Section 4.2.1, medical quality electrical isolation isrequired in ECG systems. In addition to placing the ECG electronicsin an isolated power domain, this also requires the isolation of thedigital lines to and from the AFE. Traditionally, optocouplers wereused for this purpose, but this technology suffers from aging effectsin the LED and difficulties at higher signaling speeds. The iCouplertechnology by Analog Devices overcomes these limitations, it offersfar superior performance and significantly lower power consumption[23]. The ADuM2401 device chosen for isolation offers a suitable num-ber of input and output channels, can operate at the foreseen SPI busfrequency of 10 MHz and complies to the requirements of the IEC
60601 standard.Since the two ECG AFEs share the same SPI bus, additional buffers
gated with the respective SPI Enable signals were required on theMaster In Slave Out (MISO) lines to avoid two ADuM2401 outputs con-currently driving the MISO input of the MCU.
The clock signal is routed to both AFEs from the same 2.048 MHzclock generator residing in the main power domain through the elec-trical isolators.
The analog input stage
The ADS1298R features a Delta-Sigma (∆Σ) AD converter. The ∆Σ mod-ulator and the on-chip digital decimation filters can be used to filterout the noise at higher frequencies, thus the complexity of analogantialiasing filters required can be dramatically reduced with respectto Nyquist ADCs. Effectively, the input antialiasing filters are only re-quired to filter out any interference at frequencies around multiplesof fMOD, the sampling frequency of the modulator [15]. In High Res-olution mode, fMOD = fCLK/4 = 512 kHz.
Therefore, only a simple RC filter is required for antialiasing at theinput of the ADS1298R ICs, which was adapted from the circuit designput forward by TI in [24]. Figure 4.2 shows the magnitude responseof the antialiasing filter. Complemented by the input EMI filter withinthe ADS1298R IC [15], this ensures the required antialiasing.
A neon lamp was added in parallel to the filter in order to providethe defibrillator protection required by the specifications on the in-puts of the AFE IC. The neon lamp has a maximum strike voltage of90 V [25]. In conjunction with the resistor in series, it limits transient
4.2 hardware design considerations 25
Figure 4.2: Frequency response of the input antialiasing filter. The graphconfirms that, together with the EMI filter within the ADS1298R,the filter ensures the antialiasing required.
currents reaching the inputs of the ADS1298R to approximately 9 mA,which the inputs can withstand [15].
4.2.4 Small animal respiratory measurement
Based on previous experience, the SDP1108-R differential pressure sen-sor by Sensirion has been chosen for the purpose of respiratory mea-surement. It requires a 5 V DC power supply and provides a fullycalibrated analog voltage output. It offers a full scale pressure differ-ence of 500 Pa with superior accuracy, outstanding resolution and lowtemperature dependence through internal compensation [26].
The proposed measurement configuration consists of a tiny respi-ration sensor cushion6 the small animal lies on for measurement. Theoutput of the cushion is connected to the Hi terminal of the SDP1108-R,while the Lo terminal is left open, thus atmospheric pressure is usedas reference. The performance of the SDP1108-R makes it highly suit-able for measuring the small volume flows generated by the cushionthrough the breathing of the animal.
The analog output voltage of the SDP1108-R is connected to an ADC
input of the MCU for digitization. Since the output voltages rangefrom 0.25 V to 4 V and the ADC operates with a reference voltage of3 V, a voltage divider is required between the output and the input. Inaddition, an optional RC filter has been added to the design to ensurethe possibility of limiting high-frequency noise.
6 E.g. http://www.medicare.ie/mother-baby/infant-respiration-monitors-
rental-1/graseby-respiration-sensor.html.
4.2 hardware design considerations 26
4.2.5 Microcontroller unit
The quartz oscillator, decoupling capacitors, pullup/pulldown resis-tors and other required external components have been laid out ac-cording to the guidelines of the datasheet [27].
A JTAG interface has been implemented for programming and de-bugging.
The F28M35H52C1 offers a total of five SPI interfaces. Both ECG AFEs
are connected to the only one of those that is controlled by the DSP
core to allow for readout and data processing by the same core.In addition to the output of the pressure sensors (see Section 4.2.4),
two supply voltages have been connected to ADC inputs of the Con-certo through resistive dividers to allow monitoring the state of thepower supplies.
4.2.6 Communication interfaces
Ethernet
In addition to the Ethernet MAC in the Concerto, a transceiver IC andan RJ-45 connector is required for Ethernet communication. This cir-cuitry has been adapted from the circuit design of the TI H52C1 Con-certo ControlCard [28].
USB
The F28M35H52C1 includes the implementation of the physical layer,therefore only a USB connector is required for communication. Sincethe ECG system is only meant to be used as a USB device, a simpleUSB-B connector has been implemented in the design, along with pro-tection diodes on the data lines.
CAN
The ECG system design makes use of both CAN interfaces of the Con-certo. CAN is used as an external communication interface and alsofor connecting to the slave modules (see Chapter 2). 9-pin D-SUB maleconnectors are used.
To complement the isolated power supplies, TI’s ISO1050 isolatedCAN transceivers [29] have been used in the design for CAN interfaceisolation.
In addition to protection diodes on the CAN data lines, the slavemodule (“bed”) CAN circuitry features a solution that disables thelocal CAN termination resistor if the slave modules for both channelsare attached. In contrast, the external CAN interface complies to theprotocol in use at Mediso, thus it features a termination resistor thatcan be enabled externally through the D-SUB connector.
4.3 the hardware 27
Figure 4.3: Rendered visualization of the populated ECG printed circuitboard. Note the incisions protecting the ECG circuitry from leak-age currents.
4.3 the hardware
Based on the aforementioned considerations, the schematic (see Ap-pendix A) and layout design of the final printed circuit board havebeen completed. At the time of writing this thesis, the circuit boardis being manufactured. Figure 4.3 shows a rendered visualization ofthe populated printed circuit board.
5M E A S U R E M E N T S
Since no final ECG system hardware has been produced yet at the timeof writing this thesis, all measurements described in this chapter werecarried out using the ECG prototype system. However, the expectedperformance of the final hardware is very similar.
The output of the ECG system has not yet been calibrated for con-verting output values to voltages, thus all amplitude axes show valuesin arbitrary units.
5.1 ecg signal acquisition
5.1.1 ECG simulator
Initial ECG measurements were made with an ECG simulator, a devicefeaturing connectors for ECG lead wires and capable of generatingECG waveforms. The ECG system was connected to the simulator witha standard human ECG cable in a 5-electrode configuration, as shownin Figure 3.3. The following examples illustrate the performance ofthe system with input signals from the simulator.
Figure 5.1 shows a recording of the ECG simulator output with am-plitude set to 1 mV and heart rate set to 60 bpm. The unfiltered ECG
data features very characteristic P and T waves and QRS complex,along with a negligible contribution from power line interference. Inthis case, the unfiltered data would be directly usable for QRS de-tection, however, filtering was carried out for comparison. Since thebandpass filter for human applications was used, use of the notchfilter for power line interference suppression was not required.
Suppression of the DC component by the baseline restoration filteris well visible in the filtered data. The limited delay and waveformdistortion resulting from the FIR filter are also identifiable, however,these effects are acceptable (see Section 3.3.2). As a result, the discretederivative provides a perfectly usable QRS metric.
Figure 5.2 depicts a recording of the output of the simulator withthe same settings (note the different scale on the time axis), deliber-ately corrupted by 50 Hz power line noise by placing a soldering ironclose to the ECG system. The filtering does a remarkable job in elimi-nating power line interference and a usable QRS metric is produced.
28
5.1 ecg signal acquisition 29
Figure 5.1: Recording of the ECG simulator output.
Figure 5.2: Recording of the ECG simulator output, corrupted by 50Hz noise.This figure demonstrates the efficiency of the notch filter in elim-inating power line interference.
5.1 ecg signal acquisition 30
Figure 5.3: Human ECG recording. This figure illustrates the power of the fil-ters in suppressing power line interference and baseline wander.
5.1.2 Human ECG
Figure 5.3 shows the author’s ECG recording at rest, acquired in a 3-electrode configuration with LA and RA input electrodes, and the RLfeedback electrode connected to the appropriate limbs. The record-ing exhibits noticeable power line interference and breathing causessignificant baseline wander in the ECG trace. However, these distor-tions are eliminated by the DC suppression and bandpass filters. Thisyields a very high quality metric for QRS detection. The average heartrate during the measurement can be identified as approximately 75-80 bpm.
Figure 5.4 shows the author’s stress ECG recording while perform-ing squats. The baseline restoration filter performs remarkably oncemore, however, due to muscle movement, the unfiltered ECG dataexhibits sharp transitions the currently used filters cannot eliminatesatisfactorily. This leads to the presence of spurious pulses in the QRS
metric, thus unwanted QRS triggers.
5.1.3 Small animal ECG
To verify its capability to acquire the ECG signals of rodents, the ECG
system was tested with anesthetized small animals. These recordings
5.1 ecg signal acquisition 31
Figure 5.4: Human stress ECG recording. This figure demonstrates the chal-lenges faced in the processing of stress ECG signals.
were carried out in a 3-electrode ECG configuration with LA and RAinput electrodes and a feedback electrode, RL, connected to the limbsof the animal. 3M Red Dot neonatal, pre-wired, radiolucent moni-toring electrodes were employed with a standard, unshielded ECG
cable. The bandpass filter for small animal measurements was used,therefore the notch filter for power line interference was required (seeSection 3.3.2).
The photo of an anesthetized rat fitted with ECG electrodes can beseen in Figure 5.5. Figure 5.6 shows the ECG signal recording. The un-filtered signal suffers from significant power line interference. Due tothe transients of the notch filter, this interference is only attenuated,not completely eliminated. Filtering also reduces the amplitudes ofthe QRS complexes. However, both the unfiltered and the filtered sig-nal yield a usable QRS metric through the discrete derivative filter.The heart rate of the rat was measured to be around 235 bpm (seeFigure 3.7), which is in accord with the values found in the literature[30].
Figure 5.7 shows the ECG of a hamster. Surprisingly, the signal fea-tures inverted QRS complexes due to biological reasons. The signal ex-hibits limited baseline wander and negligible power line interference.In this case, filtering yields no obvious improvement with respect tothe unfiltered signal, and the amplitudes of the QRS complexes are re-
5.1 ecg signal acquisition 32
Figure 5.5: The anesthetized rat fitted with ECG electrodes.
Figure 5.6: Rat ECG recording. This figure illustrates the capability of theprocessing to create a usable QRS metric.
5.2 performance in mri 33
Figure 5.7: Hamster ECG recording. This figure demonstrates the ability ofthe processing to create a usable QRS metric.
duced. However, as was the case with the ECG of the rat, both filteredand unfiltered signals are usable for QRS detection. The heart rate ofthe hamster was measured to be around 430 bpm, which correspondsto the values found in the literature [31].
5.2 performance in mri
Mediso assembles permanent magnet MRI machines for small animalapplications. Compared to electromagnet-based devices, these ma-chines offer the advantage of a limited fringe field [32], thus metal-lic objects may be brought into the vicinity and ordinary electronicdevices may be operated near the machine. This means that ECG mea-surement in MRI might be possible by setting up the ECG system nextto the MRI machine and using MRI-compatible ECG electrodes withinthe bore of the MRI magnet.
The manufacturer claims that the 3M electrodes I used for smallanimal ECG measurements (see Section 5.1.3) are not MRI-compatible.Nevertheless, the electrodes are non-ferrous and the pre-attached leadwires are made of carbon. Moreover, Small Animal Instruments, Inc.use the same electrodes in their MRI-compatible ECG system [10], thusthese electrodes were good candidates for measurements in MRI.
5.2 performance in mri 34
5.2.1 Initial measurements
In order to assess the viability of measuring ECG as outlined above,I carried out initial measurements with an MRI water phantom andthe 3M electrodes within the magnet bore. The two ECG electrodesthat form the ECG lead under measurement were connected togetherwith approximately 100 kΩ contact impedance, which reasonably ap-proximates usual ECG contact impedances. This setup was placed inthe MRI animal bed underneath the phantom and inserted into theMRI bore. No actual signal source was connected to the electrodes,thus the measurement was expected to register background noise andeventual MRI contributions.
Figure 5.8 shows the result of the acquisitions. The upper plot de-picts the spectrum of the background acquired within the MRI bore,with no MRI sequence in progress. The spectrum is clearly dominatedby 50 Hz power line interference with significant harmonic content.The lower plot shows the spectrum of the data acquired during theexecution of a scout MRI sequence. I deemed the results promising,since MRI contributions are only visible around 260 Hz and 480 Hz,outside of the frequency band of interest, and should therefore beeasy to attenuate. It is noteworthy that 260 Hz corresponds to therepeat period of the radio frequency pulses in the scout sequence,3.8 ms. It should also be noted that although the spectra of the 50 Hzharmonics are visually similar, MRI interference appears to amplifyseveral harmonics.
The MRI images acquired of the phantom with the electrodes inplace were only slightly deteriorated, showing limited signs of inter-ference generated by the electrodes. The electrodes were not visibleon the MRI image.
5.2.2 Small animal measurements
ECG measurements
Based on the positive experience gained with the initial measure-ments, I attempted measuring the ECG of a hamster within MRI. Iused the same electrode configuration as in Section 5.1.3. Measure-ments with the hamster inside the MRI bore, but with the gradientand radio frequency systems of the MRI machine turned off, yieldedresults in excellent agreement with those obtained outside of the ma-chine. Thus, the field of the permanent magnet appears to have nonoticeable influence on the measurement.
Figure 5.9 shows the result of the measurements. The upper plot de-picts the spectrum of the hamster ECG acquired within the MRI bore,with no MRI sequence in progress. It is interesting to note that thespectrum appears to be discrete, which suggests that the heart rateof the hamster is very steady. The lower plot shows the spectrum of
5.2 performance in mri 35
Figure 5.8: Spectrum of the background signal acquired within the MRI bore,without MRI sequence in progress and during a scout sequence.
the data acquired during the execution of a scout MRI sequence. MRI
interference appears at the same frequencies that were identified ear-lier, however, in this case, it dominates the signal. This demonstratesa weakness of the filtering in use: Figure 5.10 shows that the lowfrequency spectrum of the hamster ECG is partially preserved, withsome frequency components diminished. However, the filtering can-not sufficiently attenuate the high-frequency interference, thus QRS
detection on the filtered signal fails.Figure 5.11 shows an ECG spectrum acquired during execution of
a spin echo MRI sequence. The high frequency components of thespectrum were attenuated by the small animal band-pass filter. ThisMRI sequence appears to cause a wide-band interference within thefrequency band relevant for ECG acquisition, along with some veryhigh peaks. QRS detection on this signal yields no usable results.
MRI measurements
In contrast to the phantom measurements described in Section 5.2.1,no usable MRI images of the hamster fitted with ECG electrodes couldbe acquired with the sequences tested.
5.2 performance in mri 36
Figure 5.9: Spectrum of a hamster ECG acquired within the MRI bore. Duringthe scout sequence, MRI artifact dominates the spectrum.
Figure 5.10: Spectrum of a hamster ECG acquired within the MRI bore,zoomed in. The spectrum has been partially preserved at lowfrequencies, even despite MRI artifact from the scout sequence.
5.2 performance in mri 37
Figure 5.11: Spectrum of a hamster ECG acquired within the MRI bore duringa spin echo sequence. The high frequency components wereattenuated by the small animal band-pass filter.
5.2.3 Discussion
At present, the interactions between the ECG system and the MRI ma-chine impair both ECG signal acquisition and MRI image creation.
MRI image creation
The comparison of MRI images acquired of the phantom and of thehamster, both in presence and absence of ECG electrodes, shows adegradation in the quality of the MRI images when the electrodes arepresent. The images suggest that this effect is due to radio-frequencyinterference caused by the presence of the ECG electrodes in the MRI
bore, which probably act as antennas in guiding interference into theMRI acquisition volume. However, the degradation is much more sig-nificant in the case of hamster measurements, where it yields unus-able images, whereas for phantom measurements, it merely causesgrainy artifacts but image quality remains acceptable. This impliesthat the body of the animal amplifies the radio-frequency interferenceguided into the MRI volume by the electrodes.
Nonetheless, in the case of phantom measurements, the ECG elec-trodes were connected together with a contact impedance (see Sec-tion 5.2.1), while in the case of hamster measurements, they wereconnected to the limbs of the animal (see Section 5.2.2). In order to
5.2 performance in mri 38
evaluate the role of the animal body in the amplification of the in-terference, phantom measurements should be carried out in a similarconfiguration, ensuring that the electrodes are physically separated.Since the phantom is made of plastic and is thus probably an insu-lator, appropriate contact impedances must be ensured between thedifferent electrodes. Due to the limited amount of time available af-ter the first animal MRI experiments, this measurement has not beencarried out yet.
Colleagues with MRI-related experience suggested adding 1 nF fil-tering capacitors between all ECG electrode lines and the commonground to reduce radio-frequency interference, but our measurementscarried out with the phantom showed no quantifiable improvementnevertheless.
ECG acquisition
A tentative simulation showed that by applying a band-pass filterwith sufficient attenuation in the higher stopband to the ECG signalcorrupted by interference from the scout sequence, the remaining low-frequency components (see Figure 5.10) can be used to produce an al-most usable QRS metric. This implies that in the case of this sequence,further tweaking of the already implemented digital filtering couldprobably be used to satisfactorily eliminate MRI interference from theECG signal.
However, several other sequences that were tested, e.g. spin echo,produce wide-band interference within the frequency band of inter-est for ECG measurement (see Figure 5.11). In these cases, the filter-ing used in the ECG prototype system gives little hope of eliminatingMRI interference, thus new filtering techniques need to be considered.Since the stray magnetic field of the solenoidal radio-frequency exci-tation coil of the MRI machine can be expected to be small, and asdiscussed in Section 4.2.3, the inputs of the ECG AFE are equippedwith sufficient analog filtering against out-of-band interference, digi-tal filtering will probably be appropriate.
Measurements also revealed that the position of the ECG cable hasa great influence on the sensitivity of the system to power line inter-ference. Further investigations could help establishing good practicesof cable layout.
6S U M M A RY
6.1 accomplishments
I spent the past year working at Mediso Medical Imaging Systems.During this time, based on two different development kits, I producedan ECG prototype system, which is fully functional with the firmwareI designed. I demonstrated the ability of the system to acquire humanand small animal ECG waveforms and detect QRS complexes.
Based on the experience gained with the prototype system, we de-signed the final ECG system board. The circuit board is being man-ufactured at the time of writing this thesis. Utilizing my work donefor the prototype and a demonstration board, I have completed thedevelopment of a great part of the firmware of the future ECG system.
6.2 possibilities of improvement
The measurements carried out so far show that the system is not yetable to perform satisfactorily in the MRI environment. The interactionsof the ECG system with MRI require further investigation in order toachieve MRI-compatibility.
The firmware developed for the final ECG system needs to be com-pleted and it should be tested with the final board as soon as thelatter is available.
The filters in use should be reviewed in order to eliminate theglitches encountered during the acquisition of human stress ECG.
The ECG output amplitudes need to be calibrated in order to allowvoltage readout, taking into consideration the amplification of theband-pass filters.
By keeping the currently used filters for QRS detection, and im-plementing another set of filters which satisfactorily preserve all keyfeatures of the ECG waveform for display, a diagnostic quality ECG
could probably be produced.
39
AA P P E N D I X
Listing A.1: Configuration of the ADS1298R.
// ADS1298R register values
Uint8 ADS1298RegVal[25] =
#if ECG_SAMPLERATE == 500
0x86, // Register CONFIG1: Hi-Res EN, data rate = 500SPS
#elif ECG_SAMPLERATE == 1000
0x85, // Register CONFIG1: Hi-Res EN, data rate = 1000SPS
#elif ECG_SAMPLERATE == 2000
0x84, // Register CONFIG1: Hi-Res EN, data rate = 2000SPS
#elif ECG_SAMPLERATE == 4000
0x83, // Register CONFIG1: Hi-Res EN, data rate = 4000SPS
#else
#error Sample rate not supported!
#endif
0x10, // Register CONFIG2: variable WCT chopping frequency
0xDD, // Register CONFIG3: RLD buffer EN; Vref = 2.4V
0x03, // Register LOFF: DC current source Loff detection
0x00, // Register CH1SET: normal operation, PGA gain = 6
0x00, // Register CH2SET: id.
0x00, // Register CH3SET: id.
0x00, // Register CH4SET: id.
0x00, // Register CH5SET: id.
0x00, // Register CH6SET: id.
0x00, // Register CH7SET: id.
0x00, // Register CH8SET: id.
0x06, // Register RLD_SENSP: channels 2 and 3 (LA, LL) in RLD
0x02, // Register RLD_SENSN: channel 2 (RA) in RLD
0xFE, // Register LOFF_SENSP: Loff EN on all chans but IN1P
0x02, // Register LOFF_SENSN: Loff EN on IN2N (RA)
0x00, // Register LOFF_FLIP: Loff curr direction: default
0x00, // Register LOFF_STATP (read only)
0x00, // Register LOFF_STATN (read only)
0x0F, // Register GPIO: all GPIOs set to input (unused)
0x00, // Register PACE: PACE detect buffer off
0xF0, // Register RESP: respiration switched off
0x22, // Register CONFIG4: Loff comparators powered up
0x0A, // Register WCT1: IN2P (LA) -> WCTA
0xE3 // Register WCT2: IN3P (LL) -> WCTB, IN2N (RA) -> WCTC
;
40
<AP
PR
OV
ER
>
<><><>
<><><>
1234
<DA
TE
>
<><><>
-<><><>
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Roo
t She
et
Fö
Lap
RE
LEA
SE
D T
O P
RO
DU
CT
ION
<G
YÁ
RT
ÁS
RA
BO
CS
ÁT
OT
T>
<> o
ldal
, <>
<> o
ldal
, <>
<> o
ldal
, <>
A1
Loca
ted
at s
heet
TO
P -
Top
Top
Blo
ck R
oot P
age
1 of
1S
heet
1 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Rev
isio
n <V
álto
zat>
Cha
nge
<Vál
tozá
s>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:
App
rove
r <J
óváh
agyó
>E
CO
Dat
e <D
átum
> MW
here
<H
ol>
Doc
. Rev
.
EC
G_I
nput
s1
EC
G_C
1
EC
G_R
L1
EC
G_L
L1
EC
G_R
A1
EC
G_L
A1
ELE
C_R
A1
ELE
C_L
A1
EC
G In
terf
ace
1
ELE
C_R
A
ELE
C_L
A
CLK
ST
AR
T
CS
SC
LKDIN
DR
DY
_O
DO
UT
_OLA R
A
LL RL
C
RE
SE
T
EC
G In
terf
ace
2
ELE
C_R
A
ELE
C_L
A
CLK
ST
AR
T
CS
SC
LKDIN
DR
DY
_O
DO
UT
_OLA R
A
LL RL
C
RE
SE
T
Pow
er_S
uppl
ies
VC
C_M
EA
S
3V3_
ME
AS
DS
P
EC
G2_
MO
SI
EC
G2_
RE
SE
T
EC
G1_
RE
SE
T
EC
G1_
ST
AR
T
EC
G1_
DR
DY
EC
G1_
MIS
O
EC
G1_
CS
EC
G1_
SC
LK
EC
G1_
MO
SI
EC
G2_
MIS
O
EC
G2_
SC
LK
EC
G2_
DR
DY
EC
G2_
CS
EC
G2_
ST
AR
T
UP
1
UP
2
ET
H_M
DC
ET
H_M
DIO
ET
H_T
XD
[3:0
]E
TH
_TX
EN
ET
H_R
ST
ET
H_T
XC
LK
ET
H_R
XD
[3:0
]
ET
H_C
OL
ET
H_C
RS
ET
H_R
XE
RE
TH
_RX
CLK
ET
H_R
XD
V
ET
H_I
NT
R
US
B_D
P
CA
NT
X
CA
NR
X
MC
AN
TX
MC
AN
RX
VC
C_M
EA
S
3V3_
ME
AS
Bed
_LE
D_R
1
Bed
_LE
D_Y
1
Bed
_LE
D_G
1
EC
G_L
ED
1
EC
G_L
ED
2
Bed
_LE
D_G
2
Bed
_LE
D_Y
2
Bed
_LE
D_R
2
US
B_D
MU
SB
_VB
US
EC
G_I
nput
s2
EC
G_C
2
EC
G_R
L2
EC
G_L
L2
EC
G_R
A2
EC
G_L
A2
ELE
C_R
A2
ELE
C_L
A2
Isol
ator
s
MIS
O
DR
DY
MO
SI
SC
LK
CS
ST
AR
T
CLK
CLK
2
ST
AR
T2
CS
2
SC
LK2
MO
SI2
DR
DY
2
MIS
O2
EC
G2_
ST
AR
T
EC
G2_
DR
DY
EC
G2_
MIS
O
EC
G2_
CS
EC
G2_
SC
LK
EC
G1_
ST
AR
T
EC
G1_
DR
DY
EC
G1_
CS
EC
G1_
MIS
O
EC
G1_
SC
LK
EC
G1_
MO
SI
RE
SE
TE
CG
1_R
ES
ET
RE
SE
T2
EC
G2_
RE
SE
T
EC
G2_
MO
SI
Pre
ssur
e S
enso
rs UP
2
UP
1
Eth
erne
t Int
erfa
ce
ET
H_I
NT
R
ET
H_R
XD
VE
TH
_RX
CLK
ET
H_R
XE
R
ET
H_C
RS
ET
H_C
OL
ET
H_R
XD
[3:0
]
ET
H_T
XC
LK
ET
H_R
ST
ET
H_T
XE
NE
TH
_TX
D[3
:0]
ET
H_M
DIO
ET
H_M
DC
US
B in
terf
ace
US
B_D
MU
SB
_DP
US
B_V
BU
S
SA
Bed
CA
N In
terf
aces
CA
NR
X
CA
NT
X Mai
n C
AN
Inte
rfac
e
CA
NR
X
CA
NT
X
Fee
dbac
k_LE
Ds
Bed
_LE
D_R
1
Bed
_LE
D_Y
1
Bed
_LE
D_G
1
EC
G_L
ED
1
EC
G_L
ED
2
Bed
_LE
D_G
2
Bed
_LE
D_Y
2
Bed
_LE
D_R
2
MIS
O2
DR
DY
2
MO
SI2
SC
LK2
CS
2
EC
G2_
ST
AR
T
CLK
2
EC
G_C
2
EC
G_R
L2
EC
G_L
L2
EC
G_R
A2
EC
G_L
A2
ELE
C_R
A2
ELE
C_L
A2
ST
AR
T2
EC
G2_
CS
ELE
C_L
A1
ELE
C_R
A1
EC
G_L
A1
EC
G_R
A1
EC
G_L
L1
EC
G_R
L1
EC
G_C
1
EC
G2_
DR
DY
EC
G2_
MO
SI
EC
G2_
SC
LK
EC
G2_
MIS
O
MO
SI
SC
LK
CS
MIS
O
DR
DY
CLK
ST
AR
T
EC
G1_
MO
SI
EC
G1_
SC
LK
EC
G1_
CS
EC
G1_
MIS
O
EC
G1_
DR
DY
EC
G1_
ST
AR
T
EC
G1_
RE
SE
TR
ES
ET
EC
G2_
RE
SE
TR
ES
ET
2
UP
1
UP
2
ET
H_M
DC
ET
H_M
DIO
ET
H_T
XE
N
ET
H_R
ST
ET
H_T
XC
LKE
TH
_CO
LE
TH
_CR
S
ET
H_R
XE
RE
TH
_RX
CLK
ET
H_R
XD
V
ET
H_I
NT
R
US
B_D
P
MC
AN
TX
MC
AN
RX
CA
NT
X
CA
NR
X
3V3_
ME
AS
3V3_
ME
AS
VC
C_M
EA
S
VC
C_M
EA
S
Bed
_LE
D_R
1
Bed
_LE
D_Y
1
Bed
_LE
D_G
1
EC
G_L
ED
1
Bed
_LE
D_R
2
Bed
_LE
D_Y
2
Bed
_LE
D_G
2
EC
G_L
ED
2
US
B_D
MU
SB
_VB
US
ET
H_T
XD
[3:0
]
ET
H_R
XD
[3:0
]
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
TIT
LE
<CIM
>
A1
Loca
ted
at s
heet
1 -
A4
Blo
ck E
CG
_Inp
uts1
Pag
e 1
of 1
She
et 2
OF
18
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
EC
G_C
1
EC
G_R
L1
EC
G_L
L1
EC
G_R
A1
EC
G_L
A1
ELE
C_R
A1
ELE
C_L
A1
J1
C55
47pF
C56
47pF
R45
22K
R55
10K
GN
DIS
O1
GN
DIS
O1
DS
1
MC
0801
0000
GN
DIS
O1
22K
R46
GN
DIS
O1
GN
DIS
O1
10K
R56
MC
0801
0000
DS
2
GN
DIS
O1
47pF
C57
47pF
C58
C59
47pF
C60
47pF
GN
DIS
O1
DS
3
MC
0801
0000
R57
10K
GN
DIS
O1
GN
DIS
O1
R47
22K
22K
R48
GN
DIS
O1
GN
DIS
O1
10K
R58
MC
0801
0000
DS
4
GN
DIS
O1
47pF
C61
47pF
C62
22K
R49
GN
DIS
O1
GN
DIS
O1
10K
R59
MC
0801
0000
DS
5
GN
DIS
O1
47pF
C63
47pF
C64
J2 J3 J4 J5
ELE
C_L
A1
ELE
C_L
A1
ELE
C_R
A1
ELE
C_R
A1
EC
G_L
A1
EC
G_L
A1
EC
G_R
A1
EC
G_R
A1
ELE
C_R
L1
EC
G_R
L1
EC
G_R
L1
EC
G_L
L1
EC
G_L
L1E
LEC
_LL1
EC
G_C
1
EC
G_C
1E
LEC
_C1
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Pow
er S
uppl
ies
Táp
egys
égek
A1
Loca
ted
at s
heet
1 -
A1
Blo
ck P
ower
_Sup
plie
s P
age
1 of
3S
heet
3 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
GN
D10uF
C15
7
A3V
3
BLM
21P
G22
1SN
1D
L1
GN
D
GN
D
TP
S79
601D
CQ
R
U20
1
45
2IN
PG
OU
T
EN
15pF
C22
451
KR
139
30K
R15
4
3V3
GN
DG
ND
100n
FC
182
10uF
C15
8
GN
D
10uF
C15
9
GN
D
VC
C 100n
FC
183
C16
010
uF
J36
470n
FC
225
470n
FC
226
10nF
C22
7
10nF
C22
8
VIN
GN
D_V
IN2x
1mHL1
9
34
21
GN
D24
V
24V
GN
D1uF
C22
9C
230
1uF
GN
DIS
OM
TM
E05
05S
U21
4 3
2 1G
ND
1
VS
GN
D2
VO
UT
BLM
21A
G10
2SN
1DL6
GN
D
VC
C
C18
4
100n
F
GN
DIS
OM
22uF
C23
5
GN
DIS
OM
VC
C
GN
D
VC
C
TE
N5-
2411
WI
U23
1232214 16
32-V
IN
-VIN
2-V
OU
T
+VO
UT
+VIN
+VIN
2
R_O
N/O
FF
J26
100
R14
5
100n
FC
185
10uF
C12
0
2122
uFC
237
1N40
07
D1
U22
TM
E05
05S
4 3
2 1G
ND
1
VS
GN
D2
VO
UT
VC
CIS
OM
VC
CIS
O
GN
DIS
O
1uF
C23
1 GN
D
100n
F
C18
6
GN
DIS
O
C23
622
uF
GN
DIS
O
C23
21u
F
GN
D
L7 BLM
21A
G10
2SN
1D
R16
07.
5K
R16
213
K
GN
D
VC
C
C18
710
0nF
GN
D
GN
D
3V3_
ME
AS
AN
A
3V3
100n
FC
188
GN
D
7.5K
R15
9
13K
R16
1
VC
C_M
EA
SA
NA
3V3_
ME
AS
VC
C_M
EA
S
CA
N P
ower
s
GN
D100n
FC
189
10uF
C12
1
21
BLM
21P
G22
1SN
1D
L2V
CC
GN
D
C21
415
pF
GN
DIS
O1
R14
051
K
GN
DIS
O1
100
R14
6
100n
FC
190
GN
DIS
O1
C12
210
uF21
C15
11u
F
GN
DIS
O1
R14
710
0
C22
010
nF
U14
TP
S79
201
GN
D=G
ND
ISO
1
4
365
1IN
PG
OU
TE
N
BYP
C19
110
0nF
GN
DIS
O1
10uF
C12
3
21
GN
DIS
O1
R15
530
K
GN
DIS
O1
BLM
21A
G10
2SN
1D
L8
L9
BLM
21A
G10
2SN
1D
15pF
C21
5
GN
DIS
O1
51K
R14
1
30K
R15
6
GN
DIS
O1
A1
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Pow
er S
uppl
ies
Táp
egys
égek
Loca
ted
at s
heet
1 -
A1
Blo
ck P
ower
_Sup
plie
s P
age
2 of
3S
heet
4 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
R14
810
0
C19
210
0nF
GN
DIS
O1
C12
410
uF21
10uF
C12
5
21
L10
BLM
21A
G10
2SN
1D
GN
DIS
O1
1uF
C15
2
GN
DIS
O1
C19
310
0nFGN
DIS
O1
100
R14
9
100n
FC
194
GN
DIS
O1
BLM
21A
G10
2SN
1D
L11
GN
DIS
O1
10nF
C22
1
TP
S79
201
U15
GN
D=G
ND
ISO
1
4
365
1IN
PG
OU
TE
N
BYP
A3V
3IS
O1
3V3I
SO
1
GN
DIS
O1
VC
CIS
O1
RV
-050
5S
U18
15101312
2
1+V
in
-Vin
+Vou
t1
+Vou
t2
-Vou
t1
-Vou
t2
C25
510
0nF
GN
DIS
O1
GN
D100n
FC
195
10uF
C12
6
BLM
21P
G22
1SN
1D
L3V
CC
GN
D
C21
615
pF
GN
DIS
O2
R14
251
K
GN
DIS
O2
100
R15
0
100n
FC
196
GN
DIS
O2
C12
710
uF
C15
31u
F
GN
DIS
O2
R15
110
0
C22
210
nF
U16
TP
S79
201
GN
D=G
ND
ISO
2
INP
G
OU
TE
N
BYP
C19
710
0nF
GN
DIS
O2
10uF
C12
8
GN
DIS
O2
R15
730
K
GN
DIS
O2
BLM
21A
G10
2SN
1D
L12
15pF
C21
7
GN
DIS
O2
51K
R14
3
GN
DIS
O2
R15
210
0
C19
810
0nF
GN
DIS
O2
10uF
C12
91u
FC
154
C19
910
0nFGN
DIS
O2
100
R15
3
10nF
C22
3
RV
-050
5S
U19
+Vin
1
-Vin
2
+Vou
t1
+Vou
t2
-Vou
t1
-Vou
t2
VC
CIS
O2
GN
DIS
O2
GN
D=G
ND
ISO
2
TP
S79
201
U17 IN
PG
OU
TE
N
BYP
100n
FC
200
GN
DIS
O2
GN
DIS
O2
C13
010
uF
GN
DIS
O2
A3V
3IS
O2
A1
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Pow
er S
uppl
ies
Táp
egys
égek
Loca
ted
at s
heet
1 -
A1
Blo
ck P
ower
_Sup
plie
s P
age
3 of
3S
heet
5 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
30K
R15
8
GN
DIS
O2
L13
BLM
21A
G10
2SN
1D
L14
BLM
21A
G10
2SN
1D
BLM
21A
G10
2SN
1D
L15
3V3I
SO
2
C25
610
0nF
GN
DIS
O2
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
DS
P
DS
P
A1
Loca
ted
at s
heet
1 -
D2
Blo
ck D
SP
Pag
e 1
of 1
She
et 6
OF
18
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
EC
G2_
MO
SI
EC
G2_
RE
SE
T
EC
G1_
RE
SE
T
EC
G1_
ST
AR
T
CA
NT
X
EC
G1_
DR
DY
VC
C=3
V3
SN
74LV
C12
5AD
U24
1
32
EC
G1_
CS
EC
G1_
SC
LK
EC
G1_
MO
SI
CA
NR
X
U24
SN
74LV
C12
5AD
VC
C=3
V3
EC
G2_
SC
LK
EC
G2_
DR
DY
EC
G2_
CS
EC
G2_
ST
AR
T
R87
100
100
R88
R89
100
100
R90
R97
100
100
R91
R92
100
R93
100
MC
AN
TX
100
R94
R95
100
100
R96
DS
P_C
ore
EC
G1C
S
EC
GS
CLK
EC
GM
OS
I
EC
G_M
ISO
EC
G1_
DR
DY
EC
G1R
ES
ET
EC
G1S
TA
RT
EC
G2C
S
EC
G2_
DR
DY
EC
G2R
ES
ET
EC
G2S
TA
RT
UP
1
UP
2
ET
HT
XD
3E
TH
TX
D2
ET
HT
XD
1E
TH
TX
D0
ET
HM
DC
ET
H_M
DIO
ET
HT
XE
N
ET
H_C
RS
ET
H_C
OL
ET
H_T
XC
LK
ET
H_R
XD
[3:0
]
ET
H_R
XE
R
ET
H_R
XC
LK
ET
H_R
XD
V
ET
H_I
NT
R
ET
HR
ST
US
B_D
M
US
B_D
P
CA
NR
X
CA
N_T
X
MC
AN
_TX
MC
AN
RX
VC
C_M
EA
S
3V3_
ME
AS
Bed
_LE
D_R
2
Bed
_LE
D_Y
2
Bed
_LE
D_G
2
EC
G_L
ED
2
EC
G_L
ED
1
Bed
_LE
D_G
1
Bed
_LE
D_Y
1
Bed
_LE
D_R
1
US
BV
BU
S
UP
1A
NA
UP
2A
NA
ET
H_M
DC
ET
H_M
DIO
ET
H_T
XD
[3:0
]
ET
H_T
XE
N
ET
H_R
ST
ET
H_T
XC
LK
ET
H_R
XD
[3:0
]
ET
H_C
OL
ET
H_C
RS
ET
H_R
XE
R
ET
H_R
XC
LK
ET
H_R
XD
V
ET
H_I
NT
R
R16
349
.9
49.9
R16
4
R16
549
.9
R16
649
.949
.9R
167
R16
849
.949
.9R
169
US
B_D
M
US
B_D
P
MC
AN
RX
100
R98
U24
SN
74LV
C12
5AD
VC
C=3
V3
4
65
EC
G1_
MIS
O
EC
G2_
MIS
O
VC
C=3
V3
SN
74LV
C12
5AD
U24G
ND
GN
D
GN
D
3V3 10
0nF
C94
VC
C_M
EA
SA
NA
3V3_
ME
AS
AN
A
Bed
_LE
D_R
2
Bed
_LE
D_Y
2
Bed
_LE
D_G
2
EC
G_L
ED
2
EC
G_L
ED
1
Bed
_LE
D_G
1
Bed
_LE
D_Y
1
Bed
_LE
D_R
1
R22
310
K
R22
410
K
3V3
3V3
US
B_V
BU
SR
225
100K
T3
T4
GN
DG
ND
EC
G2_
MO
SI
EC
G2_
RE
SE
T
EC
G1_
RE
SE
T
EC
G1_
ST
AR
T
EC
G1_
DR
DY
EC
G_M
ISO
EC
G1_
CS
EC
G1_
CS
EC
G1_
SC
LK
EC
G1_
MO
SI
EC
G2_
SC
LK
EC
G2_
DR
DY
EC
G2_
CS
EC
G2_
CS
EC
G2_
ST
AR
T
EC
G1C
S
EC
GS
CLK
EC
GM
OS
I
EC
G1R
ES
ET
EC
G1S
TA
RT
EC
G2C
S
EC
G2R
ES
ET
EC
G2S
TA
RT
ET
H_M
DC
ET
H_M
DIO
ET
H_T
XE
N
ET
H_R
ST
ET
H_T
XC
LK
ET
H_C
OL
ET
H_C
RS
ET
H_R
XE
R
ET
H_R
XC
LK
ET
H_R
XD
V
ET
H_I
NT
R
ET
H_T
XD
3E
TH
_TX
D2
ET
H_T
XD
1E
TH
_TX
D0
ET
HT
XD
3E
TH
TX
D2
ET
HT
XD
1E
TH
TX
D0ET
HM
DC
ET
HT
XE
N
ET
HR
ST
UP
1
UP
2
US
B_D
M
US
B_D
P
CA
NT
X
CA
NR
X
CA
N_T
X
MC
AN
TX
MC
AN
RX
MC
AN
_TX
EC
G1_
MIS
O
EC
G2_
MIS
O
VC
C_M
EA
S
3V3_
ME
AS
Bed
_LE
D_R
2
Bed
_LE
D_Y
2
Bed
_LE
D_G
2
EC
G_L
ED
2
EC
G_L
ED
1
Bed
_LE
D_G
1
Bed
_LE
D_Y
1
Bed
_LE
D_R
1
US
B_V
BU
SU
SB
VB
US
ET
H_T
XD
[3:0
]
ET
H_R
XD
[3:0
]
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
DS
P C
ore
DS
P M
ag
A1
Loca
ted
at s
heet
6 -
D1
Blo
ck D
SP
_Cor
e P
age
1 of
2S
heet
7 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
EC
G1C
S
CA
NR
X
CA
N_T
X
MC
AN
_TX
EC
G1_
DR
DY
MC
AN
RX
EC
G1R
ES
ET
EC
G1S
TA
RT
A3V
3
GN
D
EC
G2C
S
100n
FC
97
EC
G2_
DR
DY
GN
D
EC
G2R
ES
ET
EC
G2S
TA
RT
U34
F28
M35
H52
C1R
FP
T
GP
IO19
214
0
GP
IO19
314
1
GP
IO19
414
2
GP
IO19
514
3
GP
IO19
611
2
GP
IO19
711
1
GP
IO19
811
0
GP
IO19
910
9
U34
F28
M35
H52
C1R
FP
T
VD
DA
111
9
AD
C1V
RE
FH
I12
0
AD
C1I
NA
012
1
AD
C1I
NA
2/A
IO2
122
AD
C1I
NA
312
3
AD
C1I
NA
4/A
IO4
124
AD
C1I
NA
6/A
IO6
125
AD
C1I
NA
712
6
AD
C1I
NB
011
7
AD
C1I
NB
311
6
AD
C1I
NB
4/A
IO12
115
AD
C1I
NB
711
4
AD
C1V
RE
FLO
118
VD
DA
213
4
AD
C2V
RE
FH
I13
3
AD
C2I
NA
013
2
AD
C2I
NA
2/A
IO18
131
AD
C2I
NA
313
0
AD
C2I
NA
4/A
IO20
129
AD
C2I
NA
6/A
IO22
128
AD
C2I
NA
712
7
AD
C2I
NB
013
6
AD
C2I
NB
313
7
AD
C2I
NB
4/A
IO28
138
AD
C2I
NB
713
9
AD
C2V
RE
FLO
135
U34
F28
M35
H52
C1R
FP
T
PA
0_G
PIO
05
PA
1_G
PIO
16
PA
2_G
PIO
27
PA
3_G
PIO
38
PA
4_G
PIO
49
PA
5_G
PIO
512
PA
6_G
PIO
613
PA
7_G
PIO
714
PB
0_G
PIO
815
PB
1_G
PIO
918
PB
2_G
PIO
1019
PB
3_G
PIO
1120
PE
6_G
PIO
3022
PE
7_G
PIO
3123
PB
6_G
PIO
1426
PB
7_G
PIO
1527
PD
2_G
PIO
1828
PD
3_G
PIO
1929
PB
4_G
PIO
1230
PB
5_G
PIO
1331
PE
2_G
PIO
2632
PE
3_G
PIO
2733
PH
3_G
PIO
5135
PH
2_G
PIO
5036
PC
4_G
PIO
6837
PC
5_G
PIO
6938
PC
6_G
PIO
7039
PC
7_G
PIO
7140
PH
0_G
PIO
4841
PH
1_G
PIO
4942
PE
0_G
PIO
2443
PE
1_G
PIO
2545
PH
4_G
PIO
5246
PH
5_G
PIO
5347
PF
4_G
PIO
3648
PG
0_G
PIO
4049
PG
1_G
PIO
4150
PF
5_G
PIO
3751
PJ6
_GP
IO62
53
PJ5
_GP
IO61
56
PJ4
_GP
IO60
57
PJ3
_GP
IO59
60
PJ2
_GP
IO58
61
PJ1
_GP
IO57
62
PJ0
_GP
IO56
63
PD
5_G
PIO
2164
PD
4_G
PIO
2065
PD
7_G
PIO
2368
PF
6_G
PIO
3869
PG
6_G
PIO
4670
PG
2_G
PIO
4271
PG
5_G
PIO
4572
PD
6_G
PIO
2273
PE
5_G
PIO
2976
PE
4_G
PIO
2877
PH
6_G
PIO
5479
PH
7_G
PIO
5580
PD
1_G
PIO
1798
PD
0_G
PIO
1610
2
PF
1_G
PIO
3310
3
PF
0_G
PIO
3210
4
GN
D
UP
1A
NA
UP
2A
NA
ET
HT
XD
3E
TH
TX
D2
ET
HT
XD
1E
TH
TX
D0
ET
HM
DC
ET
H_M
DIO
ET
HT
XE
N
ET
H_C
RS
ET
H_C
OL
ET
H_T
XC
LK
ET
H_R
XD
[3:0
]
ET
H_R
XE
R
ET
H_R
XC
LK
ET
H_R
XD
V
ET
H_I
NT
R
ET
HR
ST
US
B_D
MU
SB
_DP
EC
GS
CLK
EC
GM
OS
I
EC
G_M
ISO
100n
FC
98
U35
LT17
90-A
IS6-
3
64
VIN
VO
UT
L18
BLM
21A
G10
2SN
1DV
CC
1uF
C23
3 GN
D
C23
4
1uF
GN
D
VC
C_M
EA
SA
NA
3V3_
ME
AS
AN
A
A3V
3G
ND
Bed
_LE
D_R
1B
ed_L
ED
_Y1
Bed
_LE
D_G
1E
CG
_LE
D1
EC
G_L
ED
2B
ed_L
ED
_G2
Bed
_LE
D_Y
2B
ed_L
ED
_R2
US
BV
BU
S
100
R22
6
R22
7100 10
0R
228
TP
21
TP
11
TP
31
EC
G1C
S
EC
G1_
DR
DY
EC
G1R
ES
ET
EC
G1S
TA
RT
EC
G2R
ES
ET
EC
G2S
TA
RT
ET
HT
XD
3E
TH
TX
D2
ET
HT
XD
1E
TH
TX
D0
ET
HM
DC
ET
H_M
DIO
ET
HT
XE
N
ET
H_C
RS
ET
H_C
OL
ET
H_T
XC
LK
ET
H_R
XE
R
ET
H_R
XC
LKE
TH
_RX
DV
ET
H_I
NT
R
ET
HR
ST
ET
H_R
XD
3
ET
H_R
XD
2E
TH
_RX
D1
ET
H_R
XD
0
US
B_D
MU
SB
_DP
EC
GS
CLK
EC
GM
OS
IE
CG
_MIS
O
EC
G2C
SE
CG
2_D
RD
Y
CA
NR
X
CA
N_T
X
MC
AN
_TX
MC
AN
RX
VR
EF
_3V
VR
EF
_3V
VR
EF
_3V
Bed
_LE
D_R
1B
ed_L
ED
_Y1
Bed
_LE
D_G
1E
CG
_LE
D1
EC
G_L
ED
2B
ed_L
ED
_G2
Bed
_LE
D_Y
2B
ed_L
ED
_R2
US
BV
BU
S
U34
F28
M35
H52
C1R
FP
T
X2
95X
193
EM
U0
83
EM
U1
86
TR
ST
85
TM
S87
TD
I88
TD
O84
TC
K89
XR
Sn
4
AR
Sn
144
VS
SO
SC
94
XC
LKIN
97
XC
LKO
UT
_BO
OT
382
BO
OT
281
BO
OT
152
BO
OT
078
NC
91
FLT
116
FLT
221
U34
F28
M35
H52
C1R
FP
T
VR
EG
12E
N10
1
VR
EG
18E
N11
3
VD
D12
_11
11
VD
D12
_24
24
VD
D12
_55
55
VD
D12
_58
58
VD
D12
_66
66
VD
D12
_75
75
VD
D12
_90
90
VD
D12
_99
99
VD
D18
_11
VD
D18
_108
108
VD
DIO
_22
VD
DIO
_33
VD
DIO
_10
10
VD
DIO
_17
17
VD
DIO
_25
25
VD
DIO
_34
34
VD
DIO
_44
44
VD
DIO
_54
54
VD
DIO
_59
59
VD
DIO
_67
67
VD
DIO
_74
74
VD
DIO
_92
92
VD
DIO
_96
96
VD
DIO
_100
100
VD
DIO
_105
105
VD
DIO
_106
106
VD
DIO
_107
107
VS
S14
5
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
DS
P C
ore
2
DS
P M
ag 2
A1
Loca
ted
at s
heet
6 -
D1
Blo
ck D
SP
_Cor
e P
age
2 of
2S
heet
8 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
3V3
3V3
GN
D
3V3
JTA
G
HE
AD
ER
7X
2
J33
131197531
1412108642
C24
0
2.2u
F2.
2uF
C24
1
R20
02.
2K
GN
D
3V3
R20
1
2.2K
GN
D
GN
D
X2
20M
hzC
218
15pF
C21
915
pF
GN
DG
ND
GND
R22
14.
7KR
222
4.7K
GND
C24
4
470n
F47
0nF
C24
5C
246
470n
F47
0nF
C24
7C
248
470n
F47
0nF
C24
9
C25
0
470n
F
C25
1
470n
F
GN
D
C99
100n
F100n
FC
100
C10
110
0nF
100n
FC
102
C10
310
0nF10
0nF
C10
4C
105
100n
F
100n
FC
106
C10
710
0nF10
0nF
C10
8C
109
100n
F
100n
FC
110
C11
110
0nF
GN
D
C11
210
0nF
100n
FC
113
C11
410
0nF10
0nF
C11
5
GN
DG
ND
GN
DG
ND
GN
DG
ND
GN
DG
ND
GN
DG
ND
GN
DG
ND
GN
DG
ND
GN
DG
ND
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
3V3
GN
D
T1
T2
GN
DG
NDC25
410
0nF
GN
D
TD
O
TD
O
TD
I
TD
I
TM
S
TM
S
TC
K
TC
K
TR
ST
TR
ST
GN
D
X1
X1
X2
X2
EM
U0
EM
U0
EM
U1
EM
U1
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Pre
ssur
e S
enso
rs
Nyo
más
érz
ékel
ök
A1
Loca
ted
at s
heet
1 -
D2
Blo
ck P
ress
ure_
sens
or P
age
1 of
1S
heet
9 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
U25
SD
P11
08-R
2
3
1 VCC
OU
T
GND
dP
GN
D
L16
BLM
21A
G10
2SN
1DC
238
2.2u
FC
201
100n
F
GN
DG
ND
VC
C
UP
1
R18
520
K
UP
210
K
R18
7
VC
C
GN
DG
ND
GN
D
100n
FC
202
2.2u
FC
239B
LM21
AG
102S
N1D
L17
SD
P11
08-R
U26
2
3
1 VCC
OU
T
GND
dP
GN
D
C20
310
0nF
GN
D
GN
D
R18
8
10K
20K
R18
610
0nF
C20
4
GN
D
R99
100
100
R10
0U
P2
UP
1
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
TIT
LE
<CIM
>
A1
Loca
ted
at s
heet
1 -
A3
Blo
ck E
CG
_Inp
uts2
Pag
e 1
of 1
She
et 1
0 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
EC
G_C
2
EC
G_R
L2
EC
G_L
L2
EC
G_R
A2
EC
G_L
A2
ELE
C_R
A2
ELE
C_L
A2
47pF
C65
47pF
C66
GN
DIS
O2
MC
0801
0000
DS
610
K
R60
GN
DIS
O2
GN
DIS
O2
22K
R50
47pF
C67
47pF
C68
GN
DIS
O2
MC
0801
0000
DS
710
K
R61
GN
DIS
O2
GN
DIS
O2
22K
R51
R52
22K
GN
DIS
O2
GN
DIS
O2
R62
10K
DS
8
MC
0801
0000 G
ND
ISO
2
C69
47pF
C70
47pF
47pF
C71
47pF
C72
GN
DIS
O2
MC
0801
0000
DS
910
K
R63
GN
DIS
O2
GN
DIS
O2
22K
R53
R54
22K
GN
DIS
O2
GN
DIS
O2
R64
10K
DS
10
MC
0801
0000 G
ND
ISO
2
C73
47pF
C74
47pF
J6 J7 J8 J9 J10
ELE
C_R
L2
ELE
C_L
L2
ELE
C_R
A2
ELE
C_R
A2
ELE
C_L
A2
ELE
C_L
A2
EC
G_L
L2
EC
G_L
L2
EC
G_R
L2
EC
G_R
L2
EC
G_C
2
EC
G_C
2E
LEC
_C2
EC
G_R
A2
EC
G_R
A2
EC
G_L
A2
EC
G_L
A2
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
SA
Bed
CA
N In
terf
aces
Kis
álla
t ágy
CA
N in
terf
észe
k
A1
Loca
ted
at s
heet
1 -
F3
Blo
ck C
AN
_Int
f Pag
e 1
of 1
She
et 1
1 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
SD
1
BA
T75
4C
R14
451
K
R20
75.
1KR
208
5.1K
VC
CIS
O
GN
DIS
O
VC
CIS
O
Q1
2N39
04
GN
DIS
O
R20
936
0
U29
PS
7241
46
31
VC
CIS
O
VC
C1=
3V3
GN
D1=
GN
D V
CC
2=V
CC
ISO
GN
D2=
GN
DIS
O
U30
ISO
1050
7 623
TX
D
RX
DC
AN
L
CA
NH
R21
010
0
CA
NT
X
R10
1
100
CA
NR
XS
D2
NU
P21
05L
32
1
GN
DIS
O
F1
FU
SE
HO
LDE
R2
1
F2
FU
SE
HO
LDE
R2
1
VIN
VC
CIS
O C95
100n
F
GN
DIS
O
T5
GN
DIS
O
VC
C
RT
1
MIN
ISM
DC
020F
T
MIN
ISM
DC
020F
RT
2T
VC
C
GN
D100n
FC
252
3V3
SA
Bed
1 c
onne
ctor
J34
DS
UB
9 R
FE
MA
LE
1 2 3 4 56 7 8 9
GN
D_V
IN
GN
D
GN
D_V
IN
DS
UB
9 R
FE
MA
LE
J35
1 2 3 4 56 7 8 9
GN
DG
ND
ISO
GN
DIS
O
CA
NT
SW
2
GN
DIS
O
GN
DIS
O
CA
NT
SW
1
CA
NH
CA
NH
CA
NH
CA
NL
CA
NL
CA
NL
CA
NT
X
CA
NR
X
24V
PW
R1
24V
PW
R1
24V
PW
R2
24V
PW
R2
SA
Bed
1 c
onne
ctor
SA
Bed
2 c
onne
ctor
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
US
B In
terf
ace
US
B In
terf
ész
A1
Loca
ted
at s
heet
1 -
F3
Blo
ck U
SB
_Int
f Pag
e 1
of 1
She
et 1
2 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
SH
LD=G
ND
J31
US
B-B
90R
1 2 3D
AT
A-P
DA
TA
-M
VB
US
PR
TR
5V0U
2AX
D2
14
2
3
VC
C
GN
D
US
B_D
P
US
B_D
M
US
B_V
BU
SU
SB
_DM
US
B_D
P
US
B_V
BU
S
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Isol
ator
s
Gal
vani
kus
levá
lasz
tás
A1
Loca
ted
at s
heet
1 -
C3
Blo
ck Is
olat
ors
Pag
e 1
of 1
She
et 1
3 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
MIS
O
EC
G2_
ST
AR
T
MO
SI
SC
LK
DR
DY
VD
D1=
3V3
GN
D1=
GN
D V
DD
2=3V
3IS
O1
GN
D2=
GN
DIS
O1
U6
AD
uM24
01B
RW
Z
7 6
103 4 5
14 1113 12V
OC
VO
B
VID
VO
A
VIC
VIB
VIA
VE
2
VO
D
VE
1
EC
G2_
DR
DY
100
R65
ST
AR
T2
CS
R66
100
100
R67
ST
AR
T
R68
49.9
VD
D1=
3V3
GN
D1=
GN
D V
DD
2=3V
3IS
O1
GN
D2=
GN
DIS
O1
U5
AD
uM24
01B
RW
Z
7 6
103 4 5
14 1113 12V
OC
VO
B
VID
VO
A
VIC
VIB
VIA
VE
2
VO
D
VE
1
3V3I
SO
1
CLK
SC
LK2
MO
SI2
MIS
O2
3V3
3V3I
SO
1
R69
100 V
DD
1=3V
3 G
ND
1=G
ND
VD
D2=
3V3I
SO
2 G
ND
2=G
ND
ISO
2
U7
AD
uM24
01B
RW
Z
7 6
103 4 5
14 1113 12V
OC
VO
B
VID
VO
A
VIC
VIB
VIA
VE
2
VO
D
VE
1
100
R70
R71
100
49.9
R72
3V3
AD
uM24
01B
RW
Z
U8
VD
D1=
3V3
GN
D1=
GN
D V
DD
2=3V
3IS
O2
GN
D2=
GN
DIS
O2
7 6
103 4 5
14 1113 12V
OC
VO
B
VID
VO
A
VIC
VIB
VIA
VE
2
VO
D
VE
1
3V3I
SO
2
R73
100 10
0
R74
R75
100
49.9
R76
3V3
3V3 R
77
49.9
100
R78
R79
100
100
R80
3V3I
SO
2
C75
100n
F
3V3
GN
D
100n
FC
76
3V3I
SO
1
GN
DIS
O1
100n
FC
77
3V3I
SO
1
GN
DIS
O1
3V3
GN
D100n
FC
78
3V3
GN
D100n
FC
79
3V3
GN
DC80
100n
FC
8110
0nF
3V3I
SO
2
GN
DIS
O2
C82
100n
F
3V3I
SO
2
GN
DIS
O2
EC
G2_
MIS
O
CLK
2
CS
2
DR
DY
2
EC
G2_
CS
EC
G2_
SC
LK
EC
G2_
MO
SI
EC
G1_
ST
AR
T
EC
G1_
DR
DY
EC
G1_
CS
EC
G1_
MIS
O
EC
G1_
SC
LK
EC
G1_
MO
SI
RE
SE
TE
CG
1_R
ES
ET
RE
SE
T2
EC
G2_
RE
SE
T
V33
=3V
3
U9
2.04
8MH
z 3.
3V
4
1O
E
OU
T
100n
FC
83
GN
D
3V3
R81
75 75R
82
ST
AR
T2
SC
LK
MO
SI
MIS
O
SC
LK2
MO
SI2
MIS
O2
CLK
2
CS
2
DR
DY
2
DR
DY
CS
ST
AR
T
CLK
EC
G2_
ST
AR
T
EC
G2_
CLK
EC
G2_
DR
DY
EC
G2_
MIS
O
EC
G2_
CS
EC
G2_
SC
LK
EC
G2_
MO
SI
EC
G1_
ST
AR
T
EC
G1_
CLK
EC
G1_
DR
DY
EC
G1_
CS
EC
G1_
MIS
O
EC
G1_
SC
LK
EC
G1_
MO
SI
RE
SE
TE
CG
1_R
ES
ET
RE
SE
T2
EC
G2_
RE
SE
T
A1
Fö
CA
N in
terf
ész
Mai
n C
AN
Inte
rfac
e
<DA
TE
>
Cze
ller
Mik
lós
Haj
du C
s., T
örök
S.
Loca
ted
at s
heet
1 -
F3
Blo
ck C
AN
_Int
f_M
ain
Pag
e 1
of 1
She
et 1
4 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
SD
3N
UP
2105
L
32
1
GN
DIS
OM
J32
DS
UB
9 R
MA
LE
1 2 3 4 56 7 8 9
GN
DIS
OM
C96
100n
F
GN
DIS
OM
VC
CIS
OM
3V3
GN
DC25
310
0nF
T6
GN
DIS
OM
VC
C1=
3V3
GN
D1=
GN
D V
CC
2=V
CC
ISO
M G
ND
2=G
ND
ISO
M
ISO
1050
U31 TX
D
RX
DC
AN
L
CA
NH
120
CA
NT
X
R10
2
100
CA
NR
X
CA
NH
CA
NL
CA
NT
X
CA
NR
X
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Eth
erne
t Int
erfa
ce
Eth
erne
t int
erfé
sz
A1
Loca
ted
at s
heet
1 -
F4
Blo
ck E
ther
net_
Intf
Pag
e 1
of 1
She
et 1
5 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
U27
LAN
8710
A
33
3231 3029 28
27
2620 191815 13
12
111098 7
65 432
1
21 22232425 1417 16M
DIO
MD
C
CR
S
TX
D3
TX
D2
TX
D1
TX
D0
TX
EN
VDDA2
LED
2/nI
NT
SE
LLE
D1/
RE
GO
FF
XT
AL2
XT
AL1
/CLK
INVDDCR
RX
CLK
/PH
YA
D1
RX
D3/
PH
YA
D2
RX
D2/
RM
IISE
LR
XD
1/M
OD
E1
RX
D0/
MO
DE
0
VDDIO
RX
ER
/RX
D4/
PH
YA
D0
CO
L/C
RS
_DV
/MO
DE
2
nIN
T/T
XE
R/T
XD
4
nRS
T
TX
CLK
RX
DV
VDD1A
TX
NT
XP
RX
NR
XP
RB
IAS
VSSG
RO
UN
D
PO
WE
R
GN
D
R18
910
K
R19
12.
2K
3V3
3V375
R17
1R
172
7575R
173
R17
075
R17
475 75
R17
5R
176
75
R17
77575
R17
8
R20
2
12K
GN
D
X1
25M
Hz
X32
C24
2
33pF
C24
3
33pF
GN
D
R19
010
K
GN
D
R17
949
.9R
180
49.9
R18
149
.9R
182
49.9
R20
310
R20
410
C20
510
0nF
C20
610
0nF
U28
J001
1D01
BN
L
1413879101211653241T
XP
TX
CT
TX
N
RX
PR
XC
TR
XN
C_Y
LWA
_YLW
C_G
RN
A_G
RN
N.C
.C
HS
_GN
D3
CH
S_G
ND
2C
HS
_GN
D1
GN
DG
ND
L4 BLM
21P
G22
1SN
1D
L5 BLM
21P
G22
1SN
1D
3V3
2.2K
R19
2
3V3
3V3 C
161
10uF
C16
210
uFC
207
100n
F GN
DG
ND
100n
FC
208
C20
910
0nF
100n
FC
210
GN
DG
ND
GN
DG
ND
3V3
3V3
R20
533
0R
206
330
GN
D
GN
D
GN
D100n
FC
211
10uF
C16
3
GN
D
ET
H_T
XD
[3:0
]
R18
375 75
R18
4
2.2K
R19
3
3V3
3V3 R
194
2.2K
2.2K
R19
5
3V3
R19
62.
2K
GN
D
2.2K
R19
7
GN
DG
NDR
198
2.2K
2.2K
R19
9
GN
D
ET
H_T
XE
N
ET
H_R
ST
ET
H_T
XC
LK
ET
H_R
XD
[3:0
]
ET
H_C
OL
ET
H_C
RS
ET
H_R
XE
RE
TH
_RX
CLK
ET
H_R
XD
V
ET
H_I
NT
R
ET
H_M
DC
ET
H_M
DIO
EN
ET
_TX
PE
NE
T_T
XN
EN
ET
_RX
PE
NE
T_R
XN
EN
ET
_VD
DIO
EN
ET
_VD
DA
EN
ET
_VD
DA
EN
ET
_VD
DC
R
ET
H_T
XD
0E
TH
_TX
D1
ET
H_T
XD
2E
TH
_TX
D3
ETH_RXD0ET
H_R
XD
0
ETH_RXD1
ET
H_R
XD
1
ETH_RXD2
ET
H_R
XD
2
ETH_RXD3
ET
H_R
XD
3
ET
H_C
OL
ETH_COL
ET
H_C
RS
ET
H_R
XC
LK
ETH_RXCLK
ETH_RXERE
TH
_RX
ER
ET
H_R
XD
V
ET
H_I
NT
R
ET
H_R
ST
ET
H_T
XE
N
ET
H_M
DIO
ET
H_M
DC
ET
H_T
XC
LK
ET
H_T
XD
[3:0
]
ET
H_R
XD
[3:0
]
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
EC
G In
terf
ace
1
EK
G in
terf
ész
1
A1
Loca
ted
at s
heet
1 -
B4
Blo
ck E
CG
_Int
f Pag
e 1
of 1
She
et 1
6 O
F 1
8
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
GN
DIS
O1
GN
DIS
O1
10uF
C11
610
uFC
117
GN
DIS
O1
GN
DIS
O1
GN
DIS
O1
C13
11u
F
1uF
C13
2C13
31u
F
A3V
3IS
O1
C13
41u
FC
8410
0nF
GN
DIS
O1
GN
DIS
O1
GN
DIS
O1
100n
FC
853V
3IS
O1
GN
DIS
O1 1u
FC
135
10uF
C15
510
0nF
C86
GN
DIS
O1
1nF
C16
4
GN
DIS
O1
AD
S12
98R
IZX
G
U10
G8
E5
E6
E7
F5
G7
F7
F8
H8
E8
D6
B8
G6
H3H4
F6H7D7
C7
G5
A8
D5
G3
C5
B7
B5H6
A5H5D4E3F3
D8C8
A7C6B6A6B4A4C4A3B3C3
E4
G4
F4
A1
A2
B1
B2
C1
C2
D1
D2
E1
E2
F1
F2
G1
G2
H1
H2
D3
WC
TIN
1NIN
1PIN
2NIN
2PIN
3NIN
3PIN
4NIN
4PIN
5NIN
5PIN
6NIN
6PIN
7NIN
7PIN
8NIN
8PR
ES
P_M
OD
NR
ES
P_M
OD
PR
ES
V1
RLDINVRLDOUT
RLDINRLDREFAVDD_1AVDD_2AVDD_3AVDD_4AVDD_5AVDD1
DVDD1DVDD2
TESTN_PACE_OUT2TESTP_PACE_OUT1AVSS_1VCAP1AVSS_2
VCAP2AVSS_3
VCAP3
AVSS_4
VCAP4
AVSS_5
AVSS1
PWDN
DGND1
DGND2DGND3DAISY_IN
VREFNVREFP
RE
SE
T
CLK
SE
LD
RD
YD
OU
TD
INS
CLK CS
ST
AR
T
GP
IO1
GP
IO2
GP
IO3
GP
IO4
CLK
DR
DY
_O
DO
UT
_O
R10
310
K
3V3I
SO
1
R83
100
100
R84
DIN
SC
LK CS
ST
AR
T
CLK
RE
SE
T
GN
DIS
O1
R12
5
40.2
K
C16
8
2.2n
FR12
910
ME
G
2.2n
F
C16
9
10M
EG
R13
0
GN
DIS
O1
A3V
3IS
O1
A3V
3IS
O1
GN
DIS
O1
40.2
K
R12
6
2.2n
F
C17
0R13
110
ME
G
10M
EG
R13
2
C17
1
2.2n
F
C17
6
100n
F
C17
7
100n
F
ELE
C_L
A
ELE
C_R
A
C18
0
1.5n
F
R13
7
1ME
G
RL
RA LA LL
GN
DIS
O1
C10
KR
106
21
10K
R10
72
1R
108
10K
21
10K
R10
92
1R
110
10K
21
10K
R11
12
1R
112
10K
21
R11
310
K2
1
A3V
3IS
O1
VCAP1
WC
T
IN2N
IN2P
IN3P
IN4P
IN5N
IN5N
IN5P
IN5P
IN6N
IN6N
IN6P
IN6P
IN7N
IN7N
IN7P
IN7P
IN8N
IN8N
IN8P
IN8P
RE
SP
_MO
DN
RE
SP
_MO
DN
RE
SP
_MO
DP
RE
SP
_MO
DP
RLD
INV
RLDINV
RLD
OU
T
RLDIN
RLD
OU
T|R
LDIN
RLDOUT
3V3ISO1
PWDN
PWDN VR
EF
N
VREFP
RE
SE
T
RE
SE
T
DR
DY
DR
DY
DO
UT
DO
UT
DIN
DIN
SC
LK
SC
LK
CS
CS
GP
IO1
GP
IO2
GP
IO3
GP
IO4
CLK
CLK
DR
DY
_O
DO
UT
_O
IN1N
IN1N
IN1P
IN1P
ST
AR
T
ST
AR
T
<DA
TE
>
Cze
ller
Mik
lós
Haj
du C
s., T
örök
S.
EC
G In
terf
ace
2A
1E
KG
inte
rfés
z 2
Loca
ted
at s
heet
1 -
B3
Blo
ck E
CG
_Int
f2 P
age
1 of
1S
heet
17
OF
18
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
GN
DIS
O2
GN
DIS
O2
C11
810
uFC
119
10uF
GN
DIS
O2
GN
DIS
O2
GN
DIS
O2
1uF
C14
1 C14
21u
F
1uF
C14
3
A3V
3IS
O2
C14
41u
FC
8910
0nF
GN
DIS
O2
GN
DIS
O2
GN
DIS
O2
100n
FC
903V
3IS
O2
GN
DIS
O2
1uF
C14
5
10uF
C15
610
0nF
C91
GN
DIS
O2
1nF
C16
6
GN
DIS
O2
H3H4
F6H7D7
C7
G5
A8
D5
G3
C5
B7
B5H6
A5H5D4E3F3
D8C8
A7C6B6A6B4A4C4A3B3C3
E4
G4
F4
A1
A2
B1
B2
C1
C2
D1
D2
E1
E2
F1
F2
G1
G2
H1
H2
D3
AD
S12
98R
IZX
G
U11
G8
E5
E6
E7
F5
G7
F7
F8
H8
E8
D6
B8
G6
WC
TIN
1NIN
1PIN
2NIN
2PIN
3NIN
3PIN
4NIN
4PIN
5NIN
5PIN
6NIN
6PIN
7NIN
7PIN
8NIN
8PR
ES
P_M
OD
NR
ES
P_M
OD
PR
ES
V1
RLDINVRLDOUT
RLDINRLDREFAVDD_1AVDD_2AVDD_3AVDD_4AVDD_5AVDD1
DVDD1DVDD2
TESTN_PACE_OUT2TESTP_PACE_OUT1AVSS_1VCAP1AVSS_2
VCAP2AVSS_3
VCAP3
AVSS_4
VCAP4
AVSS_5
AVSS1
PWDN
DGND1
DGND2DGND3DAISY_IN
VREFNVREFP
RE
SE
T
CLK
SE
LD
RD
YD
OU
TD
INS
CLK CS
ST
AR
T
GP
IO1
GP
IO2
GP
IO3
GP
IO4
CLK
RE
SE
T
DR
DY
_O
DO
UT
_O
10K
R11
4
3V3I
SO
2
100
R85
R86
100
DIN
SC
LK CS
ST
AR
T
CLK
GN
DIS
O2
40.2
K
R12
7
2.2n
F
C17
2
10M
EG
R13
3
C17
3
2.2n
F R13
410
ME
G
GN
DIS
O2
A3V
3IS
O2
A3V
3IS
O2
GN
DIS
O2R
128
40.2
K
C17
4
2.2n
FR13
510
ME
G
R13
610
ME
G
2.2n
F
C17
5
100n
F
C17
8
100n
F
C17
9
ELE
C_L
A
ELE
C_R
A
1.5n
F
C18
1
1ME
G
R13
8
RL
RA LA LL C
R11
710
KR
118
10K
R11
910
K
R12
010
K2
1R
121
10K
21
R12
210
K2
110
KR
123
21
10K
R12
42
1
A3V
3IS
O2
GN
DIS
O2
VCAP1
WC
T
IN2N
IN2P
IN3P
IN4P
IN5N
IN5N
IN5P
IN5P
IN6N
IN6N
IN6P
IN6P
IN7N
IN7N
IN7P
IN7P
IN8N
IN8N
IN8P
IN8P
RE
SP
_MO
DN
RE
SP
_MO
DN
RE
SP
_MO
DP
RE
SP
_MO
DP
RLDINV
RLD
INV
RLDIN
RLD
OU
T
RLDOUT
RLD
OU
T|R
LDIN
3V3ISO2
PWDN
PWDN VR
EF
N
VREFP
RE
SE
T
RE
SE
T
DR
DY
DR
DY
DO
UT
DO
UT
DIN
DIN
SC
LK
SC
LK
CS
CS
GP
IO1
GP
IO2
GP
IO3
GP
IO4
CLK
CLK
DR
DY
_O
DO
UT
_O
IN1N
IN1N
IN1P
IN1P
ST
AR
T
ST
AR
T
Haj
du C
s., T
örök
S.
Cze
ller
Mik
lós
<DA
TE
>
Fee
dbac
k LE
Ds
Vis
szaj
elzö
LE
D-e
k
A1
Loca
ted
at s
heet
1 -
F2
Blo
ck F
eedb
ack_
LED
s P
age
1 of
1S
heet
18
OF
18
EL-
23-5
0-00
EC
G fo
r H
uman
& S
mal
l-Ani
mal
EK
G (
Hum
án é
s K
isál
lat)
EC
G
2012
-04-
11
AB
CD
EF
GH
1234
Con
trol
ler
<Elle
nõr>
:
Dat
e <D
átum
>:
Pag
e <O
ldal
>
Pag
e N
umbe
r <L
apsz
ám>:
Dra
win
g nu
mbe
r <R
ajzs
zám
>:
NA
ME
<N
év>:
Titl
e <C
im>:
Des
igne
r <T
erve
zõ>:
Dat
e <D
átum
>:M
Doc
. Rev
.
D3
LED
R/Y
/G 3
mm
HM
L
R21
233
0R
214
300
R21
627
0
VC
CV
CC
VC
C
U32
74H
CT
041213
74H
CT
04
U32
1011
U32
74H
CT
04
89
74H
CT
04
U33
1213
U33
74H
CT
04
89
74H
CT
04
U33
1011
VC
CV
CC
VC
C
270
R21
733
0R
213
300
R21
5 LED
R/Y
/G 3
mm
D4
HM
L
GN
D
VC
C 100n
FC
212
VC
C C21
310
0nF
GN
D
D5 LE
D D
3 G
RE
EN
270
R21
8
VC
C
74H
CT
04
U32
21
U33
74H
CT
04
65
VC
C R21
927
0 LED
D3
GR
EE
ND
6
VC
C R22
027
0 LED
D3
GR
EE
ND
7
GN
D
Bed
_LE
D_R
1
Bed
_LE
D_Y
1
Bed
_LE
D_G
1
EC
G_L
ED
1E
CG
_LE
D2
Bed
_LE
D_G
2
Bed
_LE
D_Y
2
Bed
_LE
D_R
2
74H
CT
04
U32
65
U32
74H
CT
04
43
U33
74H
CT
04
43
74H
CT
04
U33
21
Bed
_LE
D_R
1
Bed
_LE
D_Y
1
Bed
_LE
D_G
1
EC
G_L
ED
1E
CG
_LE
D2
Bed
_LE
D_G
2
Bed
_LE
D_Y
2
Bed
_LE
D_R
2
B I B L I O G R A P H Y
[1] R.E. Klabunde. Cardiovascular Physiology Concepts. WoltersKluwer Health, 2011. ISBN 9781451113846. (Cited on page 1.)
[2] B. Desjardins and E.A. Kazerooni. ECG-Gated Cardiac CT. Amer-ican Journal of Roentgenology, 182(4):993–1010, 2004. (Cited onpage 4.)
[3] A. Martinez-Moeller et al. Dual cardiac-respiratory gated PET:implementation and results from a feasibility study. EuropeanJournal of Nuclear Medicine and Molecular Imaging, 34:1447–1454,2007. ISSN 1619-7070. (Cited on page 4.)
[4] S. Skare and J.L.R. Andersson. On the effects of gating in diffu-sion imaging of the brain using single shot EPI. Magnetic Reso-nance Imaging, 19(8):1125 – 1128, 2001. ISSN 0730-725X. (Citedon page 4.)
[5] Monitoring Patients in the MRI Environment. http://www.
mrisafety.com/safety_article.asp?subject=40, . (Cited onpage 4.)
[6] MRI ECG Recording. http://www.biopac.com/
researchApplications.asp?Aid=46&AF=415&Level=3, . (Citedon page 4.)
[7] MRI monitor. http://www.gehealthcare.com/euen/patient_
monitoring/products/imm-monitoring/datex-ohmeda/mri-
monitor/index.html, . (Cited on page 4.)
[8] Advanced Clinical Solutions For MRI - Precess. http://www.
invivocorp.com/monitors/monitorinfo.php?id=1, . (Cited onpage 4.)
[9] MRI Small Animal Monitoring. http://www.biopac.com/
researchApplications.asp?Aid=46&AF=417&Level=3, . (Citedon page 4.)
[10] Model 1030 Monitoring & Gating System. http://www.i4sa.
com/web_app/main/defaultProduct.aspx?ID=76&PT=3, . (Citedon pages 4 and 33.)
[11] ECG Front-End Design is Simplified with MicroConverter.http://www.analog.com/library/analogDialogue/archives/
37-11/ecg.pdf. (Cited on page 8.)
59
bibliography 60
[12] Heart-Rate and EKG Monitor Using the MSP430FG439. http://www.ti.com/lit/an/slaa280a/slaa280a.pdf. (Cited on page 8.)
[13] ECG Implementation on the TMS320VC5505 DSP Medical De-velopment Kit (MDK). http://www.ti.com/lit/an/sprab36b/
sprab36b.pdf, . (Cited on pages 8, 12, and 13.)
[14] Competing ECG AFEs Reveal Chipmakers’ New BusinessParadigms. http://electronicdesign.com/article/analog-
and-mixed-signal/competing-ecg-afes-reveal-chipmakers-
new-business-paradigms. (Cited on page 8.)
[15] Low-Power, 8-Channel, 24-Bit Analog Front-End for Biopoten-tial Measurements - datasheet. http://www.ti.com/lit/ds/
symlink/ads1298r.pdf, . (Cited on pages 8, 9, 15, 24, and 25.)
[16] J Pan and W J Tompkins. A real-time QRS detection algorithm.IEEE Transactions on Biomedical Engineering, 32(3):230–236, March1985. (Cited on page 19.)
[17] ConcertoTM Dual-Subsystem MCU Series | 28x +ARM R© CortexTM-M3. http://www.ti.com/mcu/docs/
mcuproductcontentnp.tsp?sectionId=95&familyId=
2049&tabId=2743, . (Cited on page 21.)
[18] TEN 5WI Series, 6 Watt DC/DC converters - datasheet.http://www.tracopower.com/products/ten5wi.pdf, . (Cited onpage 23.)
[19] DC/DC-Converter 2 Watt DIP24 Miniature Single & Dual Out-put - datasheet. http://www.recom-international.com/pdf/
Econoline/RV.pdf. (Cited on page 23.)
[20] Ultralow-noise, high PSRR, fast RF 100-mA low-dropout linearregulators - datasheet. http://www.ti.com/lit/ds/slvs337b/
slvs337b.pdf, . (Cited on page 23.)
[21] TME Series, 1 Watt DC/DC converters - datasheet. http://www.
tracopower.com/products/tme.pdf, . (Cited on page 23.)
[22] Ultralow-noise, high PSRR, fast, RF, 1A low-dropout linearregulators - datasheet. http://www.ti.com/lit/ds/slvs351n/
slvs351n.pdf, . (Cited on page 23.)
[23] Quad-Channel Digital Isolators. http://www.analog.com/
static/imported-files/data_sheets/ADuM2400_2401_2402.
pdf. (Cited on page 24.)
[24] ADS1298R ECG Front-End Performance Demonstration KitUser’s Guide. http://www.ti.com/lit/ug/sbau181/sbau181.
pdf, . (Cited on page 24.)
bibliography 61
[25] multicomp Wire Ended T1 1/4 Neon lamps - datasheet. http://www.farnell.com/datasheets/320136.pdf. (Cited on page 24.)
[26] SDP1108/SDP2108 Low Differential Pressure Sensor with fastresponse time - datasheet. http://www.sensirion.com/en/pdf/
product_information/Datasheet_differential_pressure_
sensor_SDP1108_SDP2108.pdf. (Cited on page 25.)
[27] Concerto Microcontrollers - datasheet. http://www.ti.com/lit/ds/symlink/f28m35h52c.pdf. (Cited on page 26.)
[28] Concerto Control Card - F28M35xx - schematics. Avail-able through TI’s ControlSuite, http://www.ti.com/tool/
controlsuite, . (Cited on page 26.)
[29] Isolated CAN Transceiver - datasheet. http://www.ti.com/lit/ds/symlink/iso1050.pdf. (Cited on page 26.)
[30] Biology of the Rat. http://www.lvma.org/rat.html, . (Cited onpage 31.)
[31] Biology of the Hamster. http://www.lvma.org/hamster.html, .(Cited on page 33.)
[32] Westbrook, C. and Roth, C.K. and Talbot, J. MRI in Practice. JohnWiley & Sons, 2011. ISBN 9781118273869. (Cited on page 33.)