Dénes Faragó

Platform: NI sbRIO
Languages: G (LabVIEW)

The project aimed to develop firmware and PC software for a high-speed digital lock-in amplifier capable of operating simultaneously at two frequencies in the 400 kHz to 7 MHz range, with a sampling rate of 150 MHz. The hardware utilized was the NI sbRIO-9651 single-board system featuring an embedded FPGA and a processor core. My task involved designing the complete software architecture, comprising three main parts: the FPGA code, software for the embedded processor, and the PC GUI. The completed instrument is now integrated into a high-speed atomic force microscope, enhancing image and video acquisition speeds.

Corresponding scientific article:
Francesca Zuttion, Adai Colom, Prof. Stefan Matile, Denes Farago, Frédérique Pompeo, Janos Kokavecz, Anne Galinier, James N Sturgis & Ingacio Casuso "Mode of action of daptomycin: Evidence of new molecular mechanisms highlighted by high-speed atomic force microscopy," Nature Communications

Block diagram of the custom high-speed lock-in amplifier. My job was to develop the blue (digital) region. The hardware was made by Hinstra Instruments Ltd.

Graphical interface of the custom lock-in amplifier developed for high-speed atomic force microscopy

FPGA Code

• Developed using LabVIEW's FPGA Module and VHDL.
• Includes custom DDS modules, SPI engine, high-speed Fourier transforms, and digital filtering.
• Uses Xilinx IPs and fixed-point arithmetic.
• Operates at a clock frequency of 150 MHz.
• Communicates with multiple ICs and hardware components, including:
  • 3 DDS modules
  • 1 high-speed ADC with parallel DDR communication
  • 4 high-speed DACs
  • 4 serial DACs
  • Analog multiplexers and switches

Real-Time Host

• Embedded, real-time LabVIEW code running on the processor core inside the Xilinx SoC.
• Transfers commands and large data blocks between the PC and FPGA using DMA and TCP Streams in a state-machine setup.

PC Software

• Desktop application with a user-friendly windowed interface.
• Provides access to functions such as spectral analysis and setting parameters of excitation signals.
• Developed using the QMH (Queued Message Handler) design pattern.

Platform: NI compactRIO
Languages: G (LabVIEW)

With my former supervisor, we developed an efficient mixed-signal noise generation system using the well-known method of applying a filter bank to modify noise spectra. The filter bank parameters are precisely set by an optimization algorithm, achieving spectral shapes that would otherwise be unattainable. The parameters are sent to the FPGA, where the random number generation and the filtering take place. The platform is NI CompactRIO.

Corresponding conference article to read more:
R. Mingesz, D. Farago, "Implementing software defined noise generators" in R25th International Conference on Noise and Fluctuations ICNF 2019 Proceedings 18 - 21 June 2019 EPFL Neuchâtel campus - Neuchâtel, Switzerland

Differential evolution algorithm searching for the best possible filter arrangement to achieve a predefined, custom noise spectrum

Actigraphs are small devices that measure human activity by tracking wrist movements on the non-dominant hand. I was involved in developing custom actigraphs, and my role was to adapt the MCU code for different hardware versions, rewriting several existing functions of the firmware. I developed drivers for memory chips (including bad block management) and 6-axis accelerometers. Additionally, I optimized the entire MCU code for reliability and power efficiency, as the actigraphic measurements can take up to two or more weeks.

Additionally, I developed desktop applications to configure and start actigraphic measurements, read raw acceleration data from the actigraph after measurement, and visualize and calibrate the data and the accelerometer.

I have also worked on a calibration algorithm. Corresponding featured article:
D. Farago, B. Maczak and Z. Gingl, "Enhancing Accuracy in Actigraphic Measurements: A Lightweight Calibration Method for Triaxial Accelerometers," in IEEE Access, vol. 12, pp. 38102-38111, 2024, doi: 10.1109/ACCESS.2024.3374652.

Front panel of a LabVIEW program that tests the efficiency of various accelerometer calibration methods

Platform: NI sbRIO
Languages: G (LabVIEW)

End-to-end embedded system programming, similar to that of the AFM Controller project: FPGA code, Real-Time Code and PC GUI, testing and deployment.

Platform: Arduino, PC
Languages: Python (Kiwi framework) and C

The person being tested has to put their finger into a thimble-like holder that can move along a straight line. When they hear different sound effects, they need to move the thimble. After that, they have to say whether they moved the thimble themselves or if the machine moved it.

My task was to develop the firmware for the microcontroller (MCU) and create a desktop application. On the MCU, I handled tasks such as sound generation, reaction time measurement, and programming routines to control the solenoid responsible for moving the thimble. The main challenge was achieving precise timing, which couldn't be accomplished with basic Arduino libraries. Instead, I had to directly manipulate the Atmega328 registers. By utilizing timers and an external oscillator, I managed to generate sound with various frequencies, accurately measure reaction times, and control the solenoid using PWM. For the desktop software, I used the Kiwi framework.

GUI for the self-agency diagnostic equipment

Complex desktop and NI-RT control program for conducting electrochemical measurements. Some of the implemented features and essential elements include:

• - multi-window user interface


• - communication with
  • - high-current power supply
  • - humidifier
  • - gas analyzers
  • - relative temperature and humidity sensors
  • - several thermometers
  • - cooling-heating equipment
  • - flow regulators
  • - flow meters


• - communication protocols:
  • - MODBUS
  • - RS232
  • - analog interfaces (4-20 mA, 0-10 V, PWM)


• - startup, shutdown sequences
• - emergency functions
• - continuous logging
• - display of multiple graphs simultaneously
• - control of pumps and valves
• - temperature control with ON/OFF and PID controllers

Block-diagram of the Photuacoustic Controller

Platform: ATmega128

The custom, microcontroller based device receives string-based commands from a PC, triggering actions such as reading/controlling ADCs, DACs, digital thermometers, and solenoid driver circuits.

Communication protocols utilized in the project:

• RS232 (PC)
• I2C (DAC)
• SPI (ADC)
• 1-Wire (Dallas thermometer)

Platform: Raspberry PI

This project is special in the sense that it uses a Raspberry Pi single-board computer as its foundation, but its programming is done entirely in the C language instead of higher-level languages. The Raspberry Pi controls DDS ICs, ADCs and DACs, digital potentiometers, and relays from a custom PCB. Based on the initial instructions from the user, it runs complex, fully automated test sequences.