Algorithms of data processing and controlling experimental equipment

1
Algorithms of data processing and controlling
experimental equipment

• Magnetic Resonance Spectroscopy
• Popov Timophey
• Komolkin Andrey, Sukharjevskiy Stanislav

2
Content

1. Principles of Magnetic Field Resonances
2. Radio Frequency Pulse Method
3. Continuous Wave Method
4. Real-time Operating Systems Review
5. Why QNX?
6. Current ESR-project

3
Magnetic Resonance Researches
Resonances
Nuclear Magnetic Resonance
Electron Spin Resonance
Methods
Continuous Wave Method
Radio Frequency Pulse Method
4
Nuclear Magnetic Resonance
Before and after energy absorption
S
high energy
low energy
B
B
hn
N
M
M

• Relaxation
• Spin-spin relaxation
• Spin-lattice relaxation

p
p / 2
M
M
M
M magnetization vector
5
Nuclear Magnetic Resonance
NMR condition

• Field-frequency correlation for 1H-nucleus
• Magnetic field 10 000 oersted
• Resonance frequency 42.5 MHz

6
Electron Spin Resonance
Electron Spin
Magnetic potential energy of electron spin in
magnetic field
Sz
Splitting
Energy
RF-induced transition
H0
Increasing magnetic field B
7
Electron Spin Resonance
Integral intensity proportional to quantity of
unpaired electrons in a sample. Width of spectral
line characteristic of RF-energy absorption
conditions. G-factor using to initialize
optional particles, participating in reactions
with free radicals.

• X-band spectrometers (wavelength 3 cm)
• Magnetic field 3 400 oersted
• Resonance frequency 9,5 GHz

• Q-band spectrometers (wavelength 8 mm)
• Magnetic field 15 500 oersted
• Resonance frequency 35 GHz

8
Continuous Wave Method
Sweep Coils The oscillation of sweeping magnetic
field must be far less then increasing of
external field
CW-specrometer
Sweep Generator
N
S
Altering Magnetic Field We have to solve
magnetic field scan linearization problem
Receiver Amplifier
RF Generator
9
Continuous Wave Method
1. First scan cycle linearization 2. Second
cycle Slow H-field scan with sweeping H1-field
and continuous transmission data to
computer. Major condition
First derivative
10
Continuous Wave Method
Double Integrating
Trend line
Evening-out
Fourier Transform
Noise reduction
Fast Fourier Transform
If we have 2k measurements
Method is usually used for wide spectral lines
and in ESR-spectrometry
11
RF-pulse Method
Interhardware communication We need guaranteed
fast communication between all nodes in this
experiment
RF-Pulse specrometer
Smart Pulse Generator
N
S

• Features
• Constant strong external magnetic field
• Free induction decay (FID) time 10 ms 1 s
• Short RF-pulses (10-100 ms)

Receiver Amplifier
12
RF-pulse Method Realization
t
Data processing
Data processing
Controlling
Smart Pulse Generator
Receiver Amplifier
ADC
Time to data processing and updating visual
information
Generating pulse sequence
FID-waiting Data buffering
!!! Other processes are sleeping !!!
13
RF-pulse Method features
Fourier Transform of nuclear echoes (FID)
FT

• Data Collecting
• Data Collecting
• Signal-to-noise ratio reduction
• Evening-out
• Real time visualization
• Characteristic decay time 10ms 1s

14
Real Time Review
Real Time Real Philosophy
15
Real-time Review

• Criteria Requirements
• Interrupt latency (less than 1 ms)
• Context-switch time
• System Size
• Rebooting time
• Development and execution division

Hard Real Time Systems any delays and
interrupts are not allowed on any conditions
(e.g. aircraft navigation system) Soft Real Time
Systems some delays are allowed, but it results
in increase production cost and decrease of
system efficiency as a whole (e.g. computer
network)
16
Real-time Review

• RT mechanisms
• Priority system
• Scheduling algorithms
• Interprocess communication (IPC)
• Operating with timers and interrupts

• RT-System classes
• Embedded systems (VXWorks, RTEMS)
• Real-time kernels (QNX, OS9)
• Real-time UNIXes (RTLinux, LynxOS)
• Real-time Windows (Windows Embedded)

17
Why
Fundamental Principles
System Users Processes

• microkernel architecture
• message-based interprocess communication

• Process manager
• Filesystem manager
• Device Manager
• Network Manager

Kernel Architecture
Device drivers

• message passing the Kernel handles the routing
  of all messages among all processes throughout
  the entire system
• scheduling the scheduler is a part of Kernel
  and is invoked whenever a process change state as
  the result of a message or interrupt

• choose to disappear at standard processes,
  simply becoming extensions to the system process
  theyre associated with
• retain their individual identity as standard
  process

18
Why
microkernel
Device manager
File system manager
Process manager
Network manager
Interprocess communication (IPC)
Single-computer model

• QNX is message-based OS
• Message a packet of bytes passed from one
  process to another
• All messages contains information about its
  state and priority, runtime information,
  synchronizing the execution and so on.

• Entire process and message space among all
  incorporated QNX-computers
• Sensible distribution network resources amount
  executing real-time processes

19
My Current ESR Project
Klystron (high radio frequency generator)
Wave conductor
S
Detector
Strong magnetic field scan
Receiver Amplifier
Resonator with sample
ADC
20
Credits References

• Komolkin Andrey V.
• Sukharjevskiy Stanislav M.
• Quantum Magnetic Phenomena Department of Physical
  Faculty SPbSU
• SWD Software
• http//hyperphysics.phy-astr.gsu.edu/hbase/molecul
  e/esr.html
• http//www.cem.msu.edu/reusch/VirtualText/Spectrp
  y/nmr/nmr1.htm