ECT-ERT Hardware and Sensro -Lodz

7/30/2019 ECT-ERT Hardware and Sensro -Lodz

1/35

1

ECT-ERT hardware

and sensordr. Darius Styra22 months visit in TomoKIS

Computer engineering department

Technical University of Lodz


2/35

2

Content

1. Improvement of AC-Based tomography

hardware;2. Dual modality ECT-ERT sensor;

3. Compact size ECT hardware;

4. Rotatable field tomography;5. Tomogram quality estimation;

6. Conclusions;

7. Dissemination of experimental results.


3/35

3

Basics of ECT: system


4/35

4

Basics of ECT: sensor

I I

I I

Cx= 1 fF1 pF CACB= ~200 pF CA / Cx= ~2(102105)

A

A

C CB

BCx

3D ECT sensor

Equivalent circuit


5/35

5

Basics of ECT: hardware

Sex1

ELECTRODE

OUTCsw

Rf

Cf

3

2

6

+

-

Out

Sm1

Uex

Csw

Block diagram of ECT system

Picture of ECT hardwareInput circuit with switches


6/35

6

1. Improvement of ECT hardware

ECT hardware with 32 channels was designed by

Yangbo He in DENIDIA, there is still to improve:

Hardware connection to sensor;

Flexible sensitivity of input circuit; Algorithm for flexible calibration of hardware;

USB & Ethernet interfaces.


7/35

7

1.1 Hardware connection to sensor

Sex1

E

LE

CTR

ODE

OUTCsw

Rf

Cf

3

2

6

+

-

Out

Sm1

Uex

CswRf

Sm1

OUT

Uex

Sm3

Sm2

3

2

6

+

-

Out

Sex3Sex1

Sex2

ELE

CTR

ODE

Cf

Before: simple switch Improved: T-switch

Additional switches for electrode

connection to measurement circuit

prevent hardware from signal

saturation, which is accompanied by

a significant drop of the hardwaretemperature and therefore

minimization of power dissipation.

Signal decreased 1820 times at

500kHz.


8/35

8

1.2 Flexible sensitivity of input

circuit

3

2

6

+

-

Out

CswIN

Rf 1

Rf 2

Sf 1

Cf

OUT

Sf 1

Csw

OUT

Cf

CswRf 2

3

2

6

+

-

Out

Rf 1

IN

Sf 2

Csw

Sf 3

Before: simple switch Improved: T-switch

Flexible sensitivity for ECT ERT

switching.

Switchable frequency rangeincreased up to 14MHz.


9/35

9

1.3 Algorithm for flexible calibration

of hardware

High permittivity calibration

(DC PGA calculation)

Low permittivity calibration

(Offset calculation)

Calibration

End


(DC PGA calculation)

Low permittivity calibration

(Offset calculation)


(AC PGA calculation)

Calibration

End

Before: simple switch Improved: AC gainsSet DCPGA = 1;

Set Offset = 0;

Set ACPGA =1

Start

Fill sensor with high

permittivity mixture

Calculate AC gain

End

Read ADC

AC gains calibration


10/35

10

1.5 Ethernet and USB interfacesPowerful DAQ: USB & Ethernet

Prevented from signal saturation and high temperature;

Suitable for ECT and ERT measurements;

Suitable for any kind of 2D or 3D sensor;

USB & Ethernet hardware for rapid data transfer.


11/35

11

2 Dual modality ECT-ERT sensor

Two systems:

Oil and emulsion up to 40% of salt water: high resistivity, ECTmode;

Salt water and emulsion up to 50% of oil: low resistivity, ERT

mode.

Or one system:

Dual modality ECT-ERT electrodes;

Switchable ECT-ERT hardware.


12/35

12

2.1 Dual modality ECT/ERT sensor

Insulated electrode with naked pin


13/35

13

2.2 Dual modality ECT/ERT sensor

The same electrodes for ERT and ECT;

The same wires for ERT and ECT;

The same hardware for ERT and ECT.

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 200 400 600 800 1000 1200f, kHz

U, mVDM ERT ERT TW PIN ERT ERT TW

DM ECT ECT HP SHLD ECT ECT HP


14/35

14

3. Compact ECT hardware

Previous systems:

Big hardware, built in 19 Euro case;

A Lot of components;

High power consumptions;

Stray immune capacitance

measurement circuits.


15/35

15

Compact size; Low power;

Availability of supplied by battery

wireless hardware;

New stray capacitance compensation.

3.1 Hardware: a) pictures


16/35

16

3.1 Hardware: b) block diagram

PS021 from ACAM.de

measurement rate: up to

50 kHz;

resolution: up to 6 aF;

range: 0fF 10nF;

current: down to 10A;


17/35

17

1 2

1212 CC m =

3.2 Theory of stray capacitance

compensation


18/35

18

1 2

2313

23131212

CC

CCCC m +

+=

3


compensation


19/35

19

C12m C12

C13 C23

CAA+

C24

C23 C34

CAA

+

C14

C13 C34

CAA

+

C23 C34 C13 C34+ C34 C35+

CAA

C14+ C45+ C24++ +

C25

C23 C35

CAA

+

C24

C23 C34

CAA

+

C45

C34 C35

CAA

+

C23 C34 C13 C34+ C34 C35+

CAAC14+ C45+ C24+

+

C15

C13 C35

CAA

+

C14

C13 C34

CAA

+

C45

C34 C35

CAA

+

C23 C34 C13 C34+ C34 C35+

CAAC14+ C45+ C24+

+

C25 C15+C23 C35 C13 C35+

CAA

+

C23 C34 C13 C34+

CAA

C14+ C24+

C45

C34 C35

CAA

+

C23 C34 C13 C34+ C34 C35+

CAA

C14+ C45+ C24+

+

+

C AA C 13 C 23+ C 34+ C 35+


compensation: a) equation solving method


20/35

20

C12m C12C13C23

C13 C23++

C24C14

C24 C14++

C25C15

C25 C15++:=


compensation: b) neglecting method


21/35

21

Cm.i capacitance between i and allother short-circuited;

Cm.j capacitance betweenj and allother short-circuited;

Cm.i,j capacitance between connectedi with j and all other short-circuited.


compensation: c) short-circuiting method

=)(

,.

ik

kiim CC

= )(,.

jkkjjm CC

+=),(

,

),(

,,.

jik

kj

jik

kijim CCC

2

,...

,

jimjmim

ji

CCC

C

+

=


22/35

22


compensation: d) application for ECT sensor

143211413121. scswswswwm CCCCCCCCC +++++++=

243122423122. scswswswwm CCCCCCCCC +++++++=2143212423141312. scscswswwwm CCCCCCCCCCC +++++++++=

+

=

2

2 43211212

swswswsw CCCCC

C

++++

=

2

,...

,

swjimjmim

ji

CCCCC

+=


23/35

23


compensation: e) measurement speed

12

)1( ++= NNNM2/)1( = NNM

Stray capacitance immune Stray capacitance compensated

1.0671.1421.1971.2441.3211.467Difference D

52913779563722Stray compensated M

49612066452815Stray immune M

3216121086Electrode numberN

Single measurement number


24/35

24

3.3 Experimental results

a) comparison of methods

0.00%0.00%0.0000.0002.0702.2745. Isolated

15.1%12.71%0.3120.2892.3822.5634. Without compensation

0.10%0.70%0.002-0.0162.0722.2583. Short-circuiting

7.25%8.58%-0.15-0.1951.9202.0792. Neglecting

1.16%0.66%-0.024-0.0152.0462.2591. Equation solving

C45, %C12, %C45, pFC12, pF

Relative errorErrorC45, pFC12, pFMethod


25/35

25

3.4 Experimental results

b) stray capacitance compensation for ECTsensor

0.7600.23831.315Stray compensated1 72*, full1.590.494

1.0160.43330.821Stray immune1 72*, full

0.0140.3132249.867Stray compensated1 22*, full0.020.353

0.0250.5662249.514Stray immune1 22*, full

2.2680.0934.101Stray compensated1 71*, empty-1.38-0.057

4.0690.1694.158Stray immune1 71*, empty

0.0730.311424.062Stray compensated1 21*, empty0.000.004

0.1330.562424.058Stray immune1 21*, empty

diff*, %diff*, fFU*, %U*, fFC, fFMethodElectrodeSensor

U* - uncertainty type A, p=95%; diff* - difference between measurement results;

1* - sensor with 160mm diameter, electrode length 160mm and 25mm electrode width;

2* - sensor with 45mm diameter, electrode length 100mm and 9.5mm electrode width.


26/35

26

U1(t)

I1(t)

I3(t)

U3(t)U2(t)

I2(t

)

0 60 120 180 240 300 36010

5

0

5

1010

10

U 1 ( )

U 2 ( )

U 3 ( )

3600

deg

3 phase voltage

U1 ( ) Um cos 0deg+( ):= U2 ( ) Um cos 120deg+( ):= U3 ( ) Um cos 240deg+( ):=

4. Rotatable Field Tomography


27/35

27

4.1 Rotatable Field Tomography


28/35

28

4.2 HardWare for RFT

R4R

+

-

U10AD844

3

26

7

4

1

8

5

+Vs

C1C

+Vs

R3R

+Vs

-Vsin

-Vs

+

-

U8AD844

3

26

7

4

1

8

5

R2

Rvar

+Vs

-Iout

+

-

U5AD844

3

26

7

4

1

8

5

+

-

U7AD844

3

26

7

4

1

8

5

-Vs

+

-

U2AD844

3

26

7

4

1

8

5

+Vs

+Vs

+

-

U3AD844

3

26

7

4

1

8

5

-Vs

-VsC2C

-Vs

-Vs

+Iout

-Vs

R6R

+

-

U4AD844

3

26

7

4

1

8

5

+Vs

R1R

-Vs -Vs

R5R

+Vsin

+Vs+Vs

+

-

U9AD844

3

26

7

4

1

8

5

+Vs

+

-

U1OPA602

3

26

7 1

4 5

+

-

U6OPA602

3

26

7 1

4 5

-Vs

Block diagram of electronics

DDS BUF ADCAMP

DAQ

Ele1

DDS BUF ADCAMP

Ele2

DDS BUF ADCAMP

Ele3

Current excitation Voltage measurement

Schematics for current excitation

No need of switches

High quality tomogramHigh speed tomography


29/35

29

4.3 Rotatable Field Tomography

Only average value;

There is need of further research.


30/35

30

5. Tomogram quality estimationSigma L2normalized error is the root mean square of the components of the error

M, N number of point;

x exact value of point;

xTOM value of reconstructed point.

( )( )

%100

1 1

,

1 1 ,,

x

N

n

M

m

x

nm

x

N

n

M

m

x

nmTOMnm

Lx

x

xx

= =

= =

=

Cross - correlation is the root mean square of the components of the error

R1

N M 0

N 1

n 0

M 1

m

xn m,

yn m,

( )==

M, N number of point;

x exact value of point;y value of reconstructed point.


31/35

31

5. Tomogram quality estimationSftware for comparison tomogram done byIAESTEstudentDanijiel Petanovic

Software is able:

evaluate Sigma Lx erros between two

2D tomograms;

evaluate correlation between two 2D

tomograms;

enteramanuallt 2D tomogram;

Draw tomograms


32/35

32

6. ConclusionsImprovement of AC-based ECT hardware Prevented from signal saturation;

Suitable for ECT and ERT measurements;

Suitable for any kind of 2D or 3D;

Reduced noise;

USB & Ethernet hardware for rapid data transfer.Dual modality ECT-ERT sensor

The same electrodes for ERT and ECT;

The same wires for ERT and ECT;

The same hardware for ERT and ECT.

Compact ECT hardware New concept of ECT hardware

New stray capacitance compensation method;

Compact size;

Low price;

Availability of supplied by battery and/or wireless hardware.

Rotatable Field Tomography:

No success, only average value.

Error estimation in tomograms

Two tomograms comparison using signa L norms;

Two tomograms comparison using correlation;

Manual entering of tomogram


33/35

33

7. Dissemination of research results

Presentations: 6th World Congress on Industrial Process tomography in Beijing;

Electrical Engineering and Electronics 2010 in Kowno;

Marie Curie Conference ESOF 2010 in Torino;

Lecture in Technical University of Warszawa 2009.

Papers: Improvement of AC-based ECT Hardware;

Stray Capacitance Compensation for Non Stray-Immune ECT Systems;

Application of Multi Modality Tomography for Multi Phase Flow Measurements; Application of ECT/ERT/gamma ray image reconstruction for Multi Phase Flow

Measurements;

Improvement of ECT Hardware.

Few opinions: The papers are very good. Thanks for sending them to me.

I read your two papers carefully. My opinion is that the paper on improvement of ECThardware is very good, but the compact ECT paper is less impressive.


34/35

34

Acknowledgements

The author would like to thank prof. Erling A. Hammerfrom University of Bergen for valuable help.

The work is funded by the European Communitys

Sixth Framework Program Marie Curie Transfer ofKnowledge Action (DENIDIA, contract No.: MTKD-CT-2006-039546)


35/35

35

Thanks for your attention

ECT-ERT Hardware and Sensro -Lodz

Documents

Transcript of ECT-ERT Hardware and Sensro -Lodz