GTEM CABLE EMISSION STUDIES MEASUREMENT TECHNOLOGY LIMITED JUNE 29 TH 2010

GTEM CABLE EMISSION STUDIES

MEASUREMENT TECHNOLOGY LIMITED JUNE 29TH 2010

Dr. Zaid Muhi-Eldeen Al-DaherDr. Angela Nothofer

Prof. Christos Christopoulos

GTEM User Group Meeting - MTL Luton

Presentation Outline

•Repeatability Issues •Correlation Algorithm

•EUT (Antenna Gain)

•Simulation & Measurements Results

GTEM User Group Meeting - MTL Luton

Repeatability Issues: Input Reflections – S11

•Return loss due to reflections at the cell’s input port•Losses include: mismatches and discontinuities, path losses, current and RF termination characteristics•Reflection losses should be better than -20dB •Peaks occur at every λ/2 away from the characteristic frequency

0 200 400 600 800 1000 1200 1400 1600 1800 2000-55

-50

-45

-40

-35

-30

-25

-20

-15

-10

Freq (MHz)

S11 (

dB

)

GG

NPL

Termination cross over

Increasing reflections

GTEM User Group Meeting - MTL Luton

Repeatability Issues: Field Factor (Normalization) – GG Cell

0 200 400 600 800 1000 1200 1400 1600 1800 200017

18

19

20

21

22

23

24

Freq (MHz)

e0t

(dB

)

Measured

Calculated

Up to 5dB errors

GTEM User Group Meeting - MTL Luton

Repeatability Issues: Field Factor (Normalization) – NPL Cell

Up to 4.5dB errors

0 200 400 600 800 1000 1200 1400 1600 1800 20005

6

7

8

9

10

11

12

13

14

15

Freq (MHz)

e0y (

dB

)

Measured

Calculated

GTEM User Group Meeting - MTL Luton

Repeatability Issues: Field Factor (Normalization) – Both Cells

0 200 400 600 800 1000 1200 1400 1600 1800 20004

6

8

10

12

14

16

18

20

22

24

Freq (MHz)

e0y

(dB

)

NPL

GG

•Difference in performance behaviour.

•Different in induced field strengths – size factor.

•Difference in e0y mean values (10dB in this example) must be normalised further.

GTEM User Group Meeting - MTL Luton

Repeatability Issues: Field Factor – Both Cells Common Normalization

0 200 400 600 800 1000 1200 1400 1600 1800 200017

18

19

20

21

22

23

24

Freq (MHz)

e0y (

dB

)

GG

NPL

Common characteristic frequency

More ripples

GTEM User Group Meeting - MTL Luton

Repeatability Precautions

•Cables, wires, bundles, devices under test must be all positioned within the testing volume of the GTEM cell.

•The primary y-component of the electrical must be measured rather than calculated.

•Data must be normalised to a common value of e0y.

•Normalisation with respect to the input power if possible.

GTEM User Group Meeting - MTL Luton

Bundles

GTEM User Group Meeting - MTL Luton

Repeatability Measurements (Different wire lengths)

0 100 200 300 400 500 600 700 800 900 1000-60

-50

-40

-30

-20

-10

0Coil

F req (M Hz )

To

ta

l R

ad

iate

d P

ow

er (d

Bm

)

G G

NP L

0 100 200 300 400 500 600 700 800 900 1000-60

-50

-40

-30

-20

-10

0Revers e Twis ted

F req (M Hz )

To

tal

Ra

dia

ted

Po

we

r (d

Bm

)

G G

NP L100 200 300 400 500 600 700 800 900 1000

-60

-50

-40

-30

-20

-10

0In-Line Tw is ted

F req (M Hz )

To

tal

Ra

dia

ted

Po

we

r (d

Bm

)

G G

NP L

Coil

In-Line Twisted

Inverse Twisted

GTEM User Group Meeting - MTL Luton

Correlation AlgorithmTotal radiated power algorithm is based on a multi-dipole model positioned over a perfect conducting ground plane.

Assumptions:1- Any EUT may be represented by a set-up of three orthogonal dipoles 2- Dipoles are electrically short w.r.t wavelength. E.g. at 1GHz max dimension must not exceed 30cm3- Designed for frequency range of 300 to 1000MHz.4 - Any radiation elements must not have a gain greater than a dipole 5- The maximum power can be depicted from any direction

GTEM User Group Meeting - MTL Luton

Gains Using Antenna Theory

•By treating wires and bundles as potential radiating elements, studying their gains across the frequency range becomes vital.

•Strong directivities and high gains will invalidate the GTEM correlation algorithm.

•Measuring gains and antenna patterns in a GTEM cell is different to that of an anechoic chamber since there isn't really a transmitting antenna.

•By measuring the received power relative to that applied at the apex, and by accounting for any path and power losses, the gain produced by the EUT along with its cables can be determined as follows:

GTEM User Group Meeting - MTL Luton

Gains Using Antenna Thoery

eG

r AZ

EP

2

DAe 4

2

h

PZE inCG

Ginr PLPPhG 10101010 log10log10log20log208.19

A factor that compensates towards the different field strengths inside the GTEM cell.

GTEM User Group Meeting - MTL Luton

Gains – CST Simulations – Max values in 3D

0 200 400 600 800 1000 1200 1400 1600 1800 2000-25

-20

-15

-10

-5

0

5

10

15

Freq MHz

Gai

n (d

Bi)

35cm

CoilIn-Line Twisted

Reverse Twisted

GTEM User Group Meeting - MTL Luton

Gains – CST Simulations – Radiation Patterns

GTEM User Group Meeting - MTL Luton

Gains – CST Simulations – Radiation Patterns

GTEM User Group Meeting - MTL Luton

Gains – CST Simulations – Radiation Patterns

GTEM User Group Meeting - MTL Luton

Gains – CST Simulations – Radiation Patterns

GTEM User Group Meeting - MTL Luton

Gains – Measurements

Signal

Generator

Spectrum Analyser

GTEM User Group Meeting - MTL Luton

Gains – Measurements – Single Direction

0 200 400 600 800 1000 1200 1400 1600 1800 2000-60

-50

-40

-30

-20

-10

0

10

20

Freq (MHz)

Gai

n (d

Bi)

Coil Bundle

GG-0deg

NPL-0deg

GTEM User Group Meeting - MTL Luton

Cable/Wire Bundles & The Correlation Algorithm – Problems

The problem of consistency in cable positioning and bundling while using the three-position correlation algorithm for emission testing in GTEM cells seems to be unavoidable in achieving sustainable repeatable results for the following reasons:

•Cables cannot be simply held in one position without being rotated around the EUT’s axis.

•By rotating the cables around the EUT main axis, the shape and length of cables will change. Accordingly, the emission levels observed could incorrectly differ considerably at different orthogonal positions.

•In order to minimize repeatability errors cable movements must kept to a minimum.

GTEM User Group Meeting - MTL Luton

Cable/Wire Bundles & The Correlation Algorithm – Problems

•By averaging the received powers of the three orthogonal positions, emission levels could be well underestimated.

•Equipments with cables fail to meet the assumptions of the three-position correlation algorithm particularly at frequencies above 0.8GHz.

GTEM User Group Meeting - MTL Luton

Cable/Wire Bundles & The Correlation Algorithm – SolutionGoing with the viewpoint of maximizing emissions according to the standard recommendations; an alternative correlation algorithm could be arranged as follows:

•Instead of rotating the EUT and its cables in three orthogonal positions, doing a measurement in the direction of maximum emissions could be sufficient.

•This however will require a +/-180º azimuth scan inside the GTEM which could be conducted manually without the need of turntable, in steps of 30º as it is unlikely the EUT along with its cables will have a higher radiation pattern beamwidth than 30º.

•In this way, the movement of cables will be kept minimum and their effects will be low across azimuth measurements.

GTEM User Group Meeting - MTL Luton

Azimuth Measurements

0 200 400 600 800 1000 1200 1400 1600 1800 2000-80

-70

-60

-50

-40

-30

-20

Freq (MHz)

Em

issi

ons

(dB

m)

Coil

GG 0deg

NPL 0deg

Errors up to 10dB

GTEM User Group Meeting - MTL Luton

Azimuth Measurements

0 200 400 600 800 1000 1200 1400 1600 1800 2000-65

-60

-55

-50

-45

-40

-35

-30

-25

Freq (MHz)

Max

Em

issi

ons

(dB

m)

Coil

GG

NPL

Errors up to 6dB

GTEM User Group Meeting - MTL Luton

Thank You