Kenneth John Webb

1
COMPARING EMISSION MEASUREMENTS IN A
REVERBERATION CHAMBER AND A SEMI-ANECHOIC CHAMBER

• By
• Kenneth John Webb
• Principal EMC Engineer

2
Agenda

• Purpose of Project
• Overview of Project
• Dipole Data
• Laptop Data
• Summary and Conclusions

3
Purpose of Project

• Validate Reverb Chamber (RC) Calibration
• To define a possible test method for performing
  emissions in a RC
• Comparing results in a RC to conventional test
  results in an Anechoic Chamber (AC)
• Determine if equation is valid
• Power density to V/m equation
• Use of CCF, ACF, IL, CLF defined later

4
Overview of Project

• Perform Calibration in RC
• Test a known source
• Test in RC and AC
• Calculate E-field using different techniques
• Validates method
• Test and unknown source
• Test in RC and AC
• Use methods derived on known source

5
Why Measure Emissions in a Reverb Chamber?

• New Robust Test Method
• Test all sides of the unit
• Cost of reverb chamber is less than conventional
  anechoic rooms
• Measure total fields emanating from unit
• Test systems without multiple antenna positions
  or unit orientations
• Less setup time from susceptibility testing to
  emissions testing

6
Emissions in Reverb Chamber
7
Reverb Chamber Physical Characteristics
8
Typical Tuner
9
Calibration Summary

• Meets the calibration requirements of DO-160D
  Change 1 from 400MHz to 18GHz.
• Uniformity is marginal from 100-200 MHz, the
  allowed standard deviation is acceptable. Above
  1 GHz, chamber uniformity is acceptable.
• Antenna vs probe exceeds the allowed /-3 dB
  tolerances using the log periodic antenna.
• Using the EMCO 3106 antenna from 400MHz to 2GHz
  allowed acceptable results
• Obtained calibration factors needed for emissions
  testing

10
Emission Comparison Methodology
11
Emission Comparison

• Use data collected in calibration for CCF, ACF,
  CLF ,and IL
• Use equation given in IEC 61000-4-21 for power
  density to E-field conversion
• Measure a known source (dipole antenna) and an
  unknown source (Laptop computer)
• Use basic test methods in AC
• Develop a new method for RC

12
Dipole Testing
13
Dipole Antenna

• Estimated 127dBuV/m with 20dBm
• Signal generator not linear, but used same one
  for both AC and RC tests
• AC will use conventional method
• Use AF
• Place RCV antenna in H and V polarities
• Dipole in H only
• RC new method
• No direct illumination

14
Dipole Antenna in AC
15
Dipole Antenna AC Data
16
Dipole Antenna in RC
17
Summary of Equations
18
Dipole Antenna RC Data

• ERadiated (V/m) is the estimated field strength
  produced by the dipole antenna
• R was assumed to be 1m since the final
  measurement is in volts per meter.
• D is the equivalent directivity of the dipole.
• Determining the correct value would be an
  interesting experiment. For the purposes of this
  paper, 1.7 was used.
• Power into antenna was not linear
• Same signal generator was used for both anechoic
  and reverb chamber data
• ?Tx, ?Rx the antenna efficiency factors for the
  transmit and receive antenna respectively
• Used 0.75 for a log periodic antenna and 0.9 for
  a horn antenna

19
Antenna Calibration Factor

• Equation 5.4-1
• Needed for emissions calculations for max power
  radiated
• Takes into account the antenna losses, gain, and
  efficiency

20
CCF

• CCF or Chamber Calibration Factor from eqn 5.7-1
• CCF is the normalized average received power
• PAveRec is the average received power over one
  tuner rotation
• PInput is the forward power averaged over one
  tuner rotation.
• Used emission field level calculations for
  average recevied power

21
ACF and CLF

• The chamber loading factor (CLF) is calculated
  using equation 5.7-2.
• CCF is from equation 5.7-1
• ACF is from equation 5.4-1
• Also used for emissions level calculations
• Used with Insertion Loss equation 8.4-1 for max
  received power

22
IL

• IL is from equation 5.4-1
• Also used for emissions level calculations
• Use with max radiated power
• IL is the normalized maximum received power
• Calculated during calibration

23
Dipole Antenna RC Data, No CCF applied
24
Dipole RC and AC DataNo CCF
25
Dipole RC Data Retest

• The retest data collected was for information
  only
• Performed manually using mode stirred approach
• Verify data collection techinique

26
Dipole RC Data Retest, No CCF applied, Mode
Stirred
27
RC Dipole Data with CCF or CLF/IL
28
RC Dipole Data with CCF, AVG to MAX
29
RC Dipole Data with CCFAC and RC Data
30
RC Dipole Data with CCFdB Delta
31
RC Dipole Data with CCFdB Delta
32
Dipole Summary

• Excellent correlation.
• 2dB delta between rooms for the dipole
  measurements can be considered validation of the
  test method.
• Use the RSS or MAX of the horizontal and vertical
  polarities
• Must use CCF for AVG Power
• Use CLF and IL for MAX power
• Retest RC data with mode stirring also has good
  correlation to AC
• Verifies mode stirring technique using equation
  for average power

33
Dipole Summary Tuned or Stirred?
34
Laptop Testing
35
Laptop Setup in AC
36
Laptop Setup in AC
37
Laptop Setup in AC
38
Laptop Setup in AC
39
Laptop Setup in RC
40
Laptop Setup in RC
41
Sample Laptop Data RC
42
Laptop Data in RC with CCF
43
RC Laptop Data CCF vs CLF
44
Ambient in RC with CCF
45
Laptop RC Data

• Ambient is a concern
• RS testing may be leaking RF
• 50MHz emissions for info only
• CCF vs CLF about 5 to 10dB different
• Used both to compare the data

46
Laptop Measurements in AC
47
Laptop Measurements in AC
48
Laptop Measurements in AC
49
Laptop Measurements in AC
50
Laptop Measurements in AC
51
Laptop Measurements in AC
52
Laptop Measurements in AC
53
Laptop Measurements in AC
54
Laptop Measurements in AC
55
Laptop Measurements in AC
56
Laptop Measurements in AC
57
Laptop Measurements in AC
58
AC vs RC CCF Laptop Data
59
RC CCF vs AC Data
60
RC CCF vs AC Data Delta
61
RC CCF vs AC Data Delta
62
RC CCF vs AC Data
63
RC CCF vs AC Data
64
RC CCF vs AC Data
65
RC CCF vs AC Data
66
AC vs RC CCF Laptop Data
67
RC CLF vs AC Data
68
RC CLF vs AC Data
69
RC CLF vs AC Data
70
RC CLF vs AC Data
71
RC CLF vs AC Data
72
Conclusions

• Laptop data not as good a correlation as the
  dipole data
• The maximum AC data was the best correlation to
  the RC CCF or CLF data for the Laptop
• The RSS or MAX was the best for the dipole
• Several spikes that were within 10-20dB of each
  other for the Laptop data
• The dwell time of 50ms may not have been long
  enough to capture the full amplitude.
• May be due to dwell time of RC data
• Tuner speed may need to be increased/decreased
• Use of MAX vs AVG equations for E-field

73
Conclusions

• Assumption that the Laptop rotation (in six
  different orientations) could be correlated to
  the reverb chamber data may be incorrect.
• A Laptop orientation (other than the 90 degree
  changes) may have a higher amplitude emissions.
• The frequency accuracy may also have been
  different between the two test methods.
• Frequencies may have been off
• The tuner may modulate the emissions and change
  the frequency slightly.

74
Conclusions

• The method for performing the test and the
  equation used to calculate the E-fields does
  appear to have an overall correlation and
  usefulness.
• Generally, the RC had a higher amplitude when
  using average power equation and mode-stirring
• Data trend was similar
• Dipole measurements within 2dB
• Future testing is definite
• Spherical dipole radiator
• Different dwell/sweep times

75
Questions??