Susceptibility Testing - How much is Enough (Draft Version B)

rawfrogpondUrban and Civil

Nov 16, 2013 (3 years and 8 months ago)

87 views



1

DO
-
160

SECTION

20

RF SUSCEPTIBILITY TESTING


HOW MUCH IS ENOUGH?



Purpose and Scope


The purpose of this “white paper” is to explore RF susceptibility testing as specified by Section

20 of
DO
-
160
E

with respect to adequate step sizes.
While this paper is
not a proposal, it is intended to raise
the awareness of the step size issue

to the membership of RTCA SC
-
135

for consideration in
DO
-
160
F
.

The ultimate goal is

to eliminate the ambiguity now contained within the document
, and at
the same time establish a

test standard which allows for a high level of testing confidence.
The scope
of this paper is limited to Section

20, and the test requirements contained therein. Discussion of
radiated RF test methods
for EMI
is limited to anechoic chamber testing.

Rev
erberation chamber
testing is not
specifically
addressed, and is left for future discussions
.


Overview



Section I of this paper goes into detail and develops the rational for
the need for
more than
100

steps per decade when performing an EMI susceptibil
ity
(immunity)
test.


Section II of this paper discusses HIRF, and makes the argument that HIRF susceptibility
testing
and EMI susceptibility
/ immunity
testing are not necessarily the same, and that perhaps
different requirements exist
.


Section III
pr
ovided some considerations for Section

20

as to how HIRF and EMI
Susceptibility

(Immunity)
may be addressed

and economically tested.

NOTE: The term Q when used herein actually
refers to
the “Susceptibility Q,”

or n
S

which is in turn related to
the susceptibility

bandwidth
, referred to as SBW herein
.
Since
the SBW

increases

with increasing susceptibility
test levels, th
e

value of Q
de
crease with increasing test
levels.
The impl
ication is that when testing at HIRF
levels, fewer test steps, or faster sweeping is possible
when compared to traditional susceptibility / immunity
testing. Also n
ote that Q
S

is defined differently than the
traditional tuned circuit Q.



2

Section I EMI Susceptibility Testing to Ensure EMC on the
Aircraft


Definition of Terms


It is important that the terms us
ed within the context of this paper be defined and understood. A brief
definition of the
se

follows:


EMI



Electromagnetic Interference (EMI) within the context of this paper is considered to
be
caused by
onboard

electronics
,

including Passenger Electro
nic Devices (PEDs)
generating
electromagnetic energy which affects (produced susceptibility in) other
onboard

systems.
Typically, radio transmitters can affect other equipment, and other equipment can affect radio
receivers. (The notches in the
DO
-
160

Ra
diated Emission requirements are there primarily for
the protection of radio receivers,
while the lower level susceptibility requirements are there
primarily for the protection of other equipment from intended
onboard

radio transmissions.)


EM
C



Electroma
gnetic Compatibility is most easily defined as the absence of EMI.


HIRF


High Intensity Radiated Fields Susceptibility typically
results

when a high powered
transmitter external
to
the aircraft produces strong

electromagnetic

field
s

which surround the
a
ircraft.
Since such fields originate from outside the aircraft, aircraft shielding is an important
factor in reducing the effects of “HIRF.”




The “transition” point between aircraft internal EMI and external HIRF
has
not
been formally
defined in terms o
f field strength levels, but probably occurs somewhere between 20

V/m and
200

V/m.

Category R can be considered a transitional limit.


Undesired Response



An abnormal indication or output of
an
electronic equipment

caused by
the pres
ence of
“interfering”

RF voltages of currents. These RF voltages
or
currents m
ay

be
either radiated or conducted from their sources. An example may be the appearance of
additional
noise

at the output of a radio receiver,
the deviation of an indicator needle, or the
appearanc
e of distortion on a display, all caused by the presence of “
interfering
” RF voltages or
currents in or on the equipment and / or its wiring.


(As described below, not all undesired
responses are susceptibilities

or “failures.”
.)


Susceptibility



An
Undes
ired Response

which also meets a specific “criteria.” This “criteria
for susceptibility” is usually found in the equipment’s specification, but may also be found in
other documentation.
As an example, the accuracy
specification
of an indication is given
as
±

2

degrees
. Assume
the actual indication is without error under ambient
(no EMI)
conditions,
and
then
deviates +

1

degree

(an
U
ndesired
R
esponse
)

in the presence of
EMI.

T
his is
NOT

a
susceptibility because the indication is still within its
specifie
d
error tolerance. If, however, the
deviation caused by the
EMI

exceeded
±

2

degrees,
then
this would be considered
susceptibility. As a
second

example, a radio receiver is producing a 20

dB signal
-
plus
-
noise to
noise ratio ((S+N)/N) under ambient condit
ions. When the interfering RF is applied,
the (S+N)/N

degrade
s

to 11

dB. If the radio specification
is for a (S+N)/N equal to or greater than 10

dB,
then this
Undesired Response

is
NOT

a

Susceptibility
.


Note that all
susceptibilities

are
undesired res
ponses
, but NOT all
undesired
responses

are
susceptibilities
.


Guidance
for making this determination
is usually
found in the equipment specification, or regulatory documentation.




3

Susceptibility Bandwidth



The range of frequencies over which equipment is

continuously
susceptible.
Within this paper the

Susceptibility Bandwidth

is referred to
as “
S
BW
.”
(
The
SBW

would normally be

defined as
the 3

dB
bandwidth
of the response.

However, since 3

dB
data is not normally available, we are making a concession b
y including the entire observed
response bandwidth within this definition.
)
For example, if an equipment
exhibits

susceptibility
from 49

MHz

to 51

MHz
, and again from 99.5

MHz

to 100.5

MHz
, the bandwidths of
susceptibility are
2

MHz

at 50

MHz

(center freq
uency

of the range from 49

MHz to 51

MHz
)
, and
1

MHz

at 100

MHz

(center frequency

of the range from 99.5

MHz to 100.5

MHz
)
. (The center
frequency of each susceptible range is considered the
F
requency
of
S
usceptibility
,
refer
red

to
herein
as “
FoS
.”
)
While

the
SBW

is mostly independent of the equipment

type and design
,
certain circuit design
s

can result in extremely
small

SBWs
. Most commonly, equipment in this
category include
s

radio receivers which contain “tuned”
narrow bandwidth
circuits. Other types
o
f RF equipment, such as radars, DME

(
Distance Measuring Equipment
)
, etc. may also fit into
this category.



NOTE:
The

SBW

is an extremely important parameter when determining test frequency
sweep

rates and frequency
step

sizes
.
It will be shown

later th
at the

SBW

can broaden with the
application of higher test levels (HIRF), but
within this section
, our discussion is confined to
non
-
HIRF
onboard

EMI.


Sweep



Susceptibility Testing may be accomplished by applying a continuous varying
frequency (sweep)
over the frequency range of the test. Typically, sweeping is performed at a
constant sweep rate, usually expressed as frequency per
unit
time. Note that a constantly
changing frequency is more difficult to level, and normally, the specific (exact) test f
requency is
not known during the running of the test. Thus, if a susceptibility occurs, the test operator must

back
-
track” to
find
and verify it.


Step


Rather
than applying a continuous
ly

varying frequency

(which is hard to control)
,

discrete
frequency

steps may be applied
.
When stepping,
the test operator always knows
the
test frequency, and a step’s amplitude is much easier to control than a swept frequency.



Step Frequency



The test frequency actually applied during a stepped test. The
distance

b
etween
Step Frequencies

is the
Step Size
.

Each
Step Frequency

is
intended

to
span

a
range of frequencies defined by the
Step Size
. (For example, if there is a discrete test
frequency
(
S
tep
)
at 93

MHz, and the next
S
tep

is at 96

MHz, then the
Step Size

is

3

MHz,
while the step frequencies are 93

MHz and 96

MHz, respectively. The step at 93

MHz is
intended to span

±

half a
Step Size
, which is

the frequency range from 91.5

MHz to 94.5

MHz,
and the step at 96

MHz is intended to span
± half a
Step Size
, which

is the
frequency range
from 94.5

MHz to 97.5

MHz.
Note that each
intended

span
is equal in value to the
Step Size
,
and is centered on the

Step Frequency
.
)


IMPORTANT NOTE:

The frequency
span

actually

tested

by each
Step Frequency

is
not necessarily

th
e

Step Frequency

span as defined by
the
Step Size
, but rather is
a
direct function of the
SBW
.
Adequate testing within the
SBW

(that is, testing the entire
SBW
, rather than just a portion of the
SBW
)
requires that the
S
tep
S
ize

be ½ the
SBW
.
This ensures

that a step frequency is always within ±25% of the
SBW
’s

center
frequency

(
FoS
)
.

An advantage of stepping is that the test frequency is always known,
and discrete frequencies are easier to level

than continuously varying (or swept)
frequencies
.





4

An imp
ortant point to remember is that unless the
Step Size

is

½ of the

SBW
,
the
entire

Step

Frequency

S
pan

(that is, the frequency between the steps)
is NOT tested!


A
final
critically important point to remember is that
UNLESS

a frequency

Step

lands on
or

within ±25% of the
FoS

(
which is the center of the
SBW
),

th
e susceptibility is
NOT

tested

and will likely escape detection
.


SBW
s

can also
be
accurately
estimate
d

(calculate
d
)
. This estimation (calculation)
is described under the
Q

and
Q Values

definitions
(below), and also in “The Summary of Important Concepts

which follows.


Frequency Step Span



The range of frequencies between each
Step
. Mathematically, the
Frequency
Step
Span
is equal to the
Step Size
.


Dwell



During any susceptibility testing, it is important that the test frequency, whether stepped
or

swept, remain within the
SBW

long enough for the
E
quipment
U
nder
T
est

(
EUT
)
, its
monitoring
(support)
test equipment, and the test operator to respond.
A dwell time of
3

seconds is adequate for most situations, but if either the
EUT

response time, test s
upport
equipment, or operator response time is longer,
then
longer dwell times equal to or greater than
the longest of these response times
shall

be used.
As an example, if the only criteria for
susceptibility is bit error rate (BER), and the BER test tak
es 18 seconds, then an appropriate
dwell time would be a
t

20 seconds

for each frequency step. Note that this

is slightly longer than
the actual BER test

of 18

seconds
.
A case like this results in very long test durations, thus it is
prudent to find a “qu
ick” indication of susceptibility

(quick meaning less than 3 seconds to
detect
)
,
and if susceptibility is detected,
then
apply the full BER test at the
FoS

for validation.


Sweep Speed

(or
Sweep Rate
)


As in the case of step size,
sweep speed

MUST BE
a
function of
(related to)
the
SBW

if adequate test
ing

the entire swept frequency range

is desired
.
The swept test frequencies must remain within the
SBW

for the duration of the
Dwell

time.
For
example, if the
SBW

is 1

MHz
, and the required
Dwell

is 3

seco
nds, then the
Sweep Speed

is
1

MHz

every 3

seconds

or 333.33

kHz

per second.

Note: The
SBW

can be determined either
by measurement, or estimate

based on

the
Q

of the susceptibility response.

It is important to
note that
Q

as referenced in this paper is
always related to the bandwidth of the observed
susceptibility
response at the required test levels.


Q



The ratio
of the
FoS

to the
SBW

is expressed as
Q
.
Since Q as used in this paper is
ALWAYS

related to the susceptibility bandwidth at the applied R
F levels, we will refer to it as
Q
Susceptibility
, or simply Q
S
.

Mathematically,
Q
S

can be expressed
as
FoS
/
SBW
.
N
ote that
Q
S

must
increases as
FoS

increases,
and decrease as the

SBW

increas
es.


Since
the rate of
increase of
Q
S

is less than the rate of i
ncrease of
FoS
,
Q
S

is
useful for characterizing
susceptibilities

(
SBW
)

over
a
range of
frequenc
ies
.
One simplifying assumption
that

can
be
ma
d
e is to set
Q
S

to a constant over
a
specific
range of
test
frequencies
.
Although there are
some discontinuities
at the range ends, this technique simplifies calculations, and allows us to
use a single
quantity
of
logarithmically spaced
frequency steps per decade

over a relatively
wide frequency range. (Logarithmically spaced frequency steps are more efficient than
arithmetically space
d

steps
or analog sweeping
since the step size increases with frequency, in
the same manner that
SBW

increases with frequency.)


The important point here is that

Q
S

is nothing more than the ratio of

FoS

to

SBW
.

If
Q
S

is known

(
and
it

is,

see the
Q Table
), and the frequency of susceptibility (
FoS
)

is known,

then
SBW

can be calculated from
the relationship
:


SBW

=
FoS
/

Q
S



5



Q Values



Q values

have already been presented to both the RTCA
-
SC
-
135 Committee and
the Section

20 Working Group
.
These values have
also
been validated by an
ec
dotal
EMI
susceptibility

data

gathered from 2001 to date. These
Q values

are:


Q Table


Freq. Range

Q

30

Hz

-

1

MHz

10

1

MHz

-

30

MHz

50

30

MHz

-

1

GHz

100

1

GHz

-

8

GHz

500

8

GHz

-

40

GHz

1000


The
Q

values shown in this table are for relatively low levels of radiated susceptibility testing.
-
These levels may be characterized as the “day
-
to
-
day” electromagnetic environment (
EME
) and
includes
major
contributions
from internal aircraft avionics, aircraf
t radios, and Passenger
Electronic Devices (PED
s
).
However, the
Q

values in the table may be too high when
assessing susceptibility at HIRF levels, e.g., above
Category R or 2
0 to 200

V/m where the
external
EME

is the driver.


The
Q

values
in the table
above
may

be used
as
Q
S

to calculate
the
SBW
s

for immunity testing
for all DO
-
160 categories.

For
example,
at a
FoS

of 100

MHz
,

Q
,

(from the table above) = 100.
Since
Q
C

=
FoS
/
SBW
,
it
follows that

SBW

=

FoS
/
Q
S
.
Thus,
SBW

=

100

MHz
/100 = 1

MHz
.
Once
SBW

i
s

kno
w
n
,
proper step sizes

and sweep rates

can be determined
.

(Note, however,
that their use for HIRF testing may result in more test steps then necessary.)



EME



The Electromagnet
ic

Environment includes both internal and external radiated field
sourc
es.
While there is no set “boundary” between the normal
EME

and HIRF, it is generally
accepted to occur somewhere between 20

V/m and 200

V/m.

Category

R can be
considered

as
being at the transition.


Frequency Span Tested



This term should not be confus
ed with
Frequency Step Span
,
or
Step Size
. For the purposes of this paper,
the
Frequency Span Tested

is
a function of the
SBW

only, and represents the frequency range within the
SBW

which can be “tested” by a
single
Step
.

Mathematically, the
Frequency Sp
an Tested

for

one
Step

is equal to
SBW
/2.
Recall that for adequate (i.e., 100%) testing the
Step Size

must
be
½

the
SBW
.
Thus it follows
that the
Frequency Span Tested

by
one
Step

is equal to
SBW
/2.
The
percentage of the
Frequency Span Tested

is

a “
Figu
re of Merit
” for susceptibility testing
, and is defined as
follows:



(
Frequency Span Tested
)

/

(
Frequency

Step
Span
)

x

100%


IMPORTANT NOTE:

The
Frequency Span Tested

is a function of SBW, the
Susceptibility Bandwidth. The
Frequency

Step Span

is
equal t
o
the step size
actually

used during the test.



6

Example:
Assume that testing
is being performed
at 100

MHz, and that the step size

(used during the test)

is 2

MHz.

Thus, 2

MHz is the
Frequency

Step Span
.


At
100

MHz
,
Q

=

100 (see Q

Table), thus
SBW

= 1

MHz

(from the relationship,
SBW

=

FoS
/
Q
S
.
Since
SBW

is now known
,

it follows

that the

Frequency Span

Tested

by the frequency step at 100

MHz is
from 99.75

MHz to 100.25

MHz
(
from the
relationship
Frequency Span Tested

=
SBW
/2 = 1

MHz/2 = 500

kHz
)
.
Since
the
Frequency Step Span

=
Step Size

=
2

MHz,
and the

Frequency Span Tested


=

500

kHz, we can calculate the
percentage of the
Frequency Span

Tested

a
s equal
to:

(
Frequency Span Tested
)

/

(
Frequency Step Span
)

x

100
%

which
equals 500

kHz

/

2

MHz

x100% =
25
%
.


Thus,
for this example,
the

percentage of
the

Frequency Step Span

Tested

is
25
%

If
the
Step Size

is decreased
to
500

kHz
,
and
we
keep everything else the same,
then
Frequency Step Span

=
500

kHz
. The
Frequency Span

Tested

is still 500

kHz

since it
is
dependent only on the
SBW

and not on
Step Size
.

The percentage of the
Frequency Span

Tested

is now equal to

(
Frequency Span Tested
)

/

(
Frequency Step Span
)

x

100%

which
equals 500

kHz

/

500

kHz

x100% =
100
%.


Thus, the
Frequency Step Span

Tested

in
this

case is
100
%.






7

Summary
of Important Concepts


The relationship between
Step Size

(
Frequency Span
)

and
SBW

is illustrated below. Recall that

Adequate testing within the
SBW

(that is, testing the entire

SBW
, rather than just a portion of the
SBW
) requir
es that the
Step Size

(
Frequency Span
)

be ½ the
SBW
.”
In the best case, a step
frequency (
represented by the
green

v
ertical
l
ine
s
) falls exactly in the middle of the
SBW

(
blue
horizontal line
)

at “
FoS
” where
FoS

is the
(center)
frequency of susceptibility

as previously defined.”
Note that the two frequency steps immediately above and below the center “
FoS
” (in the middle of the
SBW
)

actually fall just
at the edges
of the
SBW

as shown in the left diagram (below). Thus, the
adjacent steps

(at the outside e
dges)

“miss” the
SBW
, but since the center step is at

the

FoS
, the
susceptibility is “caught.” In the worst case, where the steps completely miss the

center of the
susceptibility response,


FoS
,” there are
now
two steps
(as shown in the right diagram, bel
ow)
within
the
SBW
, and in
this
worst case, the

s
t
e
ps

are
located at approximately
FoS
±
25%.



Step Size
SBW
FoS
Step Size
SBW
FoS


Important Note: The step size must be half the
SBW

to

ensure the susceptibility is “caught” under all conditions
.



As a “mathema
tical” convenience, think in terms of one discrete
step

frequency as spanning an entire
step size
,
(
also referred to as

the
Frequency Span
)
,

but since two steps are needed to adequately test
a
SBW
,
it follows

that one
step

will actually test
½

SBW
.

This h
as been previously states as, “the
Frequency Span Tested

for one
Step

is equal to
SBW
/2.”






Important Note: The frequency “span” of a single step frequency is the step size.

Step Size
BW
F
Step Size
BW
F
Step Size
Step Size
Step Size
Frequency
Span
Step Frequency


8


It is also important to note that the “
Frequency
Step Span
” is a function of the

step size

ONLY
;

the
Frequency Span Tested

depends on the relationship between the

SBW

(a
function of the equipment
being tested,
and
the actual test frequency.


IF the
Step Size

(which is identical to the
Frequency Step Sp
an
) is
½

SBW
,

THEN 100% of
the frequency span (spectrum) is tested. On the other hand, if
½

SBW

is less than
the
step
size
, the difference between
½

SBW

an
d

the
step size

represents the portion of the spectrum
NOT being tested. Again, the
percentage of t
he
Frequency Span Tested

is the figure of merit
used to assess the “effectiveness” of the susceptibility test.




At this point,
Q

values
for non
-
HIRF testing
have been established (and validated by EMI test) for
general avionics equipment.
(This was prev
iously presented to the SC
-
135 committee.

The values
presented then, and used within this paper are available from public sources, and have been validated
by actual EMI test experience.
)
It has

also
been
shown that by knowing
Q
S
, an equipment’s
Susceptib
ility Bandwidth (
SBW
)

can be
determined
, and
from that
that one half the
SBW

is an
appropriate
Step Size

if the test goal is to determine whether or not the equipment is susceptible in an
EMI environment.

Using these parameters will result in a 100% test
of the equipment
, with the
exception of Radio receiving equipment
.

(Radio receiving equipment is addressed as an example in
Appendix

A.)


The next portion of this section discusses how this applies to DO
-
160E, and helps set some
groundwork for
considerati
on in
DO
-
160F.



9

APPLICATION TO DO
-
160
E



Now that the definitions
, concepts,

and some examples
are
behind us, let

s look at
what it takes to

establish
proper step sizes or sweep rates
according to the
stated intent

of

DO
-
160
E.
T
o do this,
t
he
appropriate p
ortions
of
DO
-
160E
Section

20
have been
extracted
for ready reference,

and appear
directly below
:



(
W
ithin this paper, the EUT susceptibility bandwidth is

called


SBW

)


From a “detection” perspectiv
e, a 3

second dwell may be more appropriate



Note that it has been shown to the SC
-
135 Committee that
the sweep
rate
described

above

is in

error
. The corrected statement should read,
“For test
equipment that generate a continu
ous frequency sweep, the sweep rate shall
be equal to the number of discrete frequencies per decade multiplied by twice
the dwell time, i.e., 100 discrete frequencies per decade times
2

times
1

second dwell time equals 200 seconds per decade sweep rate.”

The
reason for this is that a proper step size is half the
SBW
, or that it takes 2
steps to test a single
SBW
. When sweeping,
t
he interfering signal
must

remain within the
SBW

for the same length of time
period that

two steps are
within the
SBW
.

(This re
lation
ship has been
established

and illustrated in the
previous pages.)


It is from these paragraphs that the actual intent of DO
-
160 is stated. Note
that BOTH the step size and dwell time need to be justified.



10




Anecdotally
, it
is known that

at least some independent test facilities
may be

interpreting
DO
-
160
E as
ONLY
requiring 100 steps per decade (from the text directly above)
, and nothing more
.
At least one
independent test facility is known to be
performing
DO
-
160
E
at
o
nly the “minimum” number of step
s

per
decade, and ignoring the requirement
to select the number of steps per decade based on equipment
SBW
s
.

It will be shown that if aircraft EMC is a concern, then, in most instances, 100 steps per decade
are

not adequate
.


A
s an industry committee,
we need to ask

ourselves
, “
shall
100

steps per decade become “the”
Section

20 test
standard? Or,
shall
we base the number of steps per decade on “known,”
(
typical
)

equipment
susceptibility

bandwidths?


Should we differentiate

between relatively low level EMI/EMC
tests

(where Q
S

is high)
, and
high level
HIRF testing

(where Q
S

is low)
? Is
onboard

EMC

(including the
contribution from PEDs)
important, or is HIRF the over
-
riding concern?
Are the “success criteria” for
the low and

high (HIRF) levels the same? Are there different requirements at the component (module),
box, and system level?
In other words, “How much susceptibility testing
(and at what level)
is enough?
The answer to
each of
these
question
s

need
s

to be clearly st
ated in
DO
-
160

in order to avoid
any
“apparent ambiguity”
as now may
exist.


The next
portion of this
section
forms a foundation
for
onboard

EMI/EMC,
and provides
some of the
details

needed to
help us
answer
these
question
s
.
Section II

of this paper

takes

a brief look at HIRF
,
and Section III offers
”considerations”

for DO
-
160F.


It should be noted that the
SBW
s

based on
Q

values
ha
ve

been validated
(and previously presented to
the SC
-
135) for many different types of equipment. The data presented was
an
ec
dotal data
taken
during
the normal course
susceptibility evaluation
.
However,
the data was not collected specifically to
determine the 3

dB
SBW
,
and
thus
,

in
a

strict
and
rigorous

sense,
the
exact

3

dB
(
SBW
)
bandwidths



11

were not determined
.
However
, when
taken as a whole, th
is

data validates that the
Q

values

(or
bandwidths)
as
presented
in the table
are indeed
reasonable and
appear to be
correct.





12

A FIGURE of MERIT for Susceptibility Testing


If the
step size
, and the
SBW

are known
,
the

percentage of

the
Frequency Span Tested

can be
determined.

The
SBW

can be determined
from

Q

(see
Q

Table)

which was presented earlier under
Q

Values
. As an example, at 100

MHz
, Q

=

100, and from the relation
SBW

=

FoS
/
Q
S
,
it follows

that
SBW

=

1

MHz
. One hundred per
cent testing of this
SBW

(
which is
1

MHz
) requires a step size

of
SBW
/2,

or
0.5

MHz

(500

kHz
)
from a previous example
.


Thus

far,
it has been

shown

that to test 100% of the frequency spectrum near 100

MHz

requires
a step
size of 500

kHz

(0.5

MHz
).
How m
uch of the spectrum is tested when testing at 100 steps per decade?

DO
-
160
E provides
the following

formula for determining step frequencies
, from which step size

(when
performing the test at 100

steps per decade)

can be calculated
:





Applying this fo
rmula
,
and
starting at 100

MHz

as f
n
,

yields the
next

step

f
n+1

=

102.353

MHz
. The
step
size

is
equal to
102.353

MHz

minus

100

MHz

which is 2.353

MHz
.


Using the relationships previously established, t
he
frequency
step
span

of the 100

MHz

step
frequency

is 2.353

MHz
, and
from the
Q

relationship,
½

SBW

is 500

kHz

or 0.5

MHz
.
Dividing
½

SBW

(0.5

MHz
, which is the “
needed
” step size
) by the actual step size (2.353

MHz
) yields that 21.2%
as
the
percentage of the
Frequency Span Tested
. The remaining 78.8% o
f the
Frequency Range (
spectrum
)

is
untested
.


This can be further illustrated by the chart
s

which follow
:




13


Th
e

chart
above
illustrate
s
the full
SBW

in relation to two “
Steps
.” Since the
Step

frequency
completely misses the
SBW

(see chart

above
),
the susceptibility is

completely

“unde
te
cted.” (Recall
that as previously
described
,

a susceptibility can be detected ONLY when a frequency step falls at or
within
±
25% of the
SBW

center

frequency

(
FoS
).


So far, we have addressed s
tepping. Won’t sweeping catch these susceptibilities? Not likely because
of the extremely short time the swept frequency is actually within the
SBW
.

99.5
100
100.5
101
101.5
102
102.5
103
103.5
104
104.5
105
105.5
MHz
Frequency Step
Frequency Step
Frequency Step
The green horizontal line is the
SBW
or the
portion of the spectrum within which the unit
under test can respond.
The RED horizontal line is the portion of the spectrum
"spanned" during 100 steps per decade testing. (The
span of two steps is illustrated)
The difference in length (frequency)
between the
green
and
red
horizontal
lines represent the portion of the
frequency spectrum near 100 MHz
NOT tested when stepping at
100 steps per decade.
FoS


14

The chart below illustrates the
½
SBW

in relation to one “
Step
.” Since the
Step

frequency completely
mi
sses the
½
SBW

(see chart below), the susceptibility is “missed.” In addition, this chart has a yellow
horizontal line representing
the actual “
Frequency Span Tested
.”
Since the red horizontal line
indicates the
Frequency Step Span
, observe that only abou
t 20% of the
Frequency Step Span

is
actually tested

(and this is represented by the area where the red line and yellow line would overlap)
.
Approximately 80% of the
Frequency Step Span

(the portions of the red line not overlapped by the
yellow line)
is “u
ntested.”





Note that the SBW is completely missed by the frequency step. Sweeping will likely miss it as well
because of the short time the swept frequency is actually within the
SBW
.

98.5
99.0
99.5
100.0
100.5
101.0
101.5
MHz
Frequency Step
The green horizontal line is the ½
SBW
or the
portion of the spectrum within which the unit
under test can respond.
The RED horizontal line is the portion of the spectrum
"spanned" by the frequency step at 100 MHz. (The
span of only this one step is illustrated)
The difference in length (frequency)
between the
green
and
red
horizontal
lines represent the portion of the
frequency spectrum near 100 MHz
NOT tested when stepping at
100 steps per decade.
FoS
Note that unless a discrete
Step

Frequency
lands within ±25% of the
FoS
, the susceptibility in NOT detected. We can say that the
red
horizontal line
represents the intended “
Frequency Span
” (which depends on
step size
ONLY), and that the
green horizontal line
represents a susceptibility we intended to test. However, in order to test a susceptibility, a frequency
Step
must "land" at or within 25% of
the
FoS
. Since we know the
SBW
, the yellow horizontal line represents the portion of the “
Frequency Span
” actually tested.
Frequency Range Tested
(The portion of the
Frequency Span
actually tested.) Note: The
FoS
must lie anywhere on this line between the
yellow triangles.


15

To put this into yet another perspect
ive,
on
-
board radios can generate

relatively strong

(but not HIRF)

RF
levels
from 108

MHz

to 118

MHz
. The
DO
-
160
E minimum frequency steps (100 steps per decade)
within this range calculate to
109.750, 112.332, 114.976,
and
117.681

MHz
.


Looking only at on
e step
frequency at 114.976

MHz
, the approximate step size (from
114.976

MHz

to
the next step at
117.681

MHz
) is 2.705

MHz
, which is also the
Frequency Step Span
. The actual
SBW

of other
equipment within the aircraft
is approximately
1.15

MHz

(based on
Q

=

100 at
FoS

=

114.976

MHz
).
This one step effectively tests half the total
SBW
, or
1.15

MHz

/

2

=
0.5
75

MHz
. Dividing
0.5
75

MHz

(
½

SBW
) by the
frequency
span

of the test frequency
(
2.705

MHz
)
yields 21.2% of the spectrum
actually tested.


In this exam
ple, the
SBW

is

approximately 1.15

MHz
.
Once again, approximately 21.2% of the
spanned bandwidth is actually “tested
.



Looking only at the step at
114.976

MHz

(see chart below),
observe that

only

the frequency range from 114.692

MHz

to 115.260

MHz

(based

on
½
SBW
)

is actually
“tested.”
.
The range
spanned

by this step is from 113.685

MHz

to 116.329

MHz
, based on
Step Size
.
T
he
percentage of the
Frequency Span

Tested

can be calculated
as follows:


(115.260

MHz

-

114.692

MHz
) / (116.329

MHz

-

113.685

MHz
) x

100% = 21.5%.


The figure below illustrates the relationships between the frequency range
spanned
, and the portion of
the
Frequency Step Span

actually “
tested
.”




The point
? If
an
equipment

had a susceptibility a
t
, for exam
ple,
115.8

MHz, it (the susceptibility) would
go complete undetected by this test.

114.692
115.260
113.685
116.329
112
113
114
115
116
117
118
Frequency Range
"spanned" by a
single step
Portion of spectrum
tested by a single
step (Q based)
Portion of spectrum
NOT tested by a
single step
Portion of spectrum
NOT tested by a
single step
Step (100 steps per decade)
Step (100 steps per decade)
Step (100 steps per decade)


16

To save
us
from further
mathematical

ordeal, the percentages of the
Frequency Spans Tested

(spectrum tested)
when

using only the minimum steps per decade as stated in
DO
-
1
60
E has been
calculated and is

presented in graphical form below:




Observing this chart, one could infer that the most critical frequency range for an aircraft is the range
from 100

kHz

to 1

MHz

because this is the
only

frequ
ency range where 100% of the spectrum is
actually tested.
For all

other frequency ranges ONLY

2% to 43% of the spectrum

is
actually
tested
.


Percentage of Spectrum Tested by using the DO-160E
Section 20 minimum steps per decade
17.1%
100.0%
42.5%
21.2%
2.1%
4.2%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0.01
0.1
1
10
100
1000
10000
100000
MHz
Percent of Frequency Range "Tested"
10 steps / decade
100 steps / decade
100 steps / decade
100 steps / decade
100 steps / decade
MHz
Q
0.00003
10
1
50
30
100
1000
500
8000
1000
18000
1000
Assumed Q
The only region

achieving a “100%
test.”



17

How much of the spectrum
do

we
(as a committee)
want to see
test
ed
?
Is Aircraft EMC important?


To assist the co
mmittee in making this determination, the following charts illustrate the amount of
testing, in terms of steps per decade, required to test up to 100% of the spectrum.

Note that the chart
is for “immunity” testing only, and does not address “HIRF” testing
.


The table below summarizes current
DO
-
160
E minimum requirements, provides information on the
requirements for 100% testing

(of the
onboard

EMI environment)
, an
d

provide
s

a place for committee
recommendation.



Freq.
Range

Present
DO
-
160
E

Req’d

f潲 ㄰M

q敳e

C潭oitt敥 C潮獥s獵s

jez

p
p


p
p
o⨪

mpq⨪*

p
p


p
p
o⨪

p
p


p
p
o⨪

┠q敳t敤

〮〱〠


〮M





ㄷN










〮ㄠ
-

N

100

100

212%

49

49

--

--


1
-

30

100

147

42

233

343

--

--


30
-

1000

100

152

21.2

463

705

--

--


1000 8000

100

91

4.2

2305

2083

-
-

--


8000
-

18000

100

36

2.1

4608

1624

--

--


Totals


713



4853





*
S
p
D =
S
teps
p
er
(frequency)
D
ecade

** S
p
R =
S
teps
p
er
(frequency)
R
ange

*** PST =
P
ercent
(frequency)
S
pectrum
T
ested


NOTE: If the
SC
-
135
committee completes the
%

Tested

Co
lumn

(above, right)
, the
n the

S
p
D and
S
p
R can be calculated from this
value. The fourth column from the
lef
t, identified as “PST
,
” provides
the
P
ercent

(%)
S
pectrum
T
ested in the current
DO
-
160
E based on 10
steps per decade below
100

kHz
,
and 100 steps pe
r decade

above 100

kHz
.

A similar table can be used for HIRF test levels
once
Q
S

is established.


It is suggested that the committee give consideration to both “EMI

immunity

testing and “HIRF” testing.
The Steps per Decade required may not necessarily b
e the same for both. (This is developed later.)


Charts illustrating
PST, and SPD vs. Frequency follow.














18



Percent of Frequency Range (Spectrum) Tested vs. Steps per Decade
17%
39%
61%
82%
100%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0.01
0.1
1
Frequency Range, MHz
Percent Tested
10 Steps / decade
20 Steps / decade
30 Steps / decade
40 Steps / decade
49 Steps / decade
49 Steps / decade
40 Steps / decade
30 Steps / decade
20 Steps / decade
10 Steps / decade
100% testing of the 10 kHz to 1 MHz range requires a minimum of
49 steps per decade. Within this range (2 decades), there are 96
test steps.
The frequency range of 100 kHz to 1 MHz is the only range within DO-160 where more
than the theoretical minimum number of steps per decade is exceeded. The percent of
the frequency spectrum tested within this range approaches 212%. Below 100 kHz, 10
steps per decade is the minimum. These 10 steps are enough to "test" 17% of the
spectrum.


19



Percent of Frequency Range (Spectrum) Tested vs. Steps per Decade
42%
21%
64%
32%
86%
43%
100%
50%
65%
86%
100%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1
10
100
Frequency Range, MHz
Percent Tested
100 Steps / decade
150 Steps / decade
200 Steps / decade
233 Steps / decade
300 Steps / decade
400 Steps / decade
463 Steps / decade
100 Steps / decade
200 Steps / decade
233 Steps / decade
233 Steps / decade
200 Steps / decade
100 Steps / decade
300 Steps / decade
400 Steps / decade
463 Steps / decade
150 Steps / decade
150 Steps / decade
Note: The range below 1 MHz appears on the previous
chart, and above 100 MHz appears on the next chart.
100% testing of the 1 MHz to 30 MHz range requires a minimum of
233 steps per decade. Within this range (<2 decades), there are 343
test steps required.


20



Percent of Frequency Range (Spectrum) Tested vs. Steps per Decade
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
100
1000
Frequency Range, MHz
Percent Tested
100 Steps / decade
200 Steps / decade
233 Steps / decade
300 Steps / decade
400 Steps / decade
463 Steps / decade
Note: The range below 100 MHz appears on the previous
chart, and the range above 1 GHz appears on the next chart.
100% testing of the 30 MHz to 1 GHz range requires a minimum of
463 steps per decade. Within this range (<2 decades), there are 705
test steps required.
100 Steps / decade
200 Steps / decade
233 Steps / decade
300 Steps / decade
400 Steps / decade
463 Steps / decade


21



Percent of Frequency Range (Spectrum) Tested vs. Steps per Decade
4%
9%
13%
17%
20%
32%
43%
65%
82%
100%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1000
10000
Frequency Range, MHz
Percent Tested
100 Steps / decade
200 Steps / decade
300 Steps / decade
400 Steps / decade
463 Steps / decade
750 Steps / decade
1000 Steps / decade
1500 Steps / decade
1900 Steps / decade
2305 Steps / decade
3000 Steps / decade
4000 Steps / decade
4608 Steps / decade
Note: The range below 1000 MHz appears on the previous
chart, and the range above 10 GHz appears on the next chart.
100 Steps / decade
200 Steps / decade
300 Steps / decade
400 Steps / decade
500 Steps / decade
750 Steps / decade
1000 Steps / decade
1500 Steps / decade
2305 Steps / decade
1900 Steps / decade
100% testing of the 1 GHz to 8 GHz range requires a minimum of
2305 steps per decade. Within this range, there are actually 2082
test steps.


22




Percent of Frequency Range (Spectrum) Tested vs. Steps per Decade
2%
6%
10%
22%
41%
33%
50%
65%
87%
100%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
10000
100000
Frequency Range, MHz
Percent Tested
100 Steps / decade
300 Steps / decade
463 Steps / decade
1000 Steps / decade
1900 Steps / decade
1500 Steps / decade
2305 Steps / decade
3000 Steps / decade
4000 Steps / decade
4608 Steps / decade
4608 Steps / decade
4000 Steps / decade
3000 Steps / decade
2305 Steps / decade
1500 Steps / decade
1000 Steps / decade
463 Steps / decade
300 Steps / decade
100 Steps / decade
100% testing of the 8 GHz to 18 GHz range requires a minimum of
4608 steps per decade. Within this range, there are actually 1624
test steps.
Note: The range from 8 GHz to 10 GHz appears on the
previous chart.
1900 Steps / decade


23

Notes on High
-
Q
Circuits


Certain types of equipment, most notably radio receivers, contain high
-
Q
“tuned”
circuits which have
extremely narrow response bandwidths. As an example, a radio receiver operating between 100

MHz

and 200

MHz

could
have a 10

kHz

bandwidth. At
2
00

MHz
, the “Q” of this response is equal to


2
00

MHz

/

0.010

MHz

=

20,000


Applying

this Q
(20,000)
, and

starting at 100

MHz

will

need 46,054 steps per decade to search for and
find all potential susceptibilities (spurs). C
learly, this will take
too much

time
, but yet responses are
there which would be missed by testing

at less then 46,054 steps per decade.


Fortunately, these high
-
Q spurious
“tuned”
responses
can be
easily

calculated and found

The following
formula

accur
ately predicts their location assuming the radio’s IF
s

(Intermediate Frequencies)
and LOs
(Local Oscillators) or “injection frequencies”

are known.



F
spur

=
|
(p*LO
±
IF)/q
|


Where

F
spur

is the calculated spurious frequency

LO is the local oscillator freque
ncy

IF is the IF frequency,

p is an integer from 0 to


q is an integer from 1 to



This formula must be applied for each IF / LO combination, and also must be applied
if DSP (Digital
Signal Processing) is employed. For the DSP case, LO is the sample rate, IF is the digital IF, and q is a
fixed value of 1

(
DSP case only)
.


Each application of the formula results in two frequencies. As a general rule, q
may
be limited to values
of 5 or less, and p should be values which generate frequencies up to
at least
10 times the desired
receive frequency when q = 1, or

to
p


about 50,
or the highest test frequency is reached,
whichever
occurs first.


When testing High
-
Q
radio
equipment, the calculated spurious frequencies
shall
be tested
, and the
maximum step sizes not exceeded. If sweeping is performed, separate dwells wil
l be required for each
calculated spurious frequency
.




24

A
P
ossible

Requirement Presentation


The following format could be used to present the step and sweep requirements

for EMI testing
. While
equally spaced steps (on a logarithmic scale) are the norm, t
he steps need not be evenly spaced, but
shall be spaced such that the maximum step size is never exceeded. Thus, arithmetically spaced steps
are also acceptable.

(Note:
T
his table is based on a 3 second dwell. Other dwell times will change
the analog
rate.
)




The relationship between Steps per decade and the Maximum Step Size is given by the formula
:


Maximum Step Size = 10^(1/(
SpD
-
1))


1


where
SpD

= steps per decade.



The result is rounded to one significant digit, as
shown.


Arithmetic steps

If the test
frequencies are
arranged into frequency bands, and equally
(arithmetically)
space
d

steps are applied, then the lowest frequency (start frequency)
of the frequency band
determines
the step size for the entire band. For
example, in the ban
d

1 to 10

MHz
, the step size

for the
entire band would be
0.01 x f
0

=
0.01 x 1

MHz

=
1
0

kHz

(
0.
010

MHz
), and for the band from
10

MHz

to
30

MHz
, it would be
0.01 x f
0

=
0.01 x 10

MHz

=
100

kHz

(0.10

MHz
).


Logarithmic steps

Logarithmic s
teps may be determined by either of two formulae:

F
n+1

= F
n

x (Maximum Step Size + 1)

Where Maximum Step Size is from the table above


or

F
n+1

= F
n

x (10^(1/(SpD
-
1)))

where SpD is the steps per decade from the table above.

(Some insignificant rounding diff
erence may occur between the two formulae.)


NOTES:

It in not intended that an entire MOPS test be performed at each frequency step.

When using these sweeps / steps, the so called “critical frequencies” are included, so
separate “dwells” are not required.

Radio receivers
(swept testing)
are the exception
.

Analog Scans
Stepped Scans
Frequency Range
Maximum Scan Rates
Maximum Step Size
10 Hz - 1 MHz
0.0333 fo/sec
0.05 fo
49
237
6588
1 MHz - 30 MHz
0.00667 fo/sec
0.01 fo
233
343
30 MHz - 1 GHz
0.00333 fo/sec
0.005 fo
463
705
1 GHz - 8 GHz
0.000667 fo/sec
0.001 fo
2305
2082
8 GHz - 40 GHz
0.000333 fo/sec
0.0005 fo
4608
3221
Min. # Steps
in range
Tot. Min. #
Steps
Steps /
decade


25

Section II
HIRF

(High Intensity Radiated Fields)


DO
-
160

Section

20 has become extremely complicated and convoluted
(my opinion)
with numerous
category
limits and
test
techniques because it is attempting

to address both the EMI (internal)
environment and the HIRF (external) environment

simultaneously.

Let’s review the differences
in how
equipment may respond
between the
low level
EMI
and HIRF
environments.


Nearly a
ll aircraft have some sort of electroma
gnetic environment within which
most

equipment is
expected to operate
(and behave)
normally. This environment
i
s usually produced by “
onboard

equipment such as communications radios, displays,
flight control systems,

entertainment systems
,

PEDs,

etc
.

Si
nce these systems are normally “turned
-
on” for long periods of time, the “EMI” produced
by
these systems is virtually constant.
Within the context of this environment, extraneous “EMI” noises
on an aircraft radio receiver or “swimming” displays are not no
rmally tolerated.
However, since HIRF is
an “encounter,”

with an external electromagnetic field, the effects of HIRF are
generally
short
-
lived.
Thus,
radio noise or blockage of a band, or even a blanked display or autopilot disengagement may be
tolerated
,

as long as they recover once the encounter is over, and the aircraft can continue safe flight
and landing.
Note that the Section

20 “success criteria” may be different as cited in these examples.
Can we

(as an industry committee)

then say that the prim
ary concern of an EMI test
is
for
EMC
(ElectroMagnetic Compatibility)
between
all equipment
onboard

the aircraft, and the primary concern in
the HIRF environment is aircraft safety
?

If so, the criteria for “test success” would be different
between

these t
wo environments
, i
.
e.,
continued
normal
performance
vs.

aircraft
safety.


The HIRF environment is generally at least an order of magnitude (20

dB) or more, greater than the
normal aircraft onboard “EMI” environment.
At these high levels, equipment may
exh
ibit

susceptible
over a larger frequency range.
There is some limited
HIRF
data (not presented herein) that suggests
HIRF

Q
S

of 20 to 40

near 1

GHz
.


To illustrate this, let’s use a
n aircraft receiving equipment

example

based on DO
-
196
. An aircraft is
immersed in a
HIRF en
v
ironment of 200

V/m
at a frequency within the commercial FM band
(88

-

108

MHz
). The aircraft also has an
external
antenna which responds to frequencies in the range
of 108

MHz

to 117.95

MHz

and
which is connected to a receiver meeti
ng the requirements of DO
-
196.
For the purposes of this example, assume the antenna has a gain of 0

dB within the
commercial
FM
band. It can be shown that at 88

MHz
, th
is

antenna will receive and induce into the radio receiver’s
front
-
end nearly 100

Watt
s

(
60

dBm)
! DO
-
196 paragraph 2.2.8 states that the radio
receiver’s front
-
end

can tolerate
only
15

dBm (0.03

Watt) without d
esensititization
. The 100

W induced by the HIRF field
via the antenna into the radio
greatly
exceeds the d
esensitization

requireme
nt
(0.03

Watt)
by a factor of
approximately 3000! The radio WILL be d
esensitiz
ed (not able to communicate) by
the induced
HIRF.
At 107.9

MHz
, it can be shown that 200

V/m will induce a power level of 65

Watts

(48

dBm)
, while the
d
esensitization

requireme
nt is
-
10

dBm (0.0001

Watt). Thus, the
65

W
signal
induced by the HIRF field
via the antenna into the radio exceeds the d
esensitization

requirement
(0.0001

Watt)
by a factor of
approximately 650,000! Again, the radio WILL be d
esensitiz
ed (not able to com
municate) by HIRF. At
these extreme levels, it probably doesn’t matter what the exact frequency of the HIRF is. Whether it’s
at 88.5

MHz
, 96.5

MHz
, or 104.7

MHz
, it probably doesn’t matter
; the effects
(desensitization over a
broad range of frequencies)
are the same
. The strength of the signal is so overwhelming (34

to 58

dB
above the
DO
-
196
requirement) that
radio
reception will be blocked over
most of
if not all of the entire
band. Another way of looking at the overwhelming strength of HIRF is to note

that 200

V/m is 1
0

million
times radio sensitivity which is 20

microvolts

(DO
-
196)
. The radio cannot work in this environment, so
the criteria for HIRF must be, in this case, “no permanent damage
,


rather than “play through.”




26

The practical application h
ere is that the HIRF test must be treated differently than the EMI

susceptibility

test

because

the “criteria for success” is different.

The question answer
ed

by performing
an EMI test is whether or not the equipment will work
(i.e.,
be immune to, or
achie
ve EMC)
in the
specified
aircraft
onboard

Electromagnetic
environment. For HIRF, the primary question becomes, “will
the aircraft remain safe, and
be able to
continue safe flight and landing?” Since the HIRF levels are so
strong, some systems may not ope
rate in this environment. This may be acceptabl
e
, especially if
the
function

has a working backup, or
automatically resumes

after the HIRF encounter
, and the aircraft is
able to continue its flight and land safely
.


The “HIRF bandwidth” may also be much

wider that the “EMI bandwidth,”

as

illustrated below,
which
suggests that fewer steps per decade are
appropriate
to adequately test

the HIRF environment
.








SBW
HIRF BW
Bandwidth

Response



27

Illustration of the “HIRF Bandwidth (BW)



Committee Questions, “
Does data exist to support a wider “HIRF bandwidth,


and hence a lower Q
S
?”
If so, what values should we

(as a committee)

use?


Some data has been informally presented
suggesting Q
S

between 20 and 40 near 1

GHz
. A sample table has been generated using Q
S

= 50
above 1

GHz and 25 below 1

GHz.


“Shouldn’t we treat EMI Susceptibility (
onboard

EM environment compatibility) differently from

the

HIRF, High Intensity external environment

IF

the success or acceptance
criteria are

different
?”



A sample Q table fo
r HIRF testing is presented in the next section based on observed Qs near 20 to 40
in the vicinity of 1

GHz



28

Section III
Considerations
for
DO
-
160
F, Section

20


The following suggestions are based on personal opinion

based on observation

and should not be
construed as proposal
s. (S
imilar proposal
s

ha
ve

already been rejected by the committee.)

I hope this
white paper will communicate why these proposals were
originally
made, and then possibly revisit
them.

If nothing else, the existing intent of Section 2
0 needs to be clarified.


EMI and HIRF, although both are caused by electromagnetic fields,
and both are a part of the total
EME,
and
may

appear as
two different and distinct phenomena because of
equipment responses to
the
extreme differences in
amplitude
level
s
.

A
cceptable system performance
may
be
different in the HIRF
environment than it is in the non
-
HIRF environment.
It is therefore suggested that we
,

as a committee
,

consider the
separat
ion of


HIRF requirements


from

EMI requirements
” based
solely
on equipment
performance criteria

within th
e total EME

environment.
No separation is needed as long as the
equipment acceptance
criterion is

the same for both the low level EME and the HIRF EME.


For example, if a piece of equipment is expected to meet al
l requirements and to perform without
interruption throughout exposure to the entire EME spectrum, including HIRF levels, then no “division”
between HIRF and non
-
HIRF
testing
is required.
Testing is only needed at the HIRF levels, and may
proceed

using a
reduced number of steps per decade. Likewise, if the equipment has no requirement
other than to display no false or misleading data, remain safe, and recover after exposure to EME (both
HIRF and non
-
HIRF levels, then again, no “division” between HIRF and
non
-
HIRF testing is required.


On the other hand, if a piece of equipment has differing requirements depending on EME levels,
e.g., it
must remain fully functional and comply with all requirements when exposed to low level
(non
-
HIRF)
EME

(which includes
PEDs radiation)
,
while a
t HIRF levels, the equipment may
exhibit undesired
responses
, yet this may be acceptable as long as the equipment remains safe, there is no false or
misleading data
, the equipment recovers after exposure
, and the aircraft can contin
ue safe flight and
landing
.




The KEY to how much testing is enough depends on the equipment’s performance requirements
E慣捥灴慮捥⁣rit敲楡F
wit桩渠楴猠bjbK A 獥捯湤慲a 捯c獩摥rati潮 i猠t桥 捯cfi摥湣n l敶敬 t桡t t桥 t敳瑩湧
i猠慤敱畡t攠
for

fi湤
ing

p潴o
湴楡l imm畮ity L 獵獣s灴楢ility i獳略献



Alt桯畧h 扯th 愠桩g栠汥v敬 bjb EiK攮e efo䘩c慮d 愠汯w lev敬 bjb 數e獴I 慮 慩r捲aft 潰敲et敳 m潳t of t桥
tim攠楮 t桥 l潷 lev敬 bjbK q桥 l潷 l敶el bjb 慬獯⁩s捬畤敳⁣潮tri扵ti潮s fr潭 mba献 bq畩灭敮t
灥rform慮捥
req畩r敭敮t猠m慹 摩ffer 扥tw敥渠nh攠䕍b 數tr敭敳e慳 w敬lK ft i猠獵杧敳t敤 t桡t w桥渠
灥rform慮捥 req畩r敭敮t猠摩ff敲 扥tw敥渠n桥獥⁥xtr敭敳I t桡t it will 扥捯浥c湥捥獳慲s to 獥sar慴攠
efo䘠
t敳ting from
bjf

s畳捥灴楢ility Eimm畮ityF t敳瑩e朮g
A

pr散e摥nt f潲

s数慲慴ing

efo䘠from bjf 桡猠
慬r敡摹 扥敮 獥s
with
i渠

-
ㄶN
K

-
ㄶN
b p散ti潮

7 is divided into two major parts, “Operational
Shock,” and “Crash Safety.”
Both deal with shock, but the “success” criteria differ.
Sections 22 and

23
divide the lightning ph
enomena into two parts, direct and indirect effects.

Both are lightning,
Section

22
deals with low level coupled effects, and section

23 with high level direct effects, and
again
,

the
“success” criteria differs.
Is not t
he same is true for EMI

susceptibi
lity and Immunity

vs. HIRF
?



The
separation of

EMI

(immunity)

from

HIRF
based solely on equipment operational requirements
within the EME
is outlined on

the following page
.





29

EMI



Conducted and Radiated Susceptibility

(Immunity)
, Non
-
HIRF


IF the equip
ment under test has exactly the same performance requirements for both the HIRF
and non
-
HIRF Electromagnetic Environment (EME), proceed to the HIRF test. An immunity test
is not required, unless frequency ranges differ
;

then, only those frequency ranges n
ot addressed
by the HIRF test need be tested for susceptibility (immunity).

Thus, the susceptibility / immunity
test described below would only apply to equipment with no HIRF requirements, o
r

in cases
where the equipment performance requirements differed

between the HIRF and non
-
HIRF
environment.

o

1

kHz square wave modulation only, no CW, no Pulse. (This
Susceptibility

/ Immunity
Test

is strictly for the onboard EME, and is intended to include PEDs.) Any C
W

/ Pulse
requirement will be addressed by HIRF t
esting.

o

EMI Susceptibility
(Immunity)
Testing (both conducted and radiated) to be performed in
the EUT’s most sensitive mode, using the step sizes / sweep rates / steps per decade
previously established
, and shown below
. The intent is to limit the number
of “EMI
/

susceptibility

/ immunity
runs

to 1 if possible, then move on
to HIRF with its higher levels
different modulations
,

and

larger step sizes
.

o

Radiated:
30

MHz

to 18

GHz

(Some categories may not need to go to 18

GHz)



Possible 2

MHz

lower frequency o
ption for HF equipped aircraft?

Is BCI
enough?



“Limited” number of radiated limits, perhaps 5, 20, 50, 100, and 200

V/m
?

Limits
flat across the frequency range.

Would include Category

R.

o

Conducted:
10

kHz

to 200

MHz

(BCI)

(Could extend to 400

MHz

if v
alue added)



BCI Limit based on the radiated limit and is set at
1.5

milliamperes per volt/meter
.
The BCI limit shall fall
-
off at 20

dB per decade below aircraft resonance, and
10

dB per decade above aircraft resonance. Aircraft resonant frequencies shall

be set per the existing DO
-
160 at 500

kHz (lower) and 30

MHz (upper).

N.B.: Aircraft coupling (
1.5

milliamperes per volt/meter
) is a constant, and does
NOT vary with cable or aircraft length. However, the resonant frequency (break
point) does, and for

especially large aircraft
the break point
may need to be
shifted lower, or for exceptionally small aircraft, shifted higher.



BCI Test Method: Induced Current limited by Forward Power. (Only one test
method, please.)

The following table illustrates the n
umber of steps per decade required to ensure 100% testing
of each frequency range for Non
-
HIRF test levels.

Frequency
Range

Non
-
HIRF

Q
S

(max.)

Max Analog
Scan Rate*

Maximum
Step Size

Steps per
Decade

Approx no.
steps in range

10

kHz
-

1

MHz

10

0.0333

f
0
/s
ec

0.05 f
0

49

49

1

MHz
-

30

MHz

50

0.00667

f
0
/sec

0.01 f
0

233

233

30

MHz
-

1

GHz

100

0.00333 f
0
/sec

0.005 f
0

463

463

1

GHz
-

8

GHz

500

0.000667

f
0
/sec

0.001 f
0

2305

2305

8

GHz
-

18

GHz

1000

0.000333

f
0
/sec

0.0005 f
0

4608

4608

*

Based on a 3
-
second dw
ell at each step.



30


HIRF

o

Frequency Range and Levels per the FAA HIRF Rule

o

Modulations per HIRF Rule. Pulse, CW, etc.

o

Any
additional
testing required by the Rule and NOT performed during EMI Testing
(above).

o

Operational Modes as determined by the Rule and /

or MOPS

o

HIRF testing may be performed in lieu of EMI testing only if performance within the
normal aircraft EMI environment is not important
, or the “success” criteria is the same for
both the HIRF and non
-
HIRF environments.

o

Step sizes / sweep rates for H
IRF should be based on committee
HIRF
data.

An
example follows.

The following table illustrates the number of steps per decade required to ensure 100% testing
of each frequency range for HIRF test levels.


UNVETTED TABLE for HIRF Testing

Frequency
Range

HIRF
*

Q
S

(m
ax
.)

Max Analog
Scan Rate
*
*

Maximum
Step Size

Steps per
Decade

Approx no.
steps in range

10

kHz
-

1

MHz

6

0.0556 f
0
/sec

0.0833 f
0

30

59

1

MHz
-

30

MHz

12.5

0.0267 f
0
/sec

0.04 f
0

60

88

30

MHz
-

1

GHz

25

0.0133 f
0
/sec

0.02 f
0

118

179

1

GHz
-

8

GHz

50

0.00667 f
0
/sec

0.01 f
0

233

210

8

GHz
-

18

GHz

100

0.00333 f
0
/sec

0.005 f
0

463

164

* HIRF Q
S

not validated. Based on limited samples suggesting Q = 20 to 40 near 1

GHz.

*
*

Based on a 3
-
second dwell at each step.



31

As a final suggestion, one of t
he existing DO
-
160E categories can be thought of as transitioning
between EMI (
onboard

environment) and HIRF (external environment), and that is Category

R
, which is
associated with the “normal” environment
.
The committee

could do the following:


Apply
20

V/m
,

modulated

by a

1

kHz square wave
, and use

the “EMI

Susceptibility
” steps
. Follow this
by CW and Pulse using the “HIRF” steps
. Test
up to 8

GHz

or
the
highest required frequency
.


Or


Apply
20

V/m modulated
by a 1

kHz
square wave using the “
EMI Susc
eptibility

/ Immunity
” steps up to
400

MHz, followed by 1

kHz 4% pulse
modulation
from 400

MHz to 8

GHz

or the highest required
frequency,
again using the “
EMI Susceptibility / Immunity
” steps. Any additional modulations desired by
the committee
(or “the
rule”)
would be performed using the “HIRF” steps.






32

APPENDIX A


EXAMPLE

for Radio Receiver Testing


The following example is for a “multi
-
mode” receiver, and illustrates how HIRF and EMI

Susceptibility

testing may be performed economically, while still
meeting all requirements
, and not ending up with a
“test
-
a
-
thon.”


Let’s assume there are four different receivers in one package. Further, let’s assume two
receiver
outputs, one receiver having a dedicated output, and the other three
receivers
sharing on
e output which
is switched between the receivers as needed. The control and I/O circuitry is similar for
each of the
four
receivers.



How could this equipment be economically tested without losing confidence in the results?


To perform the EMI

Susceptib
ility

test, base the actual susceptible test frequencies on the

“EMI”

step
-
size table presented earlier. In addition, incl
ude the LO / IF mix frequencies as described
in
the
DO
-
160E

note on page 20
-
8
for both the dedicated output receiver, and the receive
r under test. After
completing this “run,”
(1 kHz square wave modulation, vertical and horizontal polarizations, EMI limits),
two
of the four
receivers have been tested, and the control and I/O
for these receivers
have
also
been
EMI
tested. For the next

run

, select

(and activate)

one of the two remaining receivers. Since the
control and I/O have already been tested,
(by similarity),
only the LO / IF mix frequencies for the
selected receiver need be tested. Likewise for the last receiver.

At this poin
t, one complete EMI test
“run,” and two “partial” test “runs”
have been completed to verify

all EMI test requirements.


If it is required to run an entire MOPS during the susceptibility test, it really isn’t value
-
added to run the
entire MOPS at each of th
e step frequencies

since this would result in an unreasonable test time.
Instead, run the MOPS at only one “most critical” frequency for each radio. This one frequency should
be the first IF frequency.


NOTE that the EMI Test consists of only one modulat
ion (1

kHz

square wave), and both horizontal and
vertical polarizations. This testing should consume approximately 1 8
-
hour shift for the full run, and
approximately
½

shift for the LO / IF
mix
frequencies.


Proceed with HIRF
testing using increased step
sizes (fewer steps per decade) at the levels and
modulations prescribed by “the
HIRF
rule.”





Concluding remark: By separating
the
EMI requirements from
the
HIRF requirements

based on
equipment performance requirements
,
the test requirements can be simp
lified,
and
the document made
easier to
use and
follow
.


The committee

would also
ensure adequate testing
of the

EMI requirements
(by
specifying

a sufficient number of steps per decade)
without imposing the burden of
these additional
steps onto the HIRF re
quirements.






33

Appendix B


A variation of
the HIRF

table
that
has been suggested:



The intent and function of the reverberation chamber is for system level testing per ARP5583 and LRU
testing in the case where higher field levels are required (HIRF) th
at cannot be reached in an anechoic
chamber. The step rates for RS testing in a Reverberation chamber have been broken into three
frequency ranges representing key areas of interest during testing. Table 1 lists these ranges with their
corresponding step r
ates.


Table 1: Radiated Susceptibility Reverberation Test Method Step Rates


Frequency Range

HIRF

Q
S

(max.)

Max Analog
Scan Rate*

Maximum Step
Size

Steps per
Decade

Approx no.
steps in range

100

MHz
-

1

GHz

100

0.00333 f
0
/sec

0.005 f
0

463

463

1

GHz
-

2

GHz

86.3

0.00386 f
0
/sec

0.00579 f
0

400

121

2

GHz*
-

18

GHz*

42.9

0.00777 f
0
/sec

0.0117 f
0

200*

191

4

GHz*
-
18

GHz*

21.2

0.0157 f
0
/sec

0.0236 f
0

100*

66


Note:*

Testing above 4 GHz may be lowered to 100 steps/decade with concurrence of the test
director
and the
authorities
.

Actual steps/decade rates will

be recorded in the test report.


Furthermore, a three second dwell time has been selected for each tuner step within each frequency
step.


Rationalization of this data involves three key points.

1)

That despite modern high speed digital devices, susceptibility criteria for some devices
requires that an individual monitor the equipment visually,

2)

Historical data has shown that susceptibilities are more likely to occur in the 100

MHz to
2

GHz ran
ge, and

3)

Previously obtained data from anechoic room susceptibilities emulate the same low
frequency step rates. The result of this analysis is that these new step rates offer more than
twice the coverage than most of aerospace industry that follows spec
ifications to their
minimum standards. This results in a higher quality product that is more resilient to radiated
susceptibility environments.