Rohde and Schwarz FSEA20 Data Sheet

NTP and ACP Measurements

to ETS 300 175-2 for DECT

using Spectrum Analyzer FSE

Application Note 1EF42_0E

Subject to change

24 March 1998, Robert Obertreis

Products:

NTP Measurement:

FSEx with Option FSE-B7

ACP Measurement:

FSEA 20/30 (FSE-B7 not required)

1. Introduction

DECT standard ETS 300 175-2 [1] prescribes
among others measurement of the transmit
power and the unwanted power in adjacent
channels. Following a brief introduction into
DECT, this Application Note describes the stan­dard-conforming measurement of the normal
transmit power NTP and the unwanted adjacent­channel power ACP using a spectrum analyzer
from the FSE family ( firmware version 1.63 and
higher).

Option FSE-B7 (Vector Signal Analyzer) is re­quired for measuring the normal transmit power
NTP since the signal must be demodulated for
this measurement.

Due to the wide dynamic range of FSEA20 and
FSEA30, the adjacent channel power ACP can
be measured without bandpass filters suppress­ing the transmit channel. Option FSE-B7 is not
required for measuring the adjacent-channel
power.

2. Introduction into DECT

The band between 1880 MHz and 1900 MHz is
reserved for DECT. The modulation mode used
with DECT is GFSK (Gaussian f
keying). The maximum FM deviation is 288 kHz.

The DECT band is divided into 10 equal sub­bands. The center frequency
tive channel is defined by

fcfc z=−⋅01728kH , where

and c = 01 9, ,...., .

Depending on the channel number c, the follow­ing center frequencies are obtained:

c

0 1897.344
1 1895.616
2 1893.888
3 1892.160
4 1890.432
5 1888.704
6 1886.976
7 1885.248
8 1883.520
9 1881.792

f

c

/MHz

Table 1: DECT center frequencies
With DECT, up to ten connections to a base

station can be set up simultaneously. Within a
frequency channel, transmit and receive channel
are separated by time slots. This method is re­ferred to as TDMA (time d

ivision multiple ac-

cess). A TDMA frame contains 11520 bits with a

requency shift

f

of the respec-

c

f01897344= .MHz

r

transfer rate

= 1152 bit/s1. The time period of a

frame is obtained from

Frame length

Transfer rate

11520bit

1152kbit / s

10ms==.

A frame is subdivided into 24 full slots, as shown
in the following illustration.

normal RFP transmit normal PP transmi t

slot0slot

1

slot11slot12slot

Frame, 11520 bits

13

slot
24

RFP : Radio fixed part (base station)
PP: Portable part (mobile phone)

Fig. 1: TDMA frame with DECT
Within a full slot having a length of 480 bits the

data are transferred in form of a socalled

cal packet

.

physi-

DECT standard ETS 300 175-2 defines various
physical packets which differ in their length. For
communication, the

basic physical packet P32

as shown in Fig. 2 is mostly used.

f0 f479

S0 S31 d0 d387 z0 z3

S­field

p0 p423

full-slot

D-field

packet P32

p419

Z­field

Fig. 2: Basic physical packet
The following table gives an overview of the

various physical packets with their bit length as
defined in the standard.

S-field D-field Z-field
P00
P32
P08j
P80

32 bit 64 bit 0 bit

32 bit 388 bit 4 bit

32 bit 148 bit 4 bit

32 bit 868 bit 4 bit
Table 2: Physical packets

The socalled S-field contains a 32-bit synchroni­zation sequence which is different for the base
station (RFP) and the mobile phone (PP):

RFP sequence:
1010 1010 1010 1010 1110 1001 1000 1010

PP sequence:
0101 0101 0101 0101 0001 0110 0111 0101

1

With GFSK modulation as used for DECT the bit rate is equal

to the symbol rate.

1EF42_0E 2 24.03.1998

The D-field is provided for data transmission and
the 4-bit Z-field may be used for additional syn­chronization tasks.

3 Normal Transmit Power (NTP)

3.1 Definition

NTP (normal transmit power) is the transmitted
power averaged over a physical packet from bit
p0 to the end of the packet. The mean transmit­ted power with DECT should not exceed
250 mW (24 dBm).

To enable correct power measurement, the ex­act position of the first bit p0 within a packet has
to be determined.

3.2 Measurement of NTP using FSE

This measurement can be performed quickly
and accurately with the aid of the vector signal
analyzer in the FSE. The predefined digital
standards facilitate synchronization of the burst.

Prerequisite for this measurement is an external
trigger signal to establish the time reference to
the burst, as shown in Fig. 3.

Frame­Trigger

Burst

Trigger-Delay

Fig. 3: Frame trigger
The following settings have to be made on FSE

in the stated order.

SYSTEM - PRESET:

set)

FREQUENCY - CENTER: (

frequency)

LEVEL - REF: (

• Change to vector analyzer mode:

CONFIGURATION - MODE:

VECTOR ANALYZER:

(resetting the FSE by pre-

input of channel

input of maximum signal power2)

DIGITAL STANDARDS: DECT

⇑

(menu up)

By selecting the digital standard DECT, all set­tings required for measurement of a basic physi­cal packet (423 bits) are made. If another physi­cal packet is to be measured, the result length
must be varied manually. To do so, the following
settings have to be made in the Vector Analyzer
menu:

CONFIGURATION - MODE:

MEAS RESULT:
RESULT LENGTH

3

: (length of physical

packet)

For synchronization of the signal the following
settings are required in the Trigger menu.

SWEEP - TRIGGER:

TRIGGER: EXTERN

FIND BURST: (on)
FIND SYNC: (on)

Should

SYNC NOT FOUND be indicated in the dis-

play, the synchronization pattern has to be
changed. For DECT, different patterns are al­ready available for the base station FP and the
mobile phone PP.

These patterns can be selected in the Trigger
menu as follows:

SWEEP - TRIGGER:

SYNC PATTERN:
PATTERN NAME: (dect_fp or dect_pp)

• The MAGNITUDE CAP BUFFER softkey can

be used to display the magnitude of the sig­nal stored in the measurement result memory
in the time domain.

CONFIGURATION - MODE:

MEAS RESULT: MAGNITUDE CAP
BUFFER

If prior to the synchronization sequence a
socalled start-up bit sequence is transmitted, this
sequence has to be considered in the demodu­lation. Otherwise the rising edge of the burst
would appear in the display and the power
measurement is incorrect. SYNC OFFSET can
be used to used to vary the trace position within
the demodulated signal.

2

The following settings allow the maximum signal level to be

determined:

FREQUENCY - SPAN: (10 MHz)
TRACE - 1: MAX - HOLD

3

Due to the limited capacity of the FSE memory and the given
symbol rate of 1.152 Mbit/s the length of the result is limited to
600 symbols. Physical packets containing more than 600
symbols (P80) therfore cannot be completely demodulated.

1EF42_0E 3 24.03.1998

SWEEP - TRIGGER:

µ

µ

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

A

AA

A

A

A

A

A

A

A

AA

SYNC OFFSET:

(number of bits ahead of

synchronization sequence)

The demodulated signal is shown in the illustra­tion below. The x axis extends from symbol p0
to symbol p423.

Fig. 4: DECT signal in

Magnitude Cap Buffer

display mode

• Measure the power in the Marker menu:

MARKER - SEARCH:

SUMMARY MARKER: MEAN

The MEAN power indicated on the top right is
the power averaged from bit p0 to bit p423 in
line with the standard.

4. Unwanted Power in the Adjacent
Channels due to Modulation

4.1 Definition

Emission in RF

Maximum power

channel

YM=±1
YM=±2
YM=±3

Y = any other DECT

160

1
40 nW
20 nW

W
W

channel
Table 3: Permissible adjacent-channel power

caused by modulation

The power in a frequency channel Y is defined
by integration over the power density. The inte­gration range has a width of 1 MHz and lies at

the center frequency

F

. Averaging should be

Y

made over a range of at least 60% to max. 80%
of the length of the physical packet. Measure­ment should start before 25% of the physical
packet has been emitted, but after the 32-bit
synchronization word.

FSE allows the measurement to be stopped by a
gate in case of an inactive gate signal. It is thus
possible to display the spectrum of pulsed RF
carriers without superimposed frequency compo­nents caused by switching the carrier on and off.

The permissible gate settings are graphically
shown in Fig. 5. The position of the gate lines
should be within the cross-hatched field. The
times for a basic physical packet are indicated in
parentheses.

25% ( 92 us )

Sync.

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

AAA

80%( 294 us )

60% ( 220 us )

<75%( <275 us )

60% ( 220 us )

AAA

AAA

AAA

AAA

AAA

AAA

AAA

A

A

A

A

A

A

A

AAA

A

AAA

A

AAA

A

AAA

A

AAA

A

AAA

A

AAA

A

The power emission in adjacent channels
caused by frequency modulation is to be meas­ured. Since the DECT signal is pulsed, the sig-

32Bit ( 28us )

Fig. 5: Gate settings

nal components caused by switching the carrier
on and off have to be ignored. To this end the
measurement has to be stopped outside the
burst.

The DECT standard prescribes the following
settings and limit values for the measurement:

Upon transmission of the frequency channel M
in successive frames the power in the adjacent
channels Y should not exceed the specified val­ues.

1EF42_0E 4 24.03.1998

4.2 Dynamic range of FSE

The usable dynamic range of a spectrum ana­lyzer is limited by the
oscillator, the

thermal input noise

namic response towards high levels).
Investigation of the short-term stability of an

oscillator in the frequency range leads to the
definition of phase noise. An ideal oscillator
produces a line in the spectrum while a real
oscillator produces a continuous noise spectrum
symmetrically to the carrier frequency (Fig. 6).

Power­density

Fig. 6: Signal spectrum and SSB phase noise
density

The usual definition of phase noise refers the
single-sideband noise power PRf

bandwidth at an offset frequency
carrier to the carrier power

is logarithmized to yield

Lf

() lg'()=⋅10 dBc(1Hz)

m

Lf

in dBc(Hz), ie the noise level in 1 Hz bandwidth
below the carrier level.

With a maximum transmitter power of +24 dBm,
the power emission in remote adjacent channels
caused by modulation (see Table 3) should not
exceed a value of 20 nW (^ -47 dBm), ie the
required dynamic range is 71 dBc. This value
corresponds to the required minimum spacing
between the noise power densities

PRf

'( )and

m

P

Accordingly, for a required resolution bandwidth
of 100 kHz (50 dB shape factor) the specified
phase noise of the spectrum analyzer in the
adjacent channel must be smaller than

-121 dBc(Hz).
The DECT standard prescribes measurement of

the adjacent-channel power according to the

f

m

m

.

T

phase noise

Lf

'( )

=

m

1Hz

'( ) in 1 Hz

P

. This relationship

T

of the local

(and the dy-

Pf

'( )

Rm

P

T

frequency

m

f

from the

m

max-hold method. Since the maximum phase
noise level is approx. 10 dB above the value
stated above, the phase noise of the spectrum
analyzer must be lower than -131 dBc(Hz). Ta­ble 4 gives some typical phase noise values of
different analyzer models for 5 MHz carrier
spacing at a carrier frequency

FSEA20 FSEB20

FSEM20
FSEK20

-147 dBc(Hz) -141 dBc(Hz) -145 dBc(Hz) -140 dBc(Hz)

f

= 3.5 GHz:

FSEA30 FSEB30

FSEM30
FSEK30

Table 4: Typical phase noise values of FSE

spectrum analyzers
Therefore phase noise is not a limiting parame-

ter for the measurement of adjacent channel
power with the FSE

Like phase noise, the thermal input noise also
limits the usable dynamic range of the spectrum
analyzer. With a resolution bandwidth of

100kHz

as prescribed by DECT a lower sensitivity limit
of

NF F

=− + +

174

dBm

100

++

10

1

kHz

Hz

Analyzer

dBlg .

a

0

is obtained as a function of the noise figure of
the spectrum analyzer and the attenuation

a

.

0

The noise figure of the spectrum analyzer is
influenced by the selected reference level. With
a high reference level, the internal gain is small
and the noise of the IF stages has a greater
effect on the overall noise of the FSE.

To minimize the noise figure of FSE, the first
mixer can be overdriven up to +0 dBm

4

for
measurement of the adjacent-channel power.
With a maximum input power of +24 dBm,
10 dB reference level and 30 dB attenuation of
the FSE the mixer level is -6 dBm. In this case
the IF noise is negligible.

The displayed average noise level

N

differs for

0

the various FSE models, some values being
given in the table below.

4

The intermodulation-free range for mixer levels of 0 dBm is

smaller than the dynymic range required for adjacent-channel
measurements. Since only one frequency is applied at a time,
this value may be exceeded.

1EF42_0E 5 24.03.1998

FSEA20,
FSEA30

FSEB20,
FSEB30

FSEM20,
FSEM30,

FSEK20,
FSEK30

<-75 dBm
typ. -80 dBm

(RBW = 100 kHz, 30 dB RF attenuation, f = 1.9 GHz).

<-72 dBm
typ. -77 dBm

<-68 dBm
typ. -70 dBm

Table 4: Displayed average noise level of FSE
Due to the max-hold method being used for the

measurement, the displayed noise level is ap­prox. 10 dB higher.

The power in the adjacent channels is deter­mined by integration of the noise density over
the channel bandwidth

CHBW = 1MHz . Since

without signal applied the noise density is ap­proximately constant, the FSE noise level in a
channel can be calculated as follows, taking into
account the channel bandwidth:

LN

CHBW

00

10 lg

CHBW

RBW

dB 10 dB=+ =+

N

With a selected resolution bandwidth

RBW =100kHz and a channel bandwidth
CHBW = 1MHz , the channel noise power is
10dB above the displayed noise level N

.

0

Considering all settings (max hold, integration
over

1MHz), the channel noise power is

-55 dBm The measured (single sweep) channel
noise power is -54.4 dBm and in good agree­ment with the calculated value (Fig. 7).

The phase noise and the thermal input noise will
affect the measurement result if the ratio of the
signal level to the internal noise is not at least 20
dB. It should be noted that it is not the signal
alone which is displayed, but the sum power of
the phase noise of the internal oscillators, the
thermal noise and the signal power.

The following diagram shows an error curve of
the channel power (
tion of the signal-noise ratio in the transmit
channel

5

. Due to the different signal statistics in

CHBW

= 1 MHz) as a func-

the transmit and adjacent channel, this curve
can be applied to the adjacent-channel power
with some reservations
only.

3

2,5

2

1,5

0,5

Correction in dB

0.8dB

1

0

0 2 4 6 8 10121416

Channel power over nois e in dB

Fig. 8: Diagram for correction of signal level

in transmit channel

Because of this the displayed adjacent channel
power is about 0.8dB too high.

4.3 Measurement of Adjacent­Channel Power due to Modulation
using the FSE

For this measurement, FSE is operated in the
analyzer mode and the DECT signal is applied
to the RF input. The reference level should be
set first according to the maximum signal level.

An external trigger signal is to be used for the
measurement.

SYSTEM - PRESET:

(resetting the FSE by pre-

set)

FREQUENCY - CENTER:

(input of channel

frequency)

FREQUENCY - SPAN: (15 MHz)

• • Set resolution bandwidth and video band-

width:

Fig. 7: Noise level and noise power of FSEA

without input signal

Since the minimum allowed level to DECT is ­47 dBm, the distance between noise power and
the unwanted signal due to modulation is
approx. +8dB.

5

Settings on FSE:

RBW=100 kHz, VBW=300 kHz, RF ATT=30 dB, CEN­TER=1.9 GHz, single sweep, channel power measured in
1 MHz bandwidth

1EF42_0E 6 24.03.1998

SWEEP - COUPLING:

RES BW MANUAL: (100 kHz)
VIDEO BW MANUAL: (300 kHz)

• • Set reference level and attenuation.
LEVEL - REF: (see chapter 4.2

6

)

ATTEN AUTO LOW NOISE

• Set the gate:
SWEEP - SWEEP: GATE SETTINGS:

GATE ADJUST

time domain)

(the display changes to the

:
GATE DELAY: (see Fig. 5)
GATE LENGTH: (see Fig. 5)

A DECT burst is shown in the following screen
display. The gate settings are marked by two
vertical lines designated GD and GL.

Typical values for a basic physical packet with a
normal transmit power of +24 dBm:

LEVEL REF: (+10 dBm)
GATE DELAY: (

GATE LENGTH: (
SWEEP TIME MANUAL: (

30 µs )

290 µs)

1200 290 348* µs= ms )

• Select the prescribed recording mode

TRACE 1: MAX HOLD

• Perform power measurement:

MARKER - NORMAL:

⇐ (left submenu)

POWER MEAS SETTINGS:
CHANNEL BANDWIDTH: (1 MHz)
CHANNEL SPACING: (1.728 MHz)

SET NO. OF ADJ CHAN’S: (3)

⇑ (menu up)

ADJACENT CHAN POWER
CP/ACP: (ABS)

• Carry out the measurement with the given
sweep time in single-sweep mode.

SWEEP - SWEEP: SINGLE SWEEP
The displayed channel power in the respective

channel is the channel power density integrated
between the limit lines

cu

and cl as specified in
the standard. The measured channel power CH
PWR in the transmit channel is not conform to
the standard and differs from the measured
NTP.

30 dBm

Fig. 9: DECT burst with gate settings for

modulation spectrum

7

• Set the required sweep time

SWEEP - COUPLING:

SWEEP TIME MANUAL: (1200* GATE
LENGTH

8

)

6

The following settings allow the power ramp and hence the

maximum signal level to be determined:

TRACE - 2: MAX HOLD

The trace can be blanked with softkey BLANK in menu TRACE

2.

Fig. 10: Measurement of channel power in fre-

7

To display the burst in the center, a trigger delay of 340 µs
was selected. In addition, the time scale was expanded to
100 µs/div.

8

The sweep time of 12 s prescribed by the standard refers to a
continuous sweep over 1200 TDMA frames. With FSE the

sweep is stopped outside the gate. With the same effective
measurement time, a sweep time is obtained for FSE which is
1200 times the GATE LENGTH.

quency range

1EF42_0E 7 24.03.1998

5. Unwanted Power in Adjacent

Channels due to Transients

5.1 Definition

According to the standard, the power levels of all
modulation products in the adjacent channels
including the AM products due to switching the
RF carrier on and off should be smaller than the
values given in the following table.

Emission in RF

Maximum power

channel

YM=±1
YM=±2
YM=±3

Y = any other DECT

250 µW

40 µW

4 µW
1 µW

channel

Table 6: Max. permissible adjacent-channel

power due to transients

5.2 Measurement of Adjacent-Channel
Power due to Transients using the
FSE

The measurement is similar to that of adjacent­channel power due to modulation. The only dif­ference is that the transients of the bursts have
to be measured too. For this purpose the gate
must be set so that the burst is completely within
the window as shown in Fig. 11.

The settings for the gate delay are always re­ferred to the trigger time and there are no nega­tive values as in the case of the trigger delay.

If the time between the trigger event and the
rising edge of the burst is too short, part of the
rising edge will be clipped by the gate. This can
be prevented by increasing the gate delay so
that the second burst following the trigger event
will be measured. Based on the settings for the
adjacent-channel power due to modulation, the
gate settings are to be varied accordingly.

• Select setting mode

SWEEP - SWEEP:

CONTINUOUS SWEEP
SWEEP TIME MANUAL (20 ms)

• Gate settings
SWEEP - SWEEP: GATE SETTINGS:

GATE ADJUST

time domain)

GATE DELAY: (shortly before the burst

(the display changes to the

:

9

)

GATE LENGTH: (shortly after the burst)

⇑ (menu up)

GATE EXTERN

⇑ (menu up)

GATE: (on)
Select measurement mode

SWEEP - SWEEP:

SWEEP TIME MANUAL (500 ms
SINGLE SWEEP

For measuring the power, proceed as described
for the adjacent-channel power due to modula­tion. The following illustration shows the result of
an adjacent-channel power measurement in a
spectrum widened by transients.

10

)

Fig. 11: Gate setting for measurement of tran-

sients spectrum

Compared to the adjacent-channel power due to
modulation, the permissible power in this case is

9

For the measurement it is important that all burst components
contributing to the spectrum are covered by the measurement,
whereas components caused by the counterpart are not to be
measured.

higher so that the requirements regarding the
usable dynamic range are more easy to satisfy.

10

The sweep time is not specified in ETS 300 175-2 standard
for the measurement of adjacent-channel power due to transi­ents.

1EF42_0E 8 24.03.1998

30 dB

Fig. 12: Spectrum widened by transients

References

[1] European Telecommunications Standards

Institute ETSI,

(RES); Digital European Cordless Telecommuni­cations (DECT) Common Interface Part 2:
Physical layer,

[2] Josef Wolf:

using Spectrum Analyzers of FSE Family,

cation Note, Rohde&Schwarz 1995

Radio Equipment Systems

ETS 300 175-2, September 1996

Measurement of Phase Noise

Appli-

Ordering information

Spectrum Analyzer FSEA30 1065.6000.30
Option B7: Vector Signal
Analyzer 1066.4317.02

Robert Obertreis, 1ESP

Rohde & Schwarz Munich

24.03.1998

1EF42_0E 9 24.03.1998