Transcript Mobile Phone Testing using Impedance Tuners

Mobile Phone Testing using Impedance Tuners
Roman Meierer and Steve Dudkiewicz
Your Complete Measurement & Modeling Solutions Partner
1
We Live in a 50Ω World
The components have been designed for ideal 50Ω operation!
2
When are we not at 50Ω?
1) Antennas are not always 50Ω by design
How much mismatch between phone and antenna can be tolerated?
2) Non-ideal reception conditions mimic
non-50Ω conditions
Driving inside tunnel
Broken antenna
Talking inside
elevator
How much mismatch between phone and antenna/outside-world can be tolerated?
How can we even test outside of 50Ω?
3
Mobile Phone Testing Beyond 50Ω
It is not enough to test a phone at ideal 50Ω, it must be tested under realistic conditions!
Agilent 8960
- Acts as base station
- Sets active channel and power levels
- Measures power, BER, ACPR …
Maury MT98x Tuner
- Sets non-50Ω impedance
Maury AMTS Software
- Controls 8960 and tuner
- De-embeds VSWR and power levels
to DUT reference
- Controls variable DC supply to mimic
battery voltages
- Automates measurements
4
What is the Agilent 8960?
“The Agilent 8960 is your one-box solution for wireless device development, manufacturing and repair today and tomorrow “ – Agilent Technologies
Important Tests using 8960:
TX Power Test at specific PCL levels
RX Sensitivity BER testing
ACPR
Output Spectrum
Note: 8960 testing is at 50Ω condition only
5
Tuner emulates mismatched antenna
In order to emulate all non-50Ω conditions, we remove the antenna from the phone and replace
with cable connected to impedance tuner.
Impedance tuners can create any VSWR magnitude
between 1:1 and ~10-200:1 and phase between 0-360°
6
Easily Reproducible Test Conditions
No anechoic chamber or reference antenna needed Compact laboratory desktop setup results in easily
reproducible test systems and results
Test multiple modules under identical conditions in a fraction of the time, at multiple site locations!
7
The SLIDE-SCREW TUNER
Y
X
Probe
Airline
X
Y
Probe
Airline
8
Automatic Tuner or Manual Tuner ?
Maximum Phone Power
14
12
Maximum 8960 Power
Power or Loss (dB)
10
8
Phone Power
Tuner Loss
8960 Power
6
4
2
Tuner Loss
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Tuner Reflection
-2
9
The Pre-Calibrated Automated Tuner
Motors 2 & 3 (Probes)
Motor 1 (Carriage)
Slabline
10
Maury Repeatability Data
a
b
c
0
20
dB
40
60
80
100
•a) specification
b) Worst case repeatability
c) average repeatability
11
Tuner Characterization (Calibration)
S-parameters are recorded for various X-Y positions of the tuner at each frequency of interest
loss is calculated in real-time
instant DUT de-embedding
12
Fixed Source Power LP
Vds = 5.0 V
Idsq = 150 mA
Freq = 835 MHz
At maximum PAE of 70%
the power is 33 dBm
At maximum power of 38 dBm
the PAE is 51%
13
AMTS Software for Agilent 8960
14
TX Power Testing
- Set Frequency/Channel
- Set PCL transmit power level
- Set VSWR magnitude and phase
- Set DC voltage
Measure power delivered to antenna port as
function of VSWR magnitude and phase
33
TX Power [dBm]
32
31
VSWR_2
VSWR_3
VSWR_4
VSWR_5
30
29
28
27
0
90
180
270
360
PHASE
15
RX Sensitivity Testing
- Set Frequency/Channel
- Set required BER (2% or other)
- Set VSWR magnitude and phase
- Set DC voltage
Measure sensitivity of phone for given BER as
function of VSWR magnitude and phase
Sensitivity [dBm] for BER=2%
-105
-106
VSWR=2
-107
VSWR=3
VSWR=4
-108
VSWR=5
-109
-110
0
90
180
270
360
Phase
16
Automation through Scripting
Pre-define list of frequencies/channels, power levels, battery
voltages, VSWR settings
Automatically measure each line as a separate test 
intelligent sequencing
17
What do we do with the results?
1) Antennas are not always 50Ω by design
Tell the antenna designer how much mismatch is acceptable based on real-life tests
(did 2:1 VSWR at all phases meet requirements? 3:1?...)
2) Non-ideal reception conditions mimic non-50Ω conditions
-Did our mobile phone pass minimum requirements under pre-defined VSWR and
voltage conditions?
- After a large VSWR sweep plan, does the phone return to same performance at 1:1
VSWR? (performance may degrade 0.5-1.0 dB!)
- Are there thermal stability issues? (performance may degrade 0.5-1.0 dB inside
burst  check for “smooth” curves)
- Is the phone design acceptable or must we redesign a component?
- Is this a batch manufacturing issue?
These are only suggestions on how to analyze the data, the possibilities are endless!
18
Mobile Phone Test System
19