Data Sheet ADL5605
Rev. A | Page 15 of 20
ACPR AND EVM
All adjacent channel power ratio (ACPR) and error vector
magnitude (EVM) measurements were made using a single
W-CDMA carrier and Test Model 1-64.
The signal is generated by a very low ACPR source and is meas-
ured at the output by a high dynamic range spectrum analyzer.
For ACPR measurements, the filter setting was chosen for low
ACPR; for EVM measurements, the low EVM setting was selected.
The spectrum analyzer incorporates an instrument noise correc-
tion function, and highly linear amplifiers were used to boost
the power levels for ACPR measurements.
Figure 26 shows ACPR vs. P
OUT
at 946 MHz. For power levels
up to 18 dBm, an ACPR of 51 dBc or better can be achieved
at 946 MHz.
Figure 27 shows EVM vs. P
OUT
at 946 MHz. The EVM measured
is 0.5% for power levels up to 18 dBm at 946 MHz. The baseline
composite EVM for the signal source was approximately 0.5%.
When operated in the linear region, there is little or no contribu-
tion to EVM by the amplifier.
THERMAL CONSIDERATIONS
The ADL5605 is packaged in a thermally efficient 4 mm ×
4 mm, 16-lead LFCSP. The thermal resistance from junction
to air (θ
JA
) is 52.1°C/W. The thermal resistance for the product
was extracted assuming a standard 4-layer JEDEC board with
25 copper plated thermal vias. The thermal vias are filled with
conductive copper paste (AE3030 with thermal conductivity of
7.8 W/mK and thermal expansion α1 of 4 × 10
−5
C and α2 of
8.6 × 10
−5
/°C). The thermal resistance from junction to case (θ
JC
)
is 12.1°C/W, where the case is the exposed pad of the lead frame
package.
For the best thermal performance, it is recommended that as
many thermal vias as possible be added under the exposed pad
of the LFCSP. The thermal resistance values assume a minimum
of 25 thermal vias arranged in a 5 × 5 array with a via diameter
of 8 mils, via pad of 16 mils, and a pitch of 20 mils. The vias are
plated with copper, and the drill hole is filled with a conductive
copper paste.
For optimal performance, it is recommended that the thermal
vias be filled with a conductive paste of the equivalent thermal
conductivity specified earlier in this section; alternatively, an
external heat sink can be used to dissipate heat quickly without
affecting the die junction temperature. It is also recommended
that the ground pattern be extended above and below the device
to improve thermal efficiency (see Figure 36).
SOLDERING INFORMATION AND RECOMMENDED
PCB LAND PATTERN
Figure 36 shows the recommended land pattern for the ADL5605.
To minimize thermal impedance, the exposed paddle on the
4 mm × 4 mm LFCSP is soldered to a ground plane along with
Pin 5 to Pin 8 and Pin 13 to Pin 16. To improve thermal dissi-
pation, 25 thermal vias are arranged in a 5 × 5 array under the
exposed paddle. Areas above and below the paddle are tied with
regular vias. If multiple ground layers exist, they should be tied
together using vias. For more information about land pattern
design and layout, see the AN-772 Application Note, A Design
and Manufacturing Guide for the Lead Frame Chip Scale Package
(LFCSP).
09353-035
RFOUT
RFIN
16 MIL VIA PAD
WITH 8 MIL VIA
16 13
5 8
Figure 36. Recommended Land Pattern
ADL5605 Data Sheet
Rev. A | Page 16 of 20
EVALUATION BOARD
The schematic of the ADL5605 evaluation board is shown in
Figure 37. The evaluation board uses 25 mils wide, 50 Ω traces
and is made from IS410 material with a 20 mils gap to ground.
The evaluation board is tuned for operation at 943 MHz. The
inputs and outputs should be ac-coupled with appropriately
sized capacitors; therefore, for low frequency applications, the
value of C1 and C2 may need to be increased. DC bias is
provided to the output stage via an inductor (L1) connected
to the RFOUT pin. A bias voltage of 5 V is recommended.
The evaluation board has a short, non-50 Ω line on its output
to accommodate the four output pins and to allow for easier low
inductance output matching. The pads for Pin 9 to Pin 12 are
included on this microstrip line and are included in all matches.
The evaluation board uses numbers as identifiers to aid in the
placement of matching components at both the RF input and
RF output of the device. Figure 38 and Figure 39 show images
of the board layout.
RFIN
VCC3
VCC2
DISABLE
RFOUT
12
VBIAS
11
VCC
10
DISABLE
9
RFIN
1
2
3
4
ADL5605
C
OUT
8pF
C2
100pF
C1
100pF
C
IN
N/A
C3
10pF
C4
OPEN
C10
OPEN
C5
100pF
C7
100pF
C8
0.01µF
C9
10µF
RFOUT
RFOUT
RFOUT
RFOUT
09353-036
13
8
14
7
15
6
16
5
NC NC NC NC
NC NC NC NC
C6
0.01µF
C11
10µF
C12
100pF
C13
0.01µF
C14
10µF
L1
18nH
L2
1.6nH
VCC1
R4
OPEN
R1
0Ω
R5
OPEN
R2
0Ω
Figure 37. Evaluation Board, 943 MHz Frequency Tuning Band
Table 8. Evaluation Board Configuration Options, 943 MHz Frequency Tuning Band
Component Function/Notes Default Value
C1, C2 Input/output dc blocking capacitors. C1, C2 = 100 pF
C3, C4, C5, C6, C7,
C8, C9, C10, C11,
C12, C13, C14
Power supply decoupling capacitors. Power supply decoupling capacitors are required to
filter out the high frequency noise on the power supply. The smallest capacitor should be the
closest to the ADL5605. The main bias that goes through RFOUT is the most sensitive to noise
because the bias is connected directly to the RF output.
C3 = 10 pF
C5, C7, C12 = 100 pF
C6, C8, C13 = 0.01 µF
C9, C11, C14 = 10 µF
C4, C10 = open
C
IN
Input matching capacitor. To match the ADL5605 at the 943 MHz or 881 MHz frequency tuning
band, C
IN
is not required. For the 748 MHz frequency tuning band, C
IN
is set at a specific distance
from the device so that the microstrip line can act as inductance for the matching network
(see Table 7). If space is at a premium, an inductor can take the place of the microstrip line.
C
IN
= open
C
OUT
Output matching capacitor. The output match is set for 943 MHz and is easily changed for
other frequency tuning bands. The tolerance of this capacitor should be tight. C
OUT
is set at
a specific distance from the device so that the microstrip line can act as inductance for the
matching network (see Table 7). If space is at a premium, an inductor can take the place of the
microstrip line. A short length of low impedance line on the output is embedded in the match.
C
OUT
= 8.0 pF HQ
L2 Output matching inductor. The output match is set for 943 MHz and is easily changed for other
frequency tuning bands. A high Q Coilcraft inductor with tight tolerance is recommended.
L2 = 1.6 nH HQ
L1 The main bias for the ADL5605 comes through L1 to the output stage. L1 should be high
impedance for the frequency of operation while providing low resistance for the dc current.
The evaluation board uses a Coilcraft 0603HP-18NX_LU inductor; this 18 nH inductor provides
some of the match at 943 MHz.
L1 = 18 nH
R1, R2, R4, R5 To provide bias to all stages through just one supply, set R1 and R2 to 0 Ω, and leave R4 and
R5 open. To provide separate bias to stages, set R1 and R2 to open and R4 and R5 to 0 Ω.
R1, R2 = 0 Ω
R4, R5 = open
Exposed Paddle
The paddle should be connected to both thermal and electrical ground.
Data Sheet ADL5605
Rev. A | Page 17 of 20
09353-037
Figure 38. Evaluation Board Layout, Top
09353-038
Figure 39. Evaluation Board Layout, Bottom

ADL5605ACPZ-R7

Mfr. #:
Manufacturer:
Analog Devices Inc.
Description:
RF Amplifier 1W 2-Stage, PA Driver Amp
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union

Products related to this Datasheet