AVAGO ATF-521P8-TR2

ATF-521P8
High Linearity Enhancement Mode [1] ­Pseudomorphic HEMT in
2x2 mm2 LPCC[3] ­Package
Data Sheet
Avago Technologies’ ATF‑521P8 is a single-voltage high
linearity, low noise E‑pHEMT housed in an 8-lead JEDECstandard leadless plastic chip carrier (LPCC[3]) package.
The device is ideal as a medium-power, high-linearity
amplifier. Its operating frequency range is from 50 MHz
to 6 GHz.
The thermally efficient package measures only 2mm x
2mm x 0.75mm. Its backside metalization provides ex‑
cellent thermal dissipation as well as visual evidence of
solder reflow. The device has a Point MTTF of over 300
years at a mounting temperature of +85°C. All devices
are 100% RF & DC tested.
Specifications
• 2 GHz; 4.5V, 200 mA (Typ.)
• 42 dBm output IP3
Features
• Single voltage operation
• High linearity and P1dB
• Low noise figure
• Excellent uniformity in product specifications
• Small package size:
2.0 x 2.0 x 0.75 mm3
• Point MTTF > 300 years[2]
• MSL-1 and lead-free
• Tape-and-reel packaging option available
Pin Connections and Package Marking
Pin 8
Pin 7 (Drain)
• 26.5 dBm output power at 1 dB gain compression
Pin 6
• 1.5 dB noise figure
Pin 5
• 17 dB Gain
• 12.5 dB LFOM[4]
Applications
• Front-end LNA Q2 and Q3, driver or pre-driver amplifier
for Cellular/PCS and WCDMA wireless infrastructure
• Driver amplifier for WLAN, WLL/RLL and MMDS applica‑
tions
• General purpose discrete E-pHEMT for other high linear‑
ity applications
Note:
1. Enhancement mode technology employs a single positive Vgs,
eliminating the need of negative gate voltage associated with
conventional depletion mode devices.
2. Refer to reliability datasheet for detailed MTTF data
3. Conform to JEDEC reference outline MO229 for DRP-N
4. Linearity Figure of Merit (LFOM) is essentially OIP3 divided by DC
bias power.
Source
(Thermal/RF Gnd)
Description
Pin 1 (Source)
Pin 2 (Gate)
Pin 3
Pin 4 (Source)
Bottom View
Pin 1 (Source)
Pin 2 (Gate)
Pin 3
Pin 8
2Px
Pin 4 (Source)
Pin 7 (Drain)
Pin 6
Pin 5
Top View
Note:
Package marking provides orientation and identification
“2P” = Device Code
“x” = Month code indicates the month of manufacture.
Attention:
Observe precautions for handling electrostatic
­sensitive devices.
ESD Machine Model (Class A)
ESD Human Body Model (Class 1C)
Refer to Avago Technologies Application Note A004R: Electrostatic
Discharge Damage and Control.
ATF-521P8 Absolute Maximum Ratings [1]
Parameter
VDS
Drain – Source Voltage [2]
V
7
VGS Gate­ –Source Voltage
V
-5 to 1
VGD
Gate Drain Voltage [2]
V
-5 to 1
I
Drain Current DS
Units
Absolute
Maximum
Symbol
[2]
[2]
IGS
Gate Current
P
Total Power Dissipation Pin max.
RF Input Power
TCH
TSTG
θch_b
diss
mA
500
mA
46
W
1.5
dBm
27
Channel Temperature
°C
150
Storage Temperature °C
-65 to 150
Thermal Resistance °C/W
45
[3]
[4]
Notes:
1. Operation of this device in excess of any one of these parameters may cause permanent damage.
2. Assumes DC quiescent conditions.
3. Board (package belly) temperatureTB is 25°C. Derate 22 mW/°C for TB > 83°C.
4. Channel to board thermal resistance measured using 150°C Liquid Crystal Measurement method.
5. Device can safely handle +27dBm RF Input Power provided IGS is limited to 46mA. IGS at P1dB drive level is bias circuit dependent.
Product Consistency Distribution Charts [5, 6]
180
600
0.8V
500
4
VDS (V)
IDS (mA)
180
0
6
8
0
0.5
1
1.5
2
2.5
3
NF (dB)
300
37
39
41
43
45
47
49
Figure 3. OIP3 @ 2 GHz, 4.5 V, 200 mA.
Nominal = 41.9 dBm, LSL = 38.5 dBm.
Cpk = 4.6
Stdev = 0.11
250
120
0
OIP3 (dBm)
Figure 2. NF @ 2 GHz, 4.5 V, 200 mA.
Nominal = 1.5 dB.
Cpk = 2.13
Stdev = 0.21
150
+3 Std
30
30
0.4V
2
-3 Std
60
0.5V
0
90
+3 Std
60
Vgs = 0.6V
Figure 1. Typical I-V Curves.
(VGS = 0.1 V per step)
200
-3 Std
90
+3 Std
100
30
50
15
16
17
18
GAIN (dB)
Figure 4. Gain @ 2 GHz, 4.5 V, 200 mA.
Nominal = 17.2 dB, LSL = 15.5 dB,
USL = 18.5 dB.
-3 Std
150
60
0
-3 Std
90
100
0
120
120
300
200
Cpk = 0.86
Stdev = 1.32
150
0.7V
400
150
Stdev = 0.19
19
0
25
25.5
26
+3 Std
26.5
27
27.5
P1dB (dBm)
Figure 5. P1dB @ 2 GHz, 4.5 V, 200 mA.
Nominal = 26.5 dBm, LSL = 25 dBm.
Notes:
5. Distribution data sample size is 500 samples taken from 5 different wafers. Future wafers allocated to this product may have nominal values
anywhere between the upper and lower limits.
6. Measurements are made on production test board, which represents a trade-off between optimal OIP3, P1dB and VSWR. Circuit losses have
been de‑embedded from actual measurements.
ATF-521P8 Electrical Specifications
TA = 25°C, DC bias for RF parameters is Vds = 4.5V and Ids = 200 mA unless otherwise specified.
Symbol
Parameter and Test Condition
Units
Min.
Typ.
Vgs
Operational Gate Voltage
Vds = 4.5V, Ids = 200 mA
V
—
0.62
Vth
Threshold Voltage
Vds = 4.5V, Ids = 16 mA
V
—
0.28
Idss
Saturated Drain Current
Vds = 4.5V, Vgs = 0V
µA
—
14.8
Gm
Transconductance
Vds = 4.5V, Gm = ∆Idss/∆Vgs;mmho
—
1300
Vgs = Vgs1 - Vgs2
Vgs1 = 0.55V, Vgs2 = 0.5V
Igss
Gate Leakage Current
Vds = 0V, Vgs = -4V
µA-20
0.49
[1]
NF
Noise Figure f = 2 GHz
dB­
—
1.5
f = 900 MHz
dB
—
1.2
G
Gain[1]
f = 2 GHz
dB
15.5
17
f = 900 MHz
dB
—
17.2
OIP3
Output 3rd Order
f = 2 GHz
dBm
38.5
42
Intercept Point[1] f = 900 MHz dBm
—
42.5
P1dB
Output 1dB f = 2 GHz
dBm
25
26.5
Compressed[1]
f = 900 MHz
dBm
—
26.5
PAE
Power Added Efficiency
f = 2 GHz%
45
60
—
f = 900 MHz
%
—
56
—
ACLR
Adjacent Channel Leakage
Offset BW = 5 MHz
dBc
—-51.4
Power Ratio[1,2]
Offset BW = 10 MHz
dBc
—-61.5
Max.
—
—
—
—
—
—
—
18.5
—
—
—
—
—
—
—
Notes:
1. Measurements obtained using production test board described in Figure 6.
2. ACLR test spec is based on 3GPP TS 25.141 V5.3.1 (2002-06)
– Test Model 1
– Active Channels: PCCPCH + SCH + CPICH + PICH + SCCPCH + 64 DPCH (SF=128)
– Freq = 2140 MHz
– Pin = -5 dBm
– Chan Integ Bw = 3.84 MHz
Input
50 Ohm
Transmission
Line Including
Gate Bias T
(0.3 dB loss)
Input
Matching Circuit
Γ_mag = 0.55
Γ_ang = -166°
(1.1 dB loss)
DUT
Output
Matching Circuit
Γ_mag = 0.35
Γ_ang = 168°
(0.9 dB loss)
50 Ohm
Transmission
Line and
Drain Bias T
(0.3 dB loss)
Output
Figure 6. Block diagram of the 2 GHz production test board used for NF, Gain, OIP3 , P1dB and PAE and ACLR measurements. This circuit achieves a trade-off between optimal OIP3, P1dB and VSWR. Circuit losses have been de-embedded from actual measurements.
1 pF
3.9 nH
50 Ohm
.02 λ
1.5 pF
RF Input
110 Ohm
.03 λ
DUT
110 Ohm
.03 λ
50 Ohm
.02 λ
12 nH
1.5 pF
RF Output
47 nH
15 Ohm
2.2 µF
Drain
Supply
2.2 µF
Gate
Supply
Figure 7. Simplified schematic of production test board. Primary purpose is to show 15 Ohm series resistor placement in gate supply.
Transmission line tapers, tee intersections, bias lines and parasitic values are not shown.
Gamma Load and Source at Optimum OIP3 and P1dB Tuning Conditions
The device’s optimum OIP3 and P1dB measurements were determined using a Maury load pull system at 4.5V, 200
mA quiesent bias:
Optimum OIP3
Freq
Gamma Source
Gamma Load
OIP3
(GHz)
Mag
Ang (deg)
Mag
Ang (deg)
(dBm)
Gain
(dB)
P1dB
(dBm)
PAE
(%)
0.9
0.413
10.5
0.314
42.7
16.0
27.0
54.0
2
0.368
162.0
0.538-176.0
42.5
15.8
27.5
55.3
2.4
0.318
169.0
0.566-169.0
42.0
14.1
27.4
53.5
3.9
0.463-134.0
0.495-159.0
40.3
9.6
27.3
43.9
Optimum P1dB
Freq
Gamma Source
Gamma Load
OIP3
(GHz)
Mag
Ang (deg)
Mag
Ang (deg)
(dBm)
Gain
(dB)
P1dB
(dBm)
PAE
(%)
0.9
0.587
12.7
0.613-172.1
39.1
14.5
29.3
49.6
2
0.614
126.1
0.652-172.5
39.5
12.9
29.3
49.5
2.4
0.649
145.0
0.682-171.5
40.0
12.0
29.4
46.8
3.9
0.552-162.8
0.670-151.2
38.1
9.6
27.9
39.1
179.0
ATF-521P8 Typical Performance Curves (at 25°C unless specified otherwise)
Tuned for Optimal OIP3
50
45
50
45
40
45
30
25
15
150
200
250
300
350
30
25
20
4.5V
4V
3V
20
OIP3 (dBm)
35
10
100
40
35
OIP3 (dBm)
OIP3 (dBm)
40
4.5V
4V
3V
15
10
100
400
150
200
350
25
4.5V
4V
3V
20
15
400
10
100
150
30
30
30
20
4.5V
4V
3V
15
150
200
250
300
350
P1dB (dBm)
35
P1dB (dBm)
35
25
25
20
10
100
150
200
Idq (mA)
250
300
350
10
100
400
16
11
15
15
10
12
200
250
300
350
14
13
12
4.5V
4V
3V
11
GAIN (dBm)
16
GAIN (dBm)
12
150
200
Id (mA)
Figure 14. Small Signal Gain vs Ids and Vds
at 2 GHz.
10
100
150
200
250
350
400
300
350
9
8
4.5V
4V
3V
6
400
5
100
150
200
Id (mA)
Figure 15. Small Signal Gain vs Ids and Vds
at 900 MHz.
250
300
350
400
Id (mA)
Figure 16. Small Signal Gain vs Ids and Vds
at 3.9 GHz.
Note:
Bias current for the above charts are quiescent conditions. Actual level may increase depending on amount of RF drive.
300
7
4.5V
4V
3V
11
400
250
Figure 13. P1dB vs. Idq and Vds at 3.9 GHz.
17
10
100
150
Idq (mA)
17
13
400
4.5V
4V
3V
15
Figure 12. P1dB vs. Idq and Vds at 900 MHz.
14
350
20
Idq (mA)
Figure 11. P1dB vs. Idq and Vds at 2 GHz.
300
25
4.5V
4V
3V
15
400
250
Figure 10. OIP3 vs. Ids and Vds at 3.9 GHz.
35
10
100
200
Id (mA)
Figure 9. OIP3 vs. Ids and Vds at 900 MHz.
Figure 8. OIP3 vs. Ids and Vds at 2 GHz.
P1dB (dBm)
300
30
Id (mA)
Id (mA)
GAIN (dBm)
250
35
ATF-521P8 Typical Performance Curves, continued (at 25°C unless specified otherwise)
Tuned for Optimal OIP3
70
70
60
60
50
50
50
45
40
40
4.5V
4V
3V
20
150
4.5V
4V
3V
20
200
250
300
350
10
100
400
150
27
40
25
35
30
85°C
25°C
-40°C
25
20
2.5
250
300
350
10
100
400
150
200
3
3.5
21
85°C
25°C
-40°C
1
1.5
2
2.5
3
3.5
FREQUENCY (GHz)
Figure 20. OIP3 vs. Temp and Freq
tuned for optimal OIP3 at 4.5V, 200 mA.
Figure 21. P1dB vs. Temp and Freq
tuned for optimal OIP3 at 4.5V, 200 mA.
10
85°C
25°C
-40°C
5
4
0
0.5
1
1.5
PAE (%)
50
40
30
0
0.5
85°C
25°C
-40°C
1
1.5
2
2.5
3
3.5
4
FREQUENCY (GHz)
Figure 23. PAE vs Temp and Freq
tuned for optimal OIP3 at 4.5V, 200 mA.
Note:
Bias current for the above charts are quiescent conditions. Actual level may increase depending on amount of RF drive.
2.5
3
3.5
Figure 22. Gain vs. Temp and Freq
tuned for optimal OIP3 at 4.5V, 200 mA.
60
10
2
FREQUENCY (GHz)
70
20
400
15
17
FREQUENCY (GHz)
350
Figure 19. PAE @ P1dB vs. Idq and Vds
at 3.9 GHz.
23
15
0.5
300
20
19
4
250
Idq (mA)
GAIN (dB)
45
P1dB (dBm)
OIP3 (dBm)
29
2
15
Figure 18. PAE @ P1dB vs. Idq and Vds
at 900 MHz.
50
1.5
4.5V
4V
3V
Idq (mA)
Figure 17. PAE @ P1dB vs. Idq and Vds
at 2 GHz.
1
30
20
200
Idq (mA)
15
0.5
35
25
30
30
10
100
PAE (%)
PAE (%)
PAE (%)
40
4
45
45
50
40
40
45
35
35
30
25
20
10
100
150
200
250
300
30
25
20
4.5V
4V
3V
15
40
OIP3 (dBm)
OIP3 (dBm)
OIP3 (dBm)
ATF-521P8 Typical Performance Curves (at 25°C unless specified otherwise)
Tuned for Optimal P1dB
4.5V
4.5V
4V
4V
3V
3V
15
350
10
100
400
150
200
350
25
15
10
100
400
30
30
25
20
4.5V
4V
3V
15
250
300
350
P1db (dBm)
30
P1dB (dBm)
35
200
25
20
4.5V
4V
3V
15
10
100
400
150
200
250
300
350
10
100
400
15
13
13
13
GAIN (dBm)
15
GAIN (dBm)
15
11
9
7
150
200
250
4.5V
4V
3V
150
200
300
350
Id (mA)
Figure 30. Gain vs Ids and Vds at 2 GHz.
5
100
150
200
250
350
400
300
350
4.5V
4V
3V
11
7
400
5
100
150
200
Figure 31. Gain vs Ids and Vds at 900 MHz.
250
300
350
400
Id (mA)
Id (mA)
Figure 32. Gain vs Ids and Vds at 3.9 GHz.
Note:
Bias current for the above charts are quiescent conditions. Actual level may increase depending on amount of RF drive.
300
9
4.5V
4V
3V
7
400
250
Figure 29. P1dB vs. Idq and Vds at 3.9 GHz.
17
4.5V
4V
3V
400
Idq (mA)
17
9
350
20
17
11
300
15
Figure 28. P1dB vs. Idq and Vds at 900 MHz.
Figure 27. P1dB vs. Idq and Vds at 2 GHz.
250
25
Idq (mA)
Idq (mA)
5
100
200
Figure 26. OIP3 vs. Ids and Vds at 3.9 GHz.
35
150
150
Id (mA)
35
10
100
4.5V
4V
3V
20
Figure 25. OIP3 vs. Ids and Vds at 900 MHz.
Figure 24. OIP3 vs. Ids and Vds at 2 GHz.
P1dB (dBm)
300
30
Id (mA)
Id (mA)
GAIN (dBm)
250
35
ATF-521P8 Typical Performance Curves, continued (at 25°C unless specified otherwise)
Tuned for Optimal P1dB
60
55
55
50
35
4.5V
4V
3V
30
25
150
200
PAE (%)
40
20
100
35
45
45
PAE (%)
PAE (%)
50
40
40
35
30
4.5V
4V
3V
25
250
300
350
20
100
400
150
200
250
300
350
20
100
400
150
200
50
32
45
30
Figure 35. PAE @ P1dB vs. Idq and Vds
at 3.9 GHz.
20
30
85°C
25°C
-40°C
1
1.5
2
2.5
3
3.5
GAIN (dB)
P1dB (dBm)
35
20
26
24
85°C
25°C
-40°C
22
4
20
0.5
1
1.5
2
2.5
3
3.5
FREQUENCY (GHz)
FREQUENCY (GHz)
Figure 37. P1dB vs. Temp and Freq
(tuned for optimal P1dB at 4.5V, 200 mA).
Figure 36. OIP3 vs. Temp and Freq
tuned for optimal P1dB at 4.5V, 200 mA.
10
85°C
25°C
-40°C
5
4
0
0.5
1
1.5
40
30
85°C
25°C
-40°C
10
1
1.5
2
2.5
3
3.5
2.5
3
3.5
Figure 38. Gain vs. Temp and Freq
tuned for optimal P1dB at 4.5V, 200 mA.
50
20
2
FREQUENCY (GHz)
60
PAE (%)
400
28
25
4
FREQUENCY (GHz)
Figure 39. PAE vs Temp and Freq
tuned for optimal P1dB at 4.5V.
Note:
Bias current for the above charts are quiescent conditions. Actual level may increase depending on amount of RF drive.
350
15
40
0
0.5
300
250
Idq (mA)
Figure 34. PAE @ P1dB vs. Idq and Vds
at 900 MHz.
Figure 33. PAE @ P1dB vs. Idq and Vds
at 2 GHz.
OIP3 (dBm)
4.5V
4V
3V
25
Idq (mA)
Idq (mA)
15
0.5
30
4
ATF-521P8 Typical Scattering Parameters at 25°C, VDS = 4.5V, IDS = 280 mA
Freq. S11
GHz
Mag.
Ang.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.5
2.0
2.5
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
17.0
18.0
dB
0.613-96.9
33.2
0.780-131.8
30.0
0.831-147.2
27.3
0.855-156.4
25.1
0.860-162.0
23.5
0.878-166.7
22.0
0.888-170.2
20.8
0.887-172.6
19.7
0.894-174.5
18.7
0.886-177.2
17.9
0.892
175.0
14.3
0.883
168.7
12.1
0.890
162.8
10.2
0.884
157.2
8.6
0.890
146.6
6.1
0.893
137.0
4.1
0.896
127.9
2.3
0.906
119.5
0.9
0.882
105.6-0.8
0.887
96.4-1.7
0.887
84.6-2.9
0.882
72.3-3.9
0.878
62.2-5.0
0.894
52.0-6.4
0.888
42.0-7.6
0.884
34.6-8.3
0.830
24.7-9.5
0.708
11.0-9.0
0.790-12.7-10.3
S21
Mag.
Ang.
dB
45.79
141.7-39.5
31.50
121.6-36.7
23.26
111.0-36.2
18.04
104.1-35.4
14.98
99.7-35.2
12.62
95.6-35.0
10.95
92.8-34.6
9.63
90.0-34.3
8.65
87.9-33.7
7.82
85.4-33.8
5.20
76.3-32.8
4.01
68.4-31.2
3.24
61.5-30.0
2.71
54.5-28.9
2.02
40.6-27.0
1.60
27.6-25.5
1.31
15.4-24.2
1.11
3.7-22.9
0.92-9.8-21.3
0.82-22.2-20.1
0.72-33.6-19.3
0.64-45.8-18.5
0.56-57.0-18.0
0.48-67.8-17.8
0.42-76.2-17.3
0.38-84.3-16.6
0.34-92.8-16.1
0.35-99.5-15.4
0.31-93.1-16.4
S12
Mag.
Ang.
S22 Mag. Ang.
0.011
51.3
0.015
37.1
0.015
30.6
0.017
28.2
0.017
27.4
0.018
26.1
0.019
27.4
0.019
28.9
0.021
28.5
0.020
30.3
0.023
34.6
0.027
36.7
0.032
36.8
0.036
39.2
0.045
36.1
0.053
32.4
0.061
28.2
0.071
22.9
0.086
14.5
0.098
7.2
0.109-1.0
0.119-10.5
0.126-19.8
0.130-28.6
0.137-36.1
0.147-42.9
0.156-52.4
0.169-63.8
0.152-82.8
0.317-108.3
36.2
0.423-138.5
33.2
0.466-152.4
31.9
0.483-159.9
30.3
0.488-163.8
29.5
0.496-167.0
28.5
0.497-169.9
27.6
0.500-171.7
27.0
0.501-173.6
26.1
0.502-175.7
25.9
0.502 178.8
23.5
0.492 173.6
20.2
0.490 169.8
18.5
0.494 165.7
16.2
0.505 157.8
13.8
0.529 150.3
11.9
0.551 142.9
10.4
0.570 135.5
9.6
0.567 127.3
6.8
0.585 117.8
6.2
0.593 107.3
5.0
0.617 97.1
3.9
0.636 86.0
2.8
0.662 74.7
2.1
0.697 67.5
0.9
0.732 58.7
0.3
0.752 51.9-1.8
0.816 46.1-2.2
0.660 41.2-4.3
40.0
Freq
Fmin
Γopt
Γopt
30.0
GHz
dB
Mag.
Ang.
0.5
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
1.20
1.30
1.61
1.68
2.12
2.77
2.58
2.85
3.35
0.47
170.00
0.53-177.00
0.61-166.34
0.69-155.85
0.67-146.98
0.71-134.35
0.79-125.22
0.82-115.35
0.73-105.76
Rn
2.8
2.6
2.7
4.0
8.4
19.0
26.7
47.2
65.2
Ga
dB
22.8
20.1
17.3
14.4
11.6
9.9
8.8
7.5
5.7
MSG/MAG and |S21|2 (dB)
Typical Noise Parameters at 25°C, VDS = 4.5V, IDS = 280 mA
MSG/MAG
dB
MSG
20.0
10.0
MAG
0.0
S21
-10.0
-20.0
0
5
10
15
20
FREQUENCY (GHz)
Figure 40. MSG/MAG and |S21|2 vs.
Frequency at 4.5V, 280 mA.
Notes:
1. Fmin values at 2 GHz and higher are based on measurements while the Fmins below 2 GHz have been extrapolated. The Fmin values are based on
a set of 16 noise figure measurements made at 16 different impedances using an ATN NP5 test system. From these measurements a true Fmin is
calculated. Refer to the noise parameter application section for more information.
2. S and noise parameters are measured on a microstrip line made on 0.025 inch thick alumina carrier. The input reference plane is at the end of
the gate lead. The output reference plane is at the end of the drain lead.
ATF-521P8 Typical Scattering Parameters, VDS = 4.5V, IDS = 200 mA
Freq. S11
GHz
Mag.
Ang.
dB
S21
Mag.
Ang.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
2.5
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
52.21
135.6-37.9
33.39
115.7-35.6
23.90
106.3-34.9
18.25
100.5-34.7
15.12
96.6-34.4
12.66
92.9-34.1
10.95
90.5-33.7
9.59
88.0-33.6
8.64
86.2-33.1
7.78
83.7-33.1
5.17
75.7-32.1
4.00
67.8-30.8
3.22
61.3-29.8
2.69
54.5-28.6
2.00
40.7-26.8
1.59
28.3-25.2
1.30
16.4-24.0
1.11
4.8-22.8
0.90-8.8-21.3
0.83-20.1-20.2
0.72-32.1-19.3
0.66-43.7-18.5
0.57-54.1-18.0
0.48-66.2-17.7
0.44-74.0-17.2
0.41-80.6-16.9
0.38-83.4-16.2
0.32-90.1-15.4
0.25-102.3-16.7
0.823-89.9
34.4
0.873-128.7
30.5
0.879-145.5
27.6
0.885-155.1
25.2
0.883-161.1
23.6
0.897-165.9
22.1
0.895-169.5
20.8
0.894-171.9
19.6
0.900-174.7
18.7
0.893-176.6
17.8
0.894
175.3
14.3
0.889
168.5
12.0
0.888
162.6
10.2
0.892
157.0
8.6
0.884
146.5
6.0
0.891
137.0
4.0
0.889
127.9
2.3
0.902
119.6
0.9
0.881
105.6-0.9
0.891
96.0-1.7
0.876
83.9-2.9
0.885
73.1-3.6
0.885
60.9-4.8
0.893
53.0-6.3
0.889
42.2-7.2
0.894
34.3-7.8
0.840
25.0-8.4
0.719
9.1-10.0
0.794-8.1-12.2
dB
Typical Noise Parameters, VDS = 4.5V, IDS = 200 mA
Fmin
Γopt
Γopt
Rn
GHz
dB
Mag.
Ang.
0.5
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0.60
0.72
0.96
1.11
1.44
1.75
1.99
2.12
2.36
0.30
130.00
0.35
150.00
0.47-175.47
0.57-162.03
0.62-150.00
0.69-136.20
0.74-127.35
0.80-116.83
0.69-108.38
2.8
2.6
1.9
2.1
4.5
10.0
17.0
28.5
35.6
S22
Mag. Ang.
MSG/MAG
dB
0.013
46.2
0.017
32.0
0.018
27.0
0.018
25.8
0.019
24.8
0.020
24.2
0.021
24.2
0.021
25.3
0.022
26.2
0.022
27.6
0.025
32.6
0.029
33.6
0.032
35.2
0.037
35.6
0.046
34.4
0.055
30.5
0.063
26.4
0.072
21.0
0.086
13.3
0.098
5.6
0.108-3.2
0.119-12.1
0.126-21.6
0.131-29.9
0.138-36.7
0.143-44.1
0.154-54.3
0.171-64.8
0.147-84.1
0.388-113.0
36.0
0.478-143.2
32.9
0.507-156.0
31.2
0.518-163.1
30.1
0.519-166.7
29.0
0.525-169.6
28.0
0.526-172.2
27.2
0.528-174.0
26.6
0.528-175.6
25.9
0.529-177.7
25.5
0.527 177.2
23.2
0.516 172.1
21.4
0.514 168.1
18.4
0.517 164.0
16.7
0.526 156.0
13.5
0.548 148.3
11.9
0.568 141.0
10.1
0.584 133.5
9.4
0.580 124.9
6.7
0.594 115.8
6.4
0.600 105.3
4.6
0.622 95.0
4.2
0.641 84.1
3.0
0.663 73.1
2.1
0.698 65.7
1.2
0.732 57.4
1.0
0.750 51.0-0.8
0.815 44.5-3.2
0.655 40.4-5.9
40.0
Ga
dB
20.2
18.4
16.5
13.8
11.2
9.8
8.7
7.5
5.7
MSG/MAG and |S21|2 (dB)
Freq
S12
Mag.
Ang.
30.0
MSG
20.0
10.0
MAG
0.0
S21
-10.0
-20.0
0
5
10
15
20
FREQUENCY (GHz)
Figure 41. MSG/MAG and |S21|2 vs.
Frequency at 4.5V, 200 mA.
Notes:
1. Fmin values at 2 GHz and higher are based on measurements while the Fmins below 2 GHz have been extrapolated. The Fmin values are based on
a set of 16 noise figure measurements made at 16 different impedances using an ATN NP5 test system. From these measurements a true Fmin is
calculated. Refer to the noise parameter application section for more information.
2. S and noise parameters are measured on a microstrip line made on 0.025 inch thick alumina carrier. The input reference plane is at the end of
the gate lead. The output reference plane is at the end of the drain lead.
10
ATF-521P8 Typical Scattering Parameters, VDS = 4.5V, IDS = 120 mA
Freq. S11
GHz
Mag.
Ang.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
2.5
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
S21
dB
Mag.
0.913-84.6
34.2
0.900-125.0
30.3
0.896-142.0
27.4
0.893-152.3
25.1
0.882-158.4
23.4
0.895-164.2
21.8
0.893-167.8
20.6
0.895-170.8
19.5
0.897-173.0
18.5
0.895-175.5
17.6
0.893
176.0
14.1
0.889
169.2
11.8
0.882
163.6
10.0
0.888
157.9
8.4
0.883
146.8
5.9
0.885
137.7
3.8
0.892
128.0
2.1
0.894
120.4
0.6
0.880
105.7-1.0
0.876
96.5-1.9
0.879
84.4-3.0
0.889
72.8-3.8
0.881
62.4-5.2
0.893
54.0-6.3
0.891
42.1-7.2
0.888
34.1-8.3
0.845
25.3-9.1
0.828
13.2-11.2
0.827-10.2-11.0
Ang.
S12
dB
51.26
135.4-36.4
32.80
115.4-33.9
23.39
106.1-33.4
17.89
100.3-32.9
14.75
96.3-32.6
12.36
92.9-32.7
10.71
90.5-32.4
9.39
88.0-32.3
8.44
86.1-32.2
7.59
83.6-31.8
5.07
75.3-31.1
3.89
67.8-30.0
3.15
61.2-29.0
2.62
54.6-28.2
1.97
40.7-26.5
1.55
28.2-25.2
1.28
16.7-24.0
1.08
5.1-22.8
0.89-8.7-21.2
0.81-20.8-20.1
0.71-32.7-19.3
0.65-44.3-18.6
0.55-56.0-18.1
0.48-66.6-17.7
0.44-72.6-17.3
0.39-79.2-16.8
0.35-89.6-16.1
0.28-95.9-15.6
0.28-92.5-16.6
Typical Noise Parameters, VDS = 4.5V, IDS = 120 mA
Fmin
Γopt
Γopt
GHz
dB
Mag.
Ang.
Rn
0.5
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0.60
0.72
0.81
0.92
1.24
1.50
1.60
1.88
2.02
0.19
162.00
0.30
164.00
0.44
176.97
0.56-164.98
0.59-155.51
0.70-136.55
0.75-128.59
0.81-117.31
0.68-109.54
3.0
2.6
2.0
2.0
3.4
11.1
16.0
24.0
28.8
Mag.
0.015
49.0
0.020
31.2
0.021
25.3
0.023
23.5
0.023
22.5
0.023
20.6
0.024
20.4
0.024
21.1
0.025
22.1
0.026
23.0
0.028
25.5
0.032
27.9
0.036
30.2
0.039
30.2
0.047
29.7
0.055
26.3
0.063
21.9
0.072
18.2
0.087
10.6
0.099
3.2
0.108-5.2
0.118-13.5
0.125-23.1
0.130-31.4
0.136-38.4
0.144-45.9
0.157-55.0
0.167-64.2
0.147-86.1
0.423-106.6
35.3
0.499-139.4
32.1
0.522-153.4
30.5
0.530-161.1
28.9
0.531-165.0
28.1
0.537-168.4
27.3
0.537-171.2
26.5
0.539-173.1
25.9
0.539-174.8
25.3
0.540-176.9
24.7
0.538 177.4
22.6
0.528 172.2
20.8
0.526 168.1
19.4
0.528 163.9
16.9
0.536 155.7
13.6
0.556 148.1
11.6
0.576 140.5
10.2
0.591 133.1
8.9
0.585 124.3
6.6
0.602 114.9
5.7
0.605 104.5
4.7
0.624 94.2
4.3
0.642 83.4
2.7
0.664 72.4
2.2
0.697 65.1
1.2
0.732 56.7
0.4
0.751 50.4-1.5
0.821 44.0-3.9
0.654 39.9-4.3
Ang.
dB
40.0
Ga
dB
20.0
18.3
15.9
13.6
11.1
9.7
8.7
7.6
5.6
MSG/MAG and |S21|2 (dB)
Freq
S22 MSG/MAG
Mag.
Ang.
30.0
MSG
20.0
10.0
MAG
0.0
S21
-10.0
-20.0
0
5
10
15
20
FREQUENCY (GHz)
Figure 42. MSG/MAG and |S21|2 vs.
Frequency at 4.5V, 120 mA.
Notes:
1. Fmin values at 2 GHz and higher are based on measurements while the Fmins below 2 GHz have been extrapolated. The Fmin values are based on
a set of 16 noise figure measurements made at 16 different impedances using an ATN NP5 test system. From these measurements a true Fmin is
calculated. Refer to the noise parameter application section for more information.
2. S and noise parameters are measured on a microstrip line made on 0.025 inch thick alumina carrier. The input reference plane is at the end of
the gate lead. The output reference plane is at the end of the drain lead.
11
ATF-521P8 Typical Scattering Parameters, VDS = 4V, IDS = 200 mA
Freq. S11
GHz
Mag.
Ang.
dB
S21
Mag.
Ang.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
2.5
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
51.89
134.8-37.7
32.88
115.0-35.4
23.48
105.8-35.1
17.91
100.1-34.4
14.80
96.3-34.2
12.37
92.7-34.2
10.74
90.5-33.6
9.39
88.1-33.5
8.47
85.9-33.3
7.61
84.0-32.9
5.06
75.7-32.1
3.91
68.1-30.7
3.15
61.7-29.5
2.63
55.1-28.4
1.97
41.5-26.7
1.57
29.4-25.1
1.28
17.7-23.9
1.09
6.3-22.6
0.90-7.1-21.1
0.82-19.3-20.1
0.72-30.9-19.2
0.65-42.8-18.6
0.58-53.3-18.0
0.49-63.4-17.7
0.44-73.5-17.2
0.39-80.2-16.8
0.36-85.3-16.2
0.32-90.9-15.5
0.31-95.1-16.6
0.843-90.5
34.3
0.879-129.3
30.3
0.888-146.1
27.4
0.892-155.6
25.1
0.886-161.5
23.4
0.896-165.7
21.8
0.897-169.5
20.6
0.898-172.2
19.5
0.896-174.9
18.6
0.896-176.7
17.6
0.898
175.2
14.1
0.887
168.0
11.8
0.893
162.8
10.0
0.886
156.9
8.4
0.887
146.6
5.9
0.894
136.8
3.9
0.898
127.4
2.1
0.896
119.7
0.7
0.879
105.4-0.9
0.888
95.0-1.7
0.872
84.1-2.9
0.880
72.4-3.8
0.875
60.4-4.8
0.908
52.4-6.2
0.898
41.3-7.1
0.888
34.1-8.2
0.815
24.1-8.9
0.725
11.3-9.9
0.792-9.8-10.2
dB
Typical Noise Parameters, VDS = 4V, IDS = 200 mA
Fmin
Γopt
Γopt
Rn
GHz
dB
Mag.
Ang.
0.5
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0.67
0.74
0.96
1.24
1.44
1.62
1.83
1.99
2.21
0.21
155.00
0.30
164.00
0.46-176.61
0.57-162.19
0.62-152.18
0.69-135.43
0.74-127.94
0.82-117.20
0.71-108.96
2.8
2.6
2.1
2.8
4.5
10.0
17.0
27.7
35.3
Mag.
S22
Ang.
MSG/MAG
dB
0.013
46.5
0.017
32.1
0.018
26.0
0.019
25.1
0.020
24.6
0.020
24.1
0.021
24.7
0.021
24.4
0.022
26.5
0.023
26.3
0.025
29.9
0.029
35.2
0.034
35.8
0.038
35.8
0.046
33.2
0.056
29.6
0.064
25.5
0.074
20.4
0.088
12.4
0.099
4.7
0.110-4.3
0.118-12.9
0.126-22.8
0.130-31.4
0.138-38.0
0.144-45.6
0.156-54.7
0.167-66.0
0.147-84.8
0.408-118.1
36.0
0.507-146.1
32.9
0.539-158.3
31.2
0.549-164.8
29.7
0.551-168.2
28.7
0.556-170.9
27.9
0.557-173.5
27.1
0.559-175.2
26.5
0.559-176.9
25.9
0.560-178.7
25.2
0.558 176.0
23.1
0.547 170.9
21.3
0.545 166.9
18.9
0.547 162.6
16.3
0.554 154.3
13.6
0.572 146.6
11.9
0.590 139.0
10.3
0.603 131.6
8.9
0.594 122.7
6.6
0.609 113.2
6.1
0.610 102.9
4.4
0.629 92.6
3.8
0.647 81.9
2.8
0.666 71.0
2.6
0.699 64.0
1.5
0.734 55.9
0.5
0.750 49.3-1.7
0.809 43.5-3.1
0.652 39.7-4.2
40.0
Ga
dB
20.1
18.4
16.4
13.9
11.4
10.0
8.7
7.7
5.9
MSG/MAG and |S21|2 (dB)
Freq
S12
Mag.
Ang.
30.0
MSG
20.0
10.0
MAG
0.0
S21
-10.0
-20.0
0
5
10
15
20
FREQUENCY (GHz)
Figure 43. MSG/MAG and |S21|2 vs.
Frequency at 4V, 200 mA.
Notes:
1. Fmin values at 2 GHz and higher are based on measurements while the Fmins below 2 GHz have been extrapolated. The Fmin values are based on
a set of 16 noise figure measurements made at 16 different impedances using an ATN NP5 test system. From these measurements a true Fmin is
calculated. Refer to the noise parameter application section for more information.
2. S and noise parameters are measured on a microstrip line made on 0.025 inch thick alumina carrier. The input reference plane is at the end of
the gate lead. The output reference plane is at the end of the drain lead.
12
ATF-521P8 Typical Scattering Parameters, VDS = 3V, IDS = 200 mA
Freq. S11
GHz
Mag.
Ang.
dB
S21
Mag.
Ang.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
2.5
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
48.20
132.4-36.8
29.66
113.2-34.9
21.06
104.4-34.1
16.00
99.1-34.0
13.20
95.6-33.6
11.00
92.3-33.2
9.51
90.2-33.2
8.35
87.8-33.0
7.51
86.3-32.9
6.76
84.2-32.5
4.50
76.4-31.5
3.49
69.1-29.9
2.82
63.0-29.0
2.35
56.9-27.7
1.76
43.8-26.1
1.41
32.1-24.5
1.16
21.6-23.4
0.98
10.3-22.1
0.83-2.3-20.7
0.76-13.0-19.8
0.67-26.0-18.9
0.60-36.3-18.3
0.54-47.4-17.8
0.47-57.9-17.6
0.42-62.8-17.2
0.39-74.7-16.8
0.33-78.2-16.3
0.33-90.8-15.8
0.26-92.8-17.0
0.867-94.6
33.7
0.894-132.9
29.4
0.899-148.2
26.5
0.896-157.2
24.1
0.892-162.8
22.4
0.910-167.4
20.8
0.906-170.8
19.6
0.902-173.6
18.4
0.907-175.2
17.5
0.902-177.7
16.6
0.900
174.2
13.1
0.896
168.1
10.8
0.896
162.3
9.0
0.887
156.7
7.4
0.890
145.7
4.9
0.898
136.3
3.0
0.896
127.4
1.3
0.904
119.4-0.2
0.877
104.9-1.6
0.883
94.8-2.4
0.877
83.1-3.5
0.875
71.7-4.4
0.863
60.6-5.4
0.910
51.6-6.5
0.868
40.9-7.5
0.863
33.4-8.1
0.835
25.2-9.6
0.720
11.2-9.5
0.780-7.7-11.6
dB
Typical Noise Parameters, VDS = 3V, IDS = 200 mA
Fmin
Γopt
Γopt
GHz
dB
Mag.
Ang.
Rn
0.5
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0.66
0.72
0.87
1.00
1.32
1.49
1.59
1.79
1.96
0.22
147.00
0.30
160.00
0.42-179.94
0.59-163.63
0.63-153.81
0.72-135.10
0.74-128.97
0.78-117.68
0.70-110.04
2.9
2.6
1.9
1.6
3.7
10.0
15.0
25.1
29.2
Mag.
S22
Ang.
MSG/MAG
dB
0.014
45.1
0.018
28.5
0.020
23.2
0.020
23.7
0.021
24.5
0.022
22.9
0.022
23.9
0.022
24.6
0.023
27.0
0.024
26.9
0.027
32.7
0.032
32.9
0.036
34.3
0.041
35.0
0.050
32.2
0.059
28.3
0.068
23.5
0.078
17.7
0.092
9.0
0.102
1.3
0.113-7.3
0.121-16.6
0.128-25.1
0.132-33.6
0.138-40.4
0.144-47.6
0.154-56.8
0.161-67.6
0.142-85.1
0.482-132.4
35.4
0.601-154.2
32.2
0.636-163.8
30.2
0.647-169.2
29.0
0.650-171.9
28.0
0.655-174.4
27.0
0.657-176.7
26.4
0.658-178.2
25.8
0.660-179.5
25.1
0.659 178.6
24.5
0.656 173.4
22.2
0.647 167.9
20.4
0.642 163.7
18.6
0.643 159.2
15.6
0.645 150.4
12.9
0.659 142.1
11.3
0.671 134.3
9.5
0.677 126.6
8.5
0.651 117.0
5.9
0.661 107.2
5.3
0.657 96.8
4.0
0.670 86.7
3.1
0.680 76.2
1.9
0.694 65.9
2.3
0.721 59.3
0.2
0.748 51.3-0.2
0.758 44.9-2.1
0.818 39.4-2.6
0.655 37.1-5.7
40.0
Ga
dB
20.0
18.3
16.0
13.7
11.3
9.9
8.5
7.6
5.6
MSG/MAG and |S21|2 (dB)
Freq
S12
Mag.
Ang.
30.0
MSG
20.0
10.0
MAG
0.0
S21
-10.0
-20.0
0
5
10
15
20
FREQUENCY (GHz)
Figure 44. MSG/MAG and |S21|2 vs.
Frequency at 3V, 200 mA.
Notes:
1. Fmin values at 2 GHz and higher are based on measurements while the Fmins below 2 GHz have been extrapolated. The Fmin values are based on
a set of 16 noise figure measurements made at 16 different impedances using an ATN NP5 test system. From these measurements a true Fmin is
calculated. Refer to the noise parameter application section for more information.
2. S and noise parameters are measured on a microstrip line made on 0.025 inch thick alumina carrier. The input reference plane is at the end of
the gate lead. The output reference plane is at the end of the drain lead.
13
ATF-521P8 Applications Information
Description
Avago Technologies' ATF-521P8 is an enhancement
mode PHEMT designed for high linearity and medium
power applications. With an OIP3 of 42 dBm and a 1dB
compression point of 26 dBm, ATF-521P8 is well suited
as a base station transmit driver or a first or second
stage LNA in a receive chain. Whether the design is for a
W-CDMA, CDMA, or GSM basestation, this device deliv‑
ers good linearity in the form of OIP3 or ACLR, which is
required for standards with high peak to average ratios.
Application Guidelines
The ATF-521P8 device operates as a normal FET requir‑
ing input and output matching as well as DC biasing.
Unlike a depletion mode transistor, this enhancement
mode device only requires a single positive power sup‑
ply, which means a positive voltage is placed on the
drain and gate in order for the transistor to turn on. This
application note walks through the RF and DC design
employed in a single FET amplifier. Included in this de‑
scription is an active feedback scheme to accomplish
this DC biasing.
limiting the input match close to S11*. Normally, the in‑
put return loss of a single ended amplifier is not critical
as most basestation LNA and driver amplifiers are in a
balanced configuration with 90° (quadrature) couplers.
Proceeding from the same premise, the output match
of this device becomes much simpler. As background
information, it is important to note that OIP3 is largely
dependant on the output match and that output return
loss is also required to be greater than 10 dB. So, Figure
2 shows how both good output return loss and good
linearity could be achieved simultaneously with the
same impedance point.
Of course, these points are valid only at 2 GHz, and
other frequencies will follow the same design rules but
will have different locations. Also, the location of these
points is largely due to the manufacturing process and
partly due to IC layout, but in either case beyond the
scope of this application note.
RF Input & Output Matching
In order to achieve maximum linearity, the appropriate
input (Γs) and output (ΓL) impedances must be pre‑
sented to the device. Correctly matching from these
impedances to 50Ωs will result in maximum linearity.
Although ATF-521P8 may be used in other impedance
systems, data collected for this data sheet is all refer‑
enced to a 50Ω system.
The input load pull parameter at 2 GHz is shown in Fig‑
ure 1 along with the optimum S11 conjugate match.
ΓS
S11*
3 dB
5 dB
9 dB
B
16 d
s
n Los
Retur
ΓL
S22*
Figure 2. Output Match at 2 GHz.
Figure 2. Output Match
at 2 GHz.
Once a designer has chosen the proper input and out‑
put impedance points, the next step is to choose the
correct topology to accomplish this match. For example
to perform the above output impedance transformation
from 50Ω to the given load parameter of 0.53∠-176°,
two possible solutions exist. The first potential match
is a high pass configuration accomplished by a shunt
inductor and a series capacitor shown in Figure 3 along
with its frequency response in Figure 4.
RFin
8
Figure1.1.Input
Input
Match
for ATF-521P
GHz.
Figure
Match
for ATF-521P8
at
2 GHz.
Thus, it should be obvious from the illustration above
that if this device is matched for maximum return loss
i.e. S11*, then OIP3 will be sacrificed. Conversely, if ATF521P8 is matched for maximum linearity, then return
loss will not be greater than 10 dB. For most applica‑
tions, a designer requires VSWR greater than 2:1, hence
14
C1
RFout
L1
Figure 3. High Pass Circuit Topology.
Figure 7 displays the input and output matching se‑
lected for ATF-521P8. In this example the input and out‑
put match both essentially function as high pass filters,
but the high frequency gain of the device rolls off pre‑
cipitously giving a narrow band frequency response, yet
still wide enough to accommodate a CDMA or WCDMA
transmit band. For more information on RF matching
techniques refer to MGA-53543 application note.
Amp
Frequency
Figure 4. High Pass Frequency Response.
The second solution is a low pass configuration with a
shunt capacitor and a series inductor shown in Figure 5
and 6.
RFin
L1
RFout
C1
Once the RF matching has been established, the next
step is to DC bias the device. A passive biasing example
is shown in Figure 8. In this example the voltage drop
across resistor R3 sets the drain current (Id) and is calcu‑
lated by the following equation:
R3 = Vdd – Vds
Figure 5. Low Pass Circuit Topology.
Ids + Ibb
Amp
Frequency
Figure 6. Low Pass Frequency Response.
The actual values of these components may be calcu‑
lated by hand on a Smith Chart or more accurately done
on simulation software such as ADS. There are some
advantages and disadvantages of choosing a high pass
versus a low pass. For instance, a high pass circuit cuts
off low frequency gain, which narrows the usable band‑
width of the amplifier, but consequently helps avoid
potential low frequency instability problems. A low pass
match offers a much broader frequency response, but it
has two major disadvantages. First it has the potential
for low frequency instability, and second it creates the
need for an extra DC blocking capacitor on the input
in order to isolate the device gate from the preceding
stages.
RFin
Passive Bias [1]
Zo
R1 = Vg
Ibb
R2 = (Vdd – Vg) x R1
Vg
Amp
Amp
+
Frequency
=
Frequency
Figure 7. Input and Output Match for ATF‑521P8 at 2 GHz.
15
Total Response
Output Match
Amp
+
Frequency
RFout
L1
ATF-521P8
Input Match
(3)
Often the series resistor, R4, is added to enhance the
low frequency stability. The complete passive bias ex‑
ample may be found in reference [1].
C2
Amp
(2)
p
Zo
52
p
where,
Vdd is the power supply voltage;
Vds is the device drain to source voltage;
Ids is the device drain to source current;
Ibb for DC stability is 10X the typical gate current;
A voltage divider network with R1 and R2 establishes
the typical gate bias voltage (Vg).
C3
C1
(1)
Frequency
C1
INPUT
C4
Q1
Zo
L1 C2
R4
To calculate the values of R1, R2, R3, and R4 the follow‑
ing parameters must be know or chosen first:
Ids is the device drain-to-source current;
IR is the Reference current for active bias;
Vdd is the power supply voltage available;
Vds is the device drain-to-source voltage;
Vg is the typical gate bias;
Vbe1 is the typical Base-Emitter turn on voltage for Q1 &
Q2;
OUTPUT
Zo
L4 C5
C3
R3
C6
Ib
R5
R2
R1
Vdd
Figure 8. Passive Biasing.
Therefore, resistor R3, which sets the desired device
drain current, is calculated as follows:
Active Bias [2]
Due to very high DC power dissipation and small pack‑
age constraints, it is recommended that ATF-521P8 use
active biasing. The main advantage of an active biasing
scheme is the ability to hold the drain to source current
constant over a wide range of temperature variations.
A very inexpensive method of accomplishing this is to
use two PNP bipolar transistors arranged in a current
mirror configuration as shown in Figure 9. Due to resis‑
tors R1 and R3, this circuit is not acting as a true current
mirror, but if the voltage drop across R1 and R3 is kept
identical then it still displays some of the more useful
characteristics of a current mirror. For example, transis‑
tor Q1 is configured with its base and collector tied
­together. This acts as a simple PN junction, which helps
temperature compensate the Emitter-Base junction of
Q2.
R2
Q1
R1
VE
C6
R3
Vds
C3
R6
C8
L3
L2
RFin
C1
L1
C2
Figure 9. Active Bias Circuit.
16
C7
2
2PL
RFout
7
ATF-521P8
The next three equations are used to calculate the rest
of the biasing resistors for Figure 9. Note that the volt‑
age drop across R1 must be set equal to the voltage
drop across R3, but with a current of IR.
R1 = Vdd – Vds
IR
(5)
R2 sets the bias current through Q1.
R4 = Vg
IC 2
C5
R5
IC2 is chosen for stability to be 10 times the typical gate
current and also equal to the reference current IR.
(6)
p
R4 sets the gate voltage for ATF‑521P8.
Q2
C4
(4)
p
where,
R2 = Vds – Vbe1
IR
Vdd
R4 Vg
R3 = Vdd – Vds
Ids + IC2
L4
(7)
p
Thus, by forcing the emitter voltage (VE) of transistor
Q1 equal to Vds, this circuit regulates the drain current
similar to a current mirror. As long as Q2 operates in the
forward active mode, this holds true. In other words, the
Collector-Base junction of Q2 must be kept reversed
biased.
PCB Layout
A recommended PCB pad layout for the Leadless Plastic
Chip Carrier (LPCC) package used by the ATF-521P8 is
shown in Figure 10. This layout provides plenty of plat‑
ed through hole vias for good thermal and RF ground‑
ing. It also provides a good transition from microstrip to
the device package. For more detailed dimensions refer
to Section 9 of the data sheet.
This simplifies RF grounding by reducing the amount of
inductance from the source to ground. It is also recom‑
mended to ground pins 1 and 4 since they are also con‑
nected to the device source. Pins 3, 5, 6, and 8 are not
connected, but may be used to help dissipate heat from
the package or for better alignment when soldering the
device.
This three-layer board (Figure 12) contains a 10-mil layer
and a 52-mil layer separated by a ground plane. The first
layer is Getek RG200D material with dielectric constant
of 3.8. The second layer is for mechanical rigidity and
consists of FR4 with dielectric constant of 4.2.
High Linearity Tx Driver
Figure 10. Microstripline Layout.
RF Grounding
Pin 8
Pin 7 (Drain)
Pin 6
Pin 5
Source
(Thermal/RF Gnd)
Unlike SOT packages, ATF-521P8 is housed in a lead‑
less package with the die mounted directly to the lead
frame or the belly of the package shown in Figure 11.
Pin 1 (Source)
Pin 2 (Gate)
Pin 3
The need for higher data rates and increased voice
capacity gave rise to a new third generation standard
know as Wideband CDMA or UMTS. This new standard
requires higher performance from radio components
such as higher dynamic range and better linearity. For
example, a WCDMA waveform has a very high peak to
average ratio which forces amplifiers in a transmit chain
to have very good Adjacent Channel Leakage power Ra‑
tio or ACLR, or else operate in a backed off mode. If the
amplifier is not backed off then the waveform is com‑
pressed and the signal becomes very nonlinear.
Pin 4 (Source)
This application example presents a highly linear trans‑
mit drive for use in the 2.14GHz frequency range. Using
the RF matching techniques described earlier, ATF521P8 is matched to the following input and output
impedances:
Bottom View
Figure 11. LPCC Package for ATF-521P8.
BCV62B
C5
R2
R3
0
R4
R1
C6
R6
R5
C4
C3
C1
C2
short
Figure 12. ATF-521P8 demoboard.
17
L1
0
L4
L3
L2
J1
C7
C8
J2
Input
Match
Output
Match
2PL
50 Ohm
50 Ohm
S11* = 0.89∠ -169
ΓL = 0.53∠ -176
Figure 13. ATF-521P8 Matching.
Resistor
Calculated
Actual
R1
50Ω
49.9Ω
R2
385Ω
383Ω
R3
2.38Ω
R4
62Ω
2.37Ω
61.9Ω
Table 1. Resistors for Active Bias.
As described previously the input impedance must be
matched to S11* in order to guarantee return loss great‑
er than 10 dB. A high pass network is chosen for this
match. The output is matched to ΓL with another high
pass network. The next step is to choose the proper DC
biasing conditions. From the data sheet, ATF-521P8 pro‑
duces good linearity at a drain current of 200mA and
a drain to source voltage of 4.5V. Thus to construct the
active bias circuit described, the following parameters
are given:
Ids = 200 mA
IR = 10 mA
Vdd = 5 V
Vds = 4.5V
Vg = 0.62V
Vbe1 = 0.65 V
The entire circuit schematic for a 2.14 GHz Tx driver
amplifier is shown below in Figure 14. Capacitors C4,
C5, and C6 are added as a low frequency bypass. These
terminate second order harmonics and help improve
linearity. Resistors R5 and R6 also help terminate low
frequencies, and can prevent resonant frequencies be‑
tween the two bypass capacitors.
Performance of ATF-521P8 at 2140 MHz
ATF-521P8 delivers excellent performance in the WCD‑
MA frequency band. With a drain-to-source voltage of
4.5V and a drain current of 200 mA, this device has 16.5
dB of gain and 1.55 dB of noise figure as show in Figure
15.
Using equations 4, 5, 6, and 7, the biasing resistor values
are calculated in column 2 of table 1, and the actual val‑
ues used are listed in column 3.
IR
R2=383Ω
R1=49.9Ω
Q1
Vg
Q2
R4=61.9Ω
C4=1µF
C3=4.7pF
Vbe1+
C5=1µF
Vds
R3=2.37Ω
IC2
C6=.1µF
R6=1.2Ω
R5=10Ω
C7=150pF
L3=39nH
L2=12nH
RFin
C1=1.2pF
L1=1.0nH
C2=1.5nH
Figure 14. 2140 MHz Schematic.
18
+5V
C8=1.5pF
2
2PL
7
ATF-521P8
RFout
L4=3.9nH
45
20
Gain
40
OIP3 (dBm)
GAIN and NF (dB)
15
10
35
30
5
NF
0
1.6
1.8
2.0
2.2
2.4
25
2060 2080 2100 2120 2140 2160 2180 2200
2.6
FREQUENCY (MHz)
FREQUENCY (GHz)
Input and output return loss are both greater that 10 dB.
Although somewhat narrowband, the response is ad‑
equate in the frequency range of 2110 MHz to 2170 MHz
for the WCDMA downlink. If wider band response is
need, using a balanced configuration improves return
loss and doubles OIP3.
Figure17.
17.OIP3
OIP3vs.
vs.Frequency
Frequency ininWCDMA
Figure
WCDMA Band (Pout = 12 dBm).
Band (Pout = 12 dBm).
-30
-35
-40
ACLR (dB)
Figure15.
15.Gain
Gainand
and Noise
Noise Figure
Figure vs.
Figure
vs. Frequency.
Frequency.
-45
-50
-55
0
INPUT AND OUTPUT
RETURN LOSS (dB)
-60
S11
-65
-5
-3
2
7
12
17
22
Pout (dBm)
Figure18.
18. ACLR
ACLR vs.
vs. Pout
Pout at
Figure
at55MHz
MHzOffset.
Offset.
-10
S22
-15
1.6
C1=1.2 pF
Phycomp 0402CG129C9B200
C2,C8=1.5 pF Phycomp 0402CG159C9B200
1.8
2.0
2.2
2.4
2.6
C3=4.7 pF
Phycomp 0402CG479C9B200
C4,C6=.1 µF
Phycomp 06032F104M8B200
Figure
Inputand
andOutput
Output Return
Figure
16.16.Input
ReturnLoss
Lossvs.vs. ­Frequency.
Frequency.
C5=1 µF
AVX 0805ZC105KATZA
Perhaps the most critical system level specification for
the ATF‑521P8 lies in its distortion-less output power.
Typically, amplifiers are characterized for linearity by
measuring OIP3. This is a two-tone harmonic measure‑
ment using CW signals. But because WCDMA is a modu‑
lated waveform spread across 3.84 MHz, it is difficult
to correlated good OIP3 to good ACLR. Thus, both are
measured and presented to avoid ambiguity.
C7=150 pF
Phycomp 0402CG151J9B200
L1=1.0 nH
TOKO LL1005-FH1n0S
L2=12 nH
TOKO LL1005-FS12N
L3=39 nH
TOKO LL1005-FS39
L4=3.9 nH
TOKO LL1005-FH3N9S
R1=49.9Ω
RohmRK73H1J49R9F
R2=383Ω
Rohm RK73H1J3830F
R3=2.37Ω
Rohm RK73H1J2R37F
R4=61.9Ω
Rohm RK73H1J61R9F
R5=10Ω
Rohm RK73H1J10R0F
R6=1.2Ω
Rohm RK73H1J1R21F
Q1, Q2
Philips BCV62B
J1, J2
142-0701-851
FREQUENCY (GHz)
Table 2. 2140 MHz Bill of Material.
19
Using the 3GPP standards document Release 1999 ver‑
sion 2002-6, the following channel configuration was
used to test ACLR. This table contains the power levels
of the main channels used for Test Model 1. Note that
the DPCH can be made up of 16, 32, or 64 separate
channels each at different power levels and timing off‑
sets. For a listing of power levels, channelization codes
and timing offset see the entire 3GPP TS 25.141 V3.10.0
(2002-06) standards document at: http://www.3gpp.
org/specs/specs.htm
3GPP TS 25.141 V3.10.0 (2002-06) Type
Pwr (dB)
P-CCPCH+SCH-10
Primary CPICH-10
PICH-18
S-CCPCH containing PCH (SF=256)-18
DPCH-64ch (SF=128)-1.1
where,
θb –a is the board to ambient thermal resistance;
θch–b is the channel to board thermal resistance.
The board to ambient thermal resistance thus becomes
very important for this is the designer’s major source of
heat control. To demonstrate the influence of θb-a, ther‑
mal resistance is measured for two very different sce‑
narios using the ATF-521P8 demoboard. The first case
is done with just the demoboard by itself. The second
case is the ATF demoboard mounted on a chassis or
metal casing, and the results are given below:
ATF Demoboard
θ b-a
PCB 1/8" Chassis
PCB no HeatSink 10.4°C/W
32.9°C/W
Table 4. Thermal resistance measurements.
Table 3. ACLR Channel Power Configuration.
Thermal Design
When working with medium to high power FET devices,
thermal dissipation should be a large part of the design.
This is done to ensure that for a given ambient tem‑
perature the transistor’s channel does not exceed the
maximum rating, TCH, on the data sheet. For example,
ATF‑521P8 has a maximum channel temperature of
150°C and a channel to board thermal resistance of
45°C/W, thus the entire thermal design hinges from
these key data points. The question that must be an‑
swered is whether this device can operate in a typical
environment with ambient temperature fluctuations
from -25°C to 85°C. From Figure 19, a very useful equa‑
tion is derived to calculate the temperature of the chan‑
nel for a given ambient temperature. These calculations
are all incorporated into Avago Technologies AppCAD.
Pdiss = Vds x Ids
Tch
(channel)
θch-b
Tb (board
or belly
of the part)
θb-s
Ts (sink)
θs-a
Ta (ambient)
Therefore calculating the temperature of the channel
for these two scenarios gives a good indication of what
type of heat sinking is needed.
Case 1: Chassis Mounted @ 85°C
Tch = P x (θch-b + θb-a) + Ta
=.9W x (45+10.4)°C/W +85°C
Tch = 135°C
Case 2: No Heatsink @ 85°C
Tch = P x (θch-b + θb-a) + Ta
=.9W x (45+32.9)°C/W + 85°C
Tch = 155°C
In other words, if the board is mounted to a chassis, the
channel temperature is guaranteed to be 135°C safely
below the 150°C maximum. But on the other hand, if
no heat sinking is used and the θb-a is above 27°C/W
(32.9°C/W in this case), then the power must be derated
enough to lower the temperature below 150°C. This can
be better understood with Figure 20 below. Note power
is derated at 13 mW/°C for the board with no heat sink
and no derating is required for the chassis mounted
board until an ambient temperature of 100°C.
Figure 19. Equivalent Circuit for Thermal ­Resistance.
Pdiss
(W)
Hence very similar to Ohms Law, the temperature of the
channel is calculated with equation 8 below.
0.9W
TCH = Pdiss (θch–b + θ b–s + θs–a ) + Tamb
20
No Heatsink
(13 mW/°C)
(8)
If no heat sink is used or heat sinking is incorporated
into the PCB board then equation 8 may be reduced to:
TCH = Pdiss (θch–b + θ b–a ) + Tamb (9)
Mounted on Chassis
(18 mW/°C)
0
81
100
150
Tamb (°C)
Figure 20. Derating for ATF- 521P8.
Thus, for reliable operation of ATF-521P8 and extended
MTBF, it is recommended to use some form of thermal
heatsinking. This may include any or all of the following
suggestions:
•
•
•
•
•
•
•
Maximize vias underneath and around package;
Maximize exposed surface metal;
Use 1 oz or greater copper clad;
Minimize board thickness;
Metal heat sinks or extrusions;
Fans or forced air;
Mount PCB to Chassis.
References
[1] Ward, A. (2001) Avago Technologies ATF‑54143 Low
Noise Enhancement Mode Pseudomorphic HEMT in a
Surface Mount Plastic Package, 2001 [Internet], Avail‑
able from:
<http://www.avagotech.com>
[2] Biasing Circuits and ­Considerations for GaAs MESFET Power Amplifiers, 2001 [Internet], Available from:
<http://www.rf-solutions.com/pdf/AN‑0002_ajp.pdf>
[Accessed 22 August, 2002]
Device Models
Summary
A high linearity Tx driver ­amplifier for WCDMA has been
presented and designed using Agilent’s ATF-521P8. This
includes RF, DC and good thermal dissipation practices
for reliable lifetime operation. A summary of the typical
performance for ATF-521P8 demoboard at 2140 MHz is
as follows:
Demo Board Results at 2140 MHz
Refer to Avago Technologies' Web Site:
www.avagotech.com
Ordering Information
Part Number
No. of Devices
ATF-521P8-TR1
ATF-521P8-TR2
ATF-521P8-BLK
3000
10000
100
Gain
16.5 dB
OIP3
41.2 dBm
ACLR-58 dBc
P1dB
24.8 dBm
NF
1.55 dB
2 x 2 LPCC (JEDEC DFP-N) Package Dimensions
D1
D
pin1
P
pin1
8
1
2
e
E1
3
R
2PX
4
5
Top View
Bottom View
A1
A
A2
End View
End View
DIMENSIONS
SYMBOL
A
A1
A2
b
D
D1
E
E1
e
MIN.
0.70
0
0.225
1.9
0.65
1.9
1.45
NOM.
0.75
0.02
0.203 REF
0.25
2.0
0.80
2.0
1.6
0.50 BSC
DIMENSIONS ARE IN MILLIMETERS
21
E
6
b
L
A
7
MAX.
0.80
0.05
0.275
2.1
0.95
2.1
1.75
Container
7” Reel
13”Reel
antistatic bag
PCB Land Pattern and Stencil Design
2.72 (107.09)
2.80 (110.24)
0.70 (27.56)
0.63 (24.80)
0.25 (9.84)
0.22 (8.86)
0.25 (9.84)
PIN 1
φ0.20 (7.87)
0.32 (12.79)
PIN 1
0.50 (19.68)
0.50 (19.68)
1.54 (60.61)
1.60 (62.99)
Solder
mask
RF
transmission
line
+
0.25 (9.74)
0.28 (10.83)
0.60 (23.62)
0.63 (24.80)
0.72 (28.35)
0.80 (31.50)
0.15 (5.91)
0.55 (21.65)
Stencil Layout (top view)
PCB Land Pattern (top view)
Device Orientation
4 mm
REEL
8 mm
CARRIER
TAPE
USER
FEED
DIRECTION
COVER TAPE
22
2PX
2PX
2PX
2PX
Tape Dimensions
P0
P
D
P2
E
F
W
+
+
D1
Tt
t1
K0
10° Max
10° Max
A0
DESCRIPTION
CAVITY
PERFORATION
CARRIER TAPE
COVER TAPE
DISTANCE
B0
SYMBOL
SIZE (mm)
SIZE (inches)
LENGTH
A0
2.30 ± 0.05
0.091 ± 0.004
WIDTH
B0
2.30 ± 0.05
0.091 ± 0.004
DEPTH
1.00 ± 0.05
0.039 ± 0.002
PITCH
BOTTOM HOLE DIAMETER
K0
P
D1
4.00 ± 0.10
1.00 + 0.25
0.157 ± 0.004
0.039 + 0.002
DIAMETER
D
1.50 ± 0.10
0.060 ± 0.004
PITCH
4.00 ± 0.10
0.157 ± 0.004
POSITION
P0
E
1.75 ± 0.10
0.069 ± 0.004
WIDTH
W
THICKNESS
t1
8.00 + 0.30
8.00 – 0.10
0.254 ± 0.02
0.315 ± 0.012
0.315 ± 0.004
0.010 ± 0.0008
WIDTH
C
5.4 ± 0.10
0.205 ± 0.004
TAPE THICKNESS
Tt
0.062 ± 0.001
0.0025 ± 0.0004
CAVITY TO PERFORATION
(WIDTH DIRECTION)
F
3.50 ± 0.05
0.138 ± 0.002
CAVITY TO PERFORATION
P2
2.00 ± 0.05
0.079 ± 0.002
(LENGTH DIRECTION)
For product information and a complete list of distributors, please go to our web site: www.avagotech.com
Avago, Avago Technologies, and the A logo are trademarks of Avago Technologies, Pte. in the United States and other countries.
Data subject to change. Copyright © 2006 Avago Technologies Pte. All rights reserved.
5988-9974EN - May 31, 2006
23