Download RF3934D RF OUT VD RF IN

RF3934D
RF3934D
120W GaN on
SiC Power
Amplifier Die
120W GaN ON SIC POWER AMPLIFIER DIE
Package: Die
RF OUT
VD
Features

Broadband Operation DC-4GHz

Advanced GaN HEMT Technology


RF IN
VG
Packaged Small Signal
Gain=13dB at 2GHz
48V Typical Packaged
Performance
Output Power 140W at P3dB
 Drain Efficiency 60% at P3dB
Large Signal Models Available




Chip Dimensions:
0.96mmx4.57mmx0.10mm





Functional Block Diagram
Active Area Periphery: 22.2mm
Product Description
Applications

GND
Commercial Wireless
Infrastructure
Cellular and WiMAX
Infrastructure
Civilian and Military Radar
General Purpose Broadband
Amplifiers
Public Mobile Radios
Industrial, Scientific, and
Medical
The RF3934D is a 48V, 120W, GaN on SiC high power discrete amplifier
die designed for commercial wireless infrastructure, cellular and WiMAX
infrastructure, industrial/scientific/medical and general purpose broadband amplifier applications. Using an advanced high power density Gallium Nitride (GaN) semiconductor process, the RF3934D is able to
achieve high efficiency and flat gain over a broad frequency range in a single amplifier design with proper packaging and assembly. The RF3934D is
an unmatched 0.5ìm gate, GaN transistor die suitable for many applications with >51dBm saturated power, >60% saturated drain efficiency, and
>13dB small signal gain at 2GHz.
Ordering Information
RF3934D
120W GaN on SiC Power Amplifier Die
Optimum Technology Matching® Applied
GaAs HBT
GaAs MESFET
InGaP HBT
SiGe BiCMOS
Si BiCMOS
SiGe HBT
GaAs pHEMT
Si CMOS
Si BJT
GaN HEMT
BiFET HBT
LDMOS
RF MICRO DEVICES®, RFMD®, Optimum Technology Matching®, Enabling Wireless Connectivity™, PowerStar®, POLARIS™ TOTAL RADIO™ and UltimateBlue™ are trademarks of RFMD, LLC. BLUETOOTH is a trademark owned by Bluetooth SIG, Inc., U.S.A. and licensed for use by RFMD. All other trade names, trademarks and registered trademarks are the property of their respective owners. ©2006, RF Micro Devices, Inc.
DS110520
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
1 of 9
RF3934D
Absolute Maximum Ratings
Parameter
Rating
Unit
Drain Voltage (VD)
150
V
Gate Voltage (VG)
-8 to +2
V
Gate Current (IG)
78
mA
Operational Voltage
50
V
-55 to +125
°C
200
°C
Storage Temperature Range
Operating Junction Temperature
(TJ)
Human Body Model (based on
packaged device)
Class 1A
MTTF (TJ< 200 °C, 95%
Confidence Limits)*
3E + 06
Hours
1.6
°C/W
Thermal Resistance, RTH (junction
to case)**measured at
TC =85 °C, DC bias only
Caution! ESD sensitive device.
Exceeding any one or a combination of the Absolute Maximum Rating conditions may
cause permanent damage to the device. Extended application of Absolute Maximum
Rating conditions to the device may reduce device reliability. Specified typical performance or functional operation of the device under Absolute Maximum Rating conditions is not implied.
The information in this publication is believed to be accurate and reliable. However, no
responsibility is assumed by RF Micro Devices, Inc. ("RFMD") for its use, nor for any
infringement of patents, or other rights of third parties, resulting from its use. No
license is granted by implication or otherwise under any patent or patent rights of
RFMD. RFMD reserves the right to change component circuitry, recommended application circuitry and specifications at any time without prior notice.
RoHS (Restriction of Hazardous Substances): Compliant per EU Directive
2002/95/EC.
Operation of this device beyond any one of these limits may cause permanent damage. For reliable continuous operation, the device voltage
and current must not exceed the maximum operating values.
* MTTF - median time to failure for wear-out failure mode (30% IDSS degradation) which is determined by the technology process reliability.
Refer to product qualification report for FIT (random) failure rate.
** Thermal resistance assumes AuSn die attach on1.5mm thick CPC carrier similar to Kyocera A1933. User will need to define this
specification in the final application and ensure bias conditions satisfy the following expression:PDISS T<(TJ -TC)/RTH J-C and TC =TCASE to
maintain maximum operating junction temperature and MTTF.
Parameter
Min.
Specification
Typ.
Max.
48
V
-3.7
-2.5
V
4000
MHz
Unit
Condition
Recommended Operating Conditions
Drain Voltage (VDSQ)
24
Gate Voltage (VGSQ)
-4.5
Drain Bias Current
440
Frequency of Operation
DC
mA
Die Capacitance from packaged
capacitance measurements
(package capacitance
removed during calibration)
Crss
9
pF
Ciss
40
pF
Coss
27.5
pF
VG =-8V, VD =0V
DC Functional Test
IG (on) – Forward Bias Diode Gate
Current
10
mA
VG =1.1V, VD =0V
VG =-8V, VD =0V
IG (off) – Gate Leakage
0.2
mA
ID (off) – DrainLeakage
0.2
mA
ID (off) – 48V DrainLeakage
6.0
mA
VG =-8V, VD =48V
ID (off) – 150V DrainLeakage
10.0
mA
VG =-8V, VD =150V
VGS (th) – Threshold Voltage
VDS (on) – Drain Voltage at high
current
2 of 9
-4.8
-3.4
0.25
-2.5
V
VD =48V, ID =20mA
V
VG =0V, ID =1.0A
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
DS110520
RF3934D
Parameter
Min.
Specification
Typ.
Max.
Unit
RF Typical Performance of
packaged die
Condition
[1]
VGS (q)
-3.4
V
Small Signal Gain
21
dB
VD =48V, ID =440mA
CW, F=900MHz
Small Signal Gain
13
dB
CW, F=2140MHz
Output Power at P3dB
51.6
dBm
CW, F=900MHz
Output Power at P3dB
51.46
dBm
CW, F=2140MHz
Drain Efficiency at P3dB
75
%
CW, F=900MHz
Drain Efficiency at P3dB
60
%
CW, F=2140MHz
[1] Test Conditions: CW Operation, VDSQ =48V, IDQ =440mA, T=25 °C, in standard tuned test circuit.
DS110520
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
3 of 9
RF3934D
Typical Performance of (non-internally matched) packaged die in tuned circuit
(T=25°C, unless noted)
Gain versus Output Power (F = 2140MHz)
Efficiency versus Output Power (F = 2140MHz)
(Pulsed 10% duty cycle, 10S, VD = 48V, IDQ = 440mA)
(Pulsed 10% duty cycle, 10S, VD = 48V, IDQ = 440mA)
16
70
60
Eff 85C
14
Eff 25C
Eff -40C
Drain Efficiency (%)
Gain (dB)
50
12
40
30
10
Gain 85C
Gain 25C
20
Gain -40C
8
10
36
39
42
45
48
51
36
39
42
Output Power (dBm)
45
48
Input Return Loss versus Output Power (F = 2140MHz)
Small Signal Performance versus Frequency, POUT = 30dBm
(Pulsed 10% duty cycle, 10S, VD = 48V, IDQ = 440mA)
(VD = 48V, IDQ = 440mA)
15
-6
IRL 85C
-8
IRL 25C
14
-10
51
Output Power (dBm)
IRL -40C
-6
Fixed tuned test circuit
-7
13
-8
-16
-18
-20
12
-9
11
-10
10
9
-22
-11
Gain
IRL
-12
8
-13
7
-14
-28
6
-15
-30
5
Input Return Loss (dB)
-14
Gain (dB)
IRL, Input Return Loss (dB)
-12
-24
-26
36
39
42
45
48
2110
51
-16
2120
2130
Output Power (dBm)
Gain/IRL versus Frequency, POUT = 50.8dBm
2150
2160
2170
Drain Efficiency versus Frequency, POUT = 50.8dBm
(CW, VD = 48V, IDQ = 440mA)
(CW, VD = 48V, IDQ = 440mA)
15
14
2140
Frequency (MHz)
-6
Fixed tuned test circuit
62
Fixed tuned test circuit
-7
13
-8
12
-9
11
-10
10
-11
9
-12
8
-13
61
Drain Efficiency (%)
Input Return Loss (dB)
Gain (dB)
60
59
58
57
56
7
Gain
-14
IRL
6
-15
5
2110
-16
2120
2130
2140
Frequency (MHz)
4 of 9
2150
2160
2170
Eff
55
54
2110
2120
2130
2140
2150
2160
2170
Frequency (MHz)
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
DS110520
RF3934D
Gain/ Efficiency versus POUT, F = 2140MHz
Gain/ Efficiency versus POUT, F = 2140MHz
(CW, VD = 48V, IDQ = 440mA)
(Pulsed 10% duty cycle, 10S, VD = 48V, IDQ = 440mA)
70
14
60
14
50
40
8
30
6
40
8
30
6
20
Gain
10
2
40
45
Gain
50
10
0
55
0
40
42
44
Pout, Output Power (dBm)
46
48
50
52
Pout, Output Power (dBm)
IMD3 versus POUT
Gain versus POUT
(2-Tone 1MHz Separation, VD = 48V, IDQ varied, FC = 2140MHz)
(2-Tone 1MHz Separation, VD = 48V, IDQ varied, FC = 2140MHz)
-10
16
-15
15
-20
14
-25
Gain (dB)
IMD3, Intermodulation Distortion (dBc)
Drain Eff
2
0
35
20
4
Drain Eff
4
30
50
10
Gain (dB)
10
60
12
Drain Efficiency (%)
Gain (dB)
12
70
16
Drain Efficiency (%)
16
-30
13
220mA
220mA
330mA
-35
12
330mA
440mA
440mA
550mA
550mA
11
-40
660mA
660mA
-45
10
1
10
100
1000
1
10
100
1000
POUT, Output Power (W-PEP)
POUT, Output Power (W-PEP)
IMD versus Output Power
(VD = 48V, IDQ = 440mA, F1 = 2139.5MHz, F2 = 2140.5MHz)
-10
-IMD3
IMD3
-IMD5
IMD5
-IMD7
IMD7
Intermodulation Distortion (IMD - dBc)
-15
-20
-25
-30
-35
-40
-45
-50
1
10
100
POUT, Output Power (W- PEP)
DS110520
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
5 of 9
RF3934D
Typical Performance in standard 900MHz fixed tuned test fixture (T=25°C)
Small Signal Performance versus Frequency,
POUT = 30dBm
Gain/IRL versus Frequency, POUT = 50.8dBm
(CW, VD = 48V, IDQ = 440mA)
(VD = 48V, IDQ = 440mA)
-3
Fixed tuned test circuit
-5
22
-6
21
-7
20
-8
19
-9
18
-10
Gain
17
IRL
-11
16
-12
15
-13
880
890
900
910
Gain (dB)
23
-5
24
-4
Input Return Loss (dB)
Gain (dB)
24
25
-6
Fixed tuned test circuit
23
-7
22
-8
21
-9
20
-10
19
-11
18
-12
17
-13
16
-14
15
-15
-16
14
13
Gain
-17
IRL
12
-18
11
-19
-20
10
920
Input Return Loss (dB)
25
880
885
890
Frequency (MHz)
895
900
905
910
915
920
Frequency (MHz)
Gain/ Efficiency versus POUT, F = 900MHz
Drain Efficiency versus Frequency, POUT = 50.8dBm
(CW, VD = 48V, IDQ = 440mA)
(CW, VD = 48V, IDQ = 440mA)
74
26
80
24
70
22
60
20
50
18
40
16
30
73
71
Gain (dB)
Drain Efficiency (%)
72
70
69
Gain
67
880
885
890
895
900
Frequency (MHz)
6 of 9
20
14
Eff
68
905
910
915
920
Drain Efficiency (%)
Fixed tuned test circuit
Drain Eff
12
10
10
0
30
35
40
45
50
55
Pout, Output Power (dBm)
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
DS110520
RF3934D
Die Drawing
GND
D
0.300
G
S
S
D
0.300
G
S
S
D
0.300
G
S
S
D
0.300
G
S
S
D
0.300
G
S
S
D
0.300
G
S
S
D
0.300
G
D
D
G
S
0.102
0.074
0.00
0.032
0.300
0.332
S
0.226
0.199
0.268
D
G
S
0.102
0.074
0.00
0.788
G
S
S
0.641
0.300
0.268
0.032
0.00
0.103
0.641
D
0.788
G
S
S
0.332
S
0.183
0.300
0.226
0.199
0.183
0.300
0.103
D
0.00
4.570±0.010
S
S
G
S
S
0.300
Detail of Unit Cell
S
S
0.785±0.010
0.085±0.010
0.952±0.010
Dimensions in mm
External dim. tolerance due to
dicing process development.
Bond Pad Key
G = Gate
S = Source
D = Drain
Die thickness = 0.101mm
Die backside metal = Au
Bias Instruction for RF3934D Die
ESD Sensitive Material. Please use proper ESD precautions when handling devices die. Die must be mounted with minimal die
attach voids for proper thermal dissipation. This device is a depletion mode HEMT and must have gate voltage applied for
pinched off prior to applying drain voltage.
1. Mount device on carrier or package with minimal die attach voiding and applying proper heat removal techniques.
2. Connect ground to the ground supply terminal, and ensure that both the VG and VD grounds are also connected to this
ground terminal.
3. Apply -8V to VG.
4. Apply 48V to VD.
5. Increase VG until drain current reaches desired bias point.
6. Apply RF input.
DS110520
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
7 of 9
RF3934D
Assembly Notes
Die Storage
• Individual bare die should be held in appropriately sized ESD waffle trays or ESD GEL packs.
• Die should be stored in CDA/N2 cabinets and in a controlled temperature and humidity environment.
Die Handling
• Die should only be picked using an auto or semi-automated pick system and an appropriate pick tool.
• Pick parameters will need to be carefully defined so not to cause damage to either the top or bottom die surface.
• GaN HEMT devices are ESD sensitive materials. Please use proper ESD precautions when handling devices or evaluation boards.
• RFMD does not recommend operating this device with typical drain voltage applied and the gate pinched off in a high
humidity, high temperature environment.
Caution: The use of inappropriate or worn-out ejector needle and improper ejection parameter settings can
cause die backside tool marks or micro-cracks that can eventually lead to die cracking.
Die Attach
There are two commonly applied die attach processes: adhesive die attach and eutectic die attach. Both processes use special equipment and tooling to mount the die.
EUTECTIC ATTACH
•
•
•
•
•
•
•
•
•
80/20 AuSn preform, 0.5 - 1mil thickness, made from virgin melt gold.
Pulsed heat or die scrub attach process using auto / semi-automatic equipment.
Attach process carried out in an inert atmosphere.
Custom die pick collets are required that match the outline of the die and the specific process employed using either
pulsed, fixed heat, or scrub.
Maximum temperature during die attach should be no greater than 320 C and for less than 30 seconds.
Key parameters that need to be considered include: die placement force, die scrub profile and heat profile.
Minimal amount of voiding is desired to ensure maximum heat transfer to the carrier and no voids should be present
under the active area of the die.
Voiding can be measured using X-ray or Acoustic microscopy.
The acceptable level of voiding should be determined using thermal modeling analysis.
ADHESIVE ATTACH
• High thermal silver filled epoxy is dispensed in a controlled manner and die is placed using an appropriate collet.
Assembled parts are cured at temperatures between 150C and 180C.
• Always refer to epoxy manufacturer's data sheet.
• Industry recognized standards for epoxy die attach are clearly defined within MIL-883.
Early Life Screen Conditions
RFMD recommends an Early Life Screen test that subjects this die to TJ =250C (junction temperature) for at least 6 hours
prior to field deployment.
Mounting and Thermal Considerations
The thermal resistance provided as RTH (junction to case) represents only the packaged device thermal characteristics. This is
measured using IR microscopy capturing the device under test temperature at the hottest spot of the die. At the same time, the
package temperature is measured using a thermocouple touching the backside of the die embedded in the device heatsink
but sized to prevent the measurement system from impacting the results. Knowing the dissipated power at the time of the
measurement, the thermal resistance is calculated.
8 of 9
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
DS110520
RF3934D
Mounting and Thermal Considerations (continued)
In order to achieve the advertised MTTF, proper heat removal must be considered to maintain the junction at or below the maximum of 200C. Proper thermal design includes consideration of ambient temperature and the thermal resistance from ambient to the back of the package including heatsinking systems and air flow mechanisms. Incorporating the dissipated DC power,
it is possible to calculate the junction temperature of the device.
DC Bias
The GaN HEMT device is a depletion mode high electron mobility transistor (HEMT). At zero volts VGS the drain of the device is
saturated and uncontrolled drain current will destroy the transistor. The gate voltage must be taken to a potential lower than
the source voltage to pinch off the device prior to applying the drain voltage, taking care not to exceed the gate voltage maximum limits. RFMD recommends applying VGS =-5V before applying any VDS.
RF Power transistor performance capabilities are determined by the applied quiescent drain current. This drain current can be
adjusted to trade off power, linearity, and efficiency characteristics of the device. The recommended quiescent drain current
(IDQ) shown in the RF typical performance table is chosen to best represent the operational characteristics for this device, considering manufacturing variations and expected performance. The user may choose alternate conditions for biasing this device
based on performance trade-offs.
GaN HEMT Capacitances
The physical structure of the GaN HEMT results in three terminal capacitors similar to other FET technologies. These capacitances exist across all three terminals of the device. The physical manufactured characteristics of the device determine the
value of the CDS (drain to source), CGS (gate to source) and CGD (gate to drain). These capacitances change value as the terminal voltages are varied. RFMD presents the three terminal capacitances measured with the gate pinched off (VGS =-8V) and
zero volts applied to the drain. During the measurement process, the parasitic capacitances of the package that holds the
amplifier is removed through a calibration step. Any internal matching is included in the terminal capacitance measurements.
The capacitance values presented in the typical characteristics table of the device represent the measured input (CISS), output
(COSS), and reverse (CRSS) capacitance at the stated bias voltages. The relationship to three terminal capacitances is as follows:
CISS =CGD +CGS
COSS = CGD + CDS
CRSS = CGD
DS110520
7628 Thorndike Road, Greensboro, NC 27409-9421 · For sales or technical
support, contact RFMD at (+1) 336-678-5570 or [email protected].
www.BDTIC.com/RFMD
9 of 9