AD8016
we can use symmetry to simplify the computation for a dc input
signal.
of the die, allowing more drivers/square-inch within the CO
design. The AD8016, whether in a PSOP3 (ARP) or batwing
(ARB) package, can be designed to operate in the CO solution
using prudent measures to manage the power dissipation through
careful PCB design. The PSOP3 package is available for use in
designing the highest density CO solutions. Maximum heat trans-
fer to the PCB can be accomplished using the PSOP3 package
when the thermal slug is soldered to an exposed copper pad
directly beneath the AD8016. Optimum thermal performance
can be achieved in the ARE package only when the back of the
package is soldered to a PCB designed for maximum thermal
capacity (see Figure 44). Thermal experiments with the PS0P3
package were conducted without soldering the heat slug to the
PCB. Heat transfer was through physical contact only. The
following offers some insight into the AD8016 power dissipation
and relative junction temperature, the effects of PCB size and
composition on the junction-to-air thermal resistance or θJA.
VO
PD = 2 × IQ × VS + 4 × (VS – VO )
RL
where
V
O is the peak output voltage of an amplifier.
This formula is slightly pessimistic due to the fact that some of
the quiescent supply current is commutated during sourcing or
sinking current into the load. For a sine wave source, integration
over a half cycle yields:
2
4 VO VS VO
PD = 2 × IQ × V + 2
−
S
π RL
R
L
The situation is more complicated with a complex modulated
signal. In the case of a DMT signal, taking the equivalent sine
wave power overestimates the power dissipation by ~23%. For
example:
THERMAL TESTING
A wind tunnel study was conducted to determine the relationship
between thermal capacity (i.e., printed circuit board copper area),
air flow and junction temperature. Junction-to-ambient ther-
mal resistance, θJA, was also calculated for the AD8016ARP,
AD8016ARE, and AD8016ARB packages. The AD8016 was
operated in a noninverting differential driver configuration, typical
of an xDSL application yet isolated from any other modem
components. Testing was conducted using a 1 ounce copper
board in an ambient temperature of ~24°C over air flows of
200, 150, 100, and 50 (0.200 and 400 for AD8016ARE) linear
feet per minute (LFM) and for ARP and ARB packages as well
as in still air. The four-layer PCB was designed to maximize the
area of copper on the outer two layers of the board while the
inner layers were used to configure the AD8016 in a differential
driver circuit. The PCB measured 3 × 4 inches in the beginning
of the study and was progressively reduced in size to approxi-
mately 2 × 2 inches. The testing was performed in a wind tunnel to
control air flow in units of LFM. The tunnel is approximately
11 inches in diameter.
P
OUT = 23.4 dBm = 220 mW
V
OUT @ 50 Ω = 3.31 V rms
VO = 2.354 V
at each amplifier output, which yields a PD of 1.81 W.
Through measurement, a DMT signal of 23.4 dBm requires
1.47 W of power to be dissipated by the AD8016. Figure 41
shows the results of calculation and actual measurements
detailing the relationship between the power dissipated by the
AD8016 versus the total output power delivered to the back
termination resistors and the load combined. A 1:2 transformer
turns ratio was used in the calculations and measurements.
2.5
2.0
CALCULATED
1.5
MEASURED
SINE
AIR FLOW TEST CONDITIONS
DUT Power: Typical DSL DMT signal produces about 1.5 W
of power dissipation in the AD8016 package. The fully biased
(PWDN0 and PWDN1 = Logic 1) quiescent current of the
AD8016 is ~25 mA. A 1 MHz differential sine wave at an ampli-
tude of 8 V p-p/amplifier into an RLOAD of 100 Ω differential
(50 Ω per side) will produce the 1.5 W of power typical in the
AD8016 device. (See the Power Dissipation section for details.)
MEASURED
DMT
1.0
0.5
0
0
100
200
300
OUTPUT POWER – mW
Thermal Resistance: The junction-to-case thermal resistance
(θJC) of the AD8016ARB or batwing package is 8.6°C/W,
AD8016ARE is 5.6°C/W, and the AD8016ARP or PSOP3
package is 0.86°C/W. These package specifications were used in
this study to determine junction temperature based on the mea-
sured case temperature.
Figure 41. Power Dissipation vs. Output Power (Including
Back Terminations). See Figure 7 for Test Circuit
THERMAL ENHANCEMENTS AND PCB LAYOUT
There are several ways to enhance the thermal capacity of the
CO solution. Additional thermal capacity can be created using
enhanced PCB layout techniques such as interlacing (sometimes
referred to as stitching or interconnection) of the layers immedi-
ately beneath the line driver. This technique serves to increase
the thermal mass or capacity of the PCB immediately beneath
the driver. (See AD8016-EVAL boards for an example of this
method of thermal enhancement.) A cooling fan that draws
moving air over the PCB and xDSL drivers, while not always
required, may be useful in reducing the operating temperature
PCB Dimensions of a Differential Driver Circuit: Several
components are required to support the AD8016 in a differential
driver circuit. The PCB area necessary for these components (i.e.,
feedback and gain resistors, ac coupling and decoupling capaci-
tors, termination and load resistors) dictated the area of the
smallest PCB in this study, 4.7 square inches. Further reduction
in PCB area, although possible, will have consequences in terms
of the maximum operating junction temperature.
–14–
REV. A
ADI [ ADI ]
