AAT3131 Datasheet, PDF (10/14 Pages) Advanced Analogic Technologies – High Efficiency 1X/1.5X Fractional Charge Pump for White LED Applications

EN/SET Timing Diagram
EN/SET
Code OFF
AAT3131
High Efficiency 1X/1.5X Fractional
Charge Pump for White LED Applications
t_HI
t_LO
t_OFF
1
2
3
OFF
LED Selection
The AAT3131 is specifically intended for driving
white LEDs. However, the device design will allow
the AAT3131 to drive most types of LEDs with for-
ward voltage specifications ranging from 2.0V to
4.3V. LED applications may include main and sub-
LCD display backlighting, camera photo-flash appli-
cations, color (RGB) LEDs, infrared (IR) diodes for
remotes, and other loads benefiting from a con-
trolled output current generated from a varying input
voltage. Since the D1 to D4 output current sources
are matched with negligible voltage dependence,
the LED brightness will be matched regardless of
the specific LED forward voltage (V_F) levels.
In some instances (e.g., in high luminous output
applications such as photo flash), it may be neces-
sary to drive high-V_F type LEDs. The low dropout
current sources in the AAT3131 makes it capable
of driving LEDs with forward voltages as high as
4.3V at full current from an input supply as low as
3.0V. Outputs can be paralleled to drive high-cur-
rent LEDs without complication.
Device Switching Noise Performance
The AAT3131 operates at a fixed frequency of
approximately 1MHz to control noise and limit har-
monics that can interfere with the RF operation of
cellular telephone handsets or other communica-
tion devices. Back-injected noise appearing on the
input pin of the charge pump is 20mV peak-to-
peak, typically ten times less than inductor-based
DC/DC boost converter white LED backlight solu-
tions. The AAT3131 soft-start feature prevents
noise transient effects associated with inrush cur-
rents during start-up of the charge pump circuit.
Power Efficiency and Device
Evaluation
The charge pump efficiency discussion in the fol-
lowing sections only accounts for efficiency of the
charge pump section itself. Due to the unique circuit
architecture and design of the AAT3131, it is very
difficult to measure efficiency in terms of a percent
value comparing input power over output power.
Since the AAT3131 outputs are pure constant cur-
rent sources and typically drive individual loads, it
is difficult to measure the output voltage for a given
output (D1 to D4) to derive an overall output power
measurement. For any given application, white
LED forward voltage levels can differ, yet the out-
put drive current will be maintained as a constant.
This makes quantifying output power a difficult task
when taken in the context of comparing to other white
LED driver circuit topologies. A better way to quantify
total device efficiency is to observe the total input
power to the device for a given LED current drive
level. The best white LED driver for a given applica-
tion should be based on trade-offs of size, external
component count, reliability, operating range, and
total energy usage...not just % efficiency.
The AAT3131 efficiency may be quantified under
very specific conditions and is dependent upon the
input voltage versus the output voltage seen
across the loads applied to outputs D1 through D4
for a given constant current setting. Depending
upon the case of V_IN being greater than the specif-
ic voltage seen across the load on D1 (or D4), the
device will operate in load switch mode. If V_IN is
less than the voltage required on the constant cur-
rent source, the device will operate in 1.5X charge
pump mode. Each of these two modes will yield
10
3131.2005.12.1.2