MIC2297_08 Datasheet, PDF (9/11 Pages) Micrel Semiconductor – 40V PWM Boost Regulator White LED Driver

Micrel
MIC2297
Over-Voltage Protection
The MIC2297 has an over-voltage protection function. If an
LED is disconnected from the circuit or the feedback pin is
shorted to ground, the feedback pin will fall to ground
potential. This will cause the MIC2297 to switch at full duty
cycle in an attempt to maintain the feedback voltage. As a
result, the output voltage will climb out of control. This may
cause the switch node voltage to exceed its maximum
voltage rating, possibly damaging the IC and the external
components. To ensure the highest level of protection, the
MIC2297 OVP pin will shut the switch off when an over-
voltage condition is detected, saving itself and the output
capacitor.
Brightness Control
In the MIC2297, the reference to the voltage error amplifier
is pinned out. The BRT pin and REF pin form a voltage
divider off the internal 1.245V reference. The voltage is
such that with nothing connected to the BRT pin, the REF
voltage is 0.2V and the BRT voltage is 1V. The REF
voltage is 1/5 the BRT voltage.
The minimum REF voltage with BRT pulled to ground is
typically 10mV. With a 10Ω sense resistor, the LED
current is typically 1mA with the BRT pin pulled to ground.
An analog DC voltage can be connected to the BRT pin.
The MIC2297 will create an LED current proportional to
the BRT voltage according to the following equation:
I _LED
=
BRT
5 • R_sense
Where BRT is the voltage applied to the BRT pin, and
Rsense is the sense resistor used in the LED string. It’s
important to use a 1uF ceramic capacitor on the REF pin
to filter any noise.
An external PWM signal can be applied to the BRT for
dimming. The 1uF REF capacitor and internal BRT 124kΩ
resistor form an RC that filters the voltage to the REF pin.
The LED current is proportional the PWM duty cycle
according to the following equation:
I _LED
=
V_peak • D
5 • R_sense
Where Vpeak is the peak PWM voltage and D is the duty
cycle of the PWM signal.
Component Selection
Inductor
Inductor selection is a balance between efficiency,
stability, cost, size, and rated current. For most
applications a 22µH is the recommended inductor value. It
is usually a good balance between these considerations.
Larger inductance values reduce the peak-to-peak ripple
current, affecting efficiency. This has the effect of reducing
both the DC losses and the transition losses. There is also
a secondary effect of an inductor’s DC resistance (DCR).
The DCR of an inductor will be higher for more inductance
in the same package size. This is due to the longer
windings required for an increase in inductance. Since the
majority of input current (minus the MIC2297 operating
current) is passed through the inductor, higher DCR
inductors will reduce efficiency. To maintain stability,
increasing inductor size will have to be met with an
increase in output capacitance. This is due to the
unavoidable “right half plane zero” effect for the continuous
current boost converter topology. The frequency at which
the right half plane zero occurs can be calculated as
follows:
f _rhpz
=
V_out
V_in²
⋅ L ⋅ I _out
⋅ 2π
The right half plane zero has the undesirable effect of
increasing gain, while decreasing phase. This requires that
the loop gain is rolled off before this has significant effect
on the total loop response. This can be accomplished by
either reducing inductance (increasing RHPZ frequency) or
increasing the output capacitor value (decreasing loop
gain).
Output Capacitor
Output capacitor selection is also a trade-off between
performance, size, and cost. Increasing output
capacitance will lead to an improved transient response,
but also an increase in size and cost. X5R or X7R
dielectric ceramic capacitors are recommended for
designs with the MIC2297.
The output capacitor sets the frequency of the pole and
zero in the power stage. The zero is given by:
f_z
=
C
1
⋅ R_esr
⋅ 2π
For ceramic capacitors, the ESR is very small. This puts
the zero at a very high frequency where it can be ignored.
The frequency of the pole caused by the output capacitor
is given by.
f_p
=
I _out
C ⋅V_out
⋅π
Mach 2008
9
M9999-032708