Microsemi

Integrated Products Division

11861 Western Avenue, Garden Grove, CA. 92841, 714-898-8121, Fax: 714-893-2570

Page 6

Copyright

© 2000

Rev. 1.0b, 2005-03-03

LX1993

High Efficiency LED Driver

I N T E GRA T ED

PR ODUC T S

APPLI CATI ON I N FORMATI O N

()

()

0.5

V

I

I

C

L

∆V

IN

OUT

PK

OUT

DROOP

−

×

×

=

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

The output overshoot can be estimated as follows where the

0.5 value in the denominator is an estimate of the voltage

drop across the diode:

()

()

IN

OUT

2

OUT

PK

OUT

2

1

OVERSHOOT

V

0.5

V

I

I

C

L

∆V

−

+

−

×

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

×

=

()

()

mV

0

.

2

0.5

13.0

mA

8

.

2

1

200mA

F

4.7

H

47

≅

−

×

×

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

µ

µ

=

()

()

mV

4

.

18

0

.

3

0.5

0

.

13

mA

8

.

12

mA

200

F

4.7

H

47

∆V

2

2

1

OVERSHOOT

≅

−

+

−

×

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

µ

µ

×

=

Therefore,

A Schottky diode is recommended for most applications

(e.g., Microsemi UPS5817). The low forward voltage drop

and fast recovery time associated with this device supports

the switching demands associated with this circuit

topology. The designer is encouraged to consider the

diode’s average and peak current ratings with respect to the

application’s

output

and

peak

inductor

current

requirements.

Further, the diode’s reverse breakdown

voltage characteristic must be capable of withstanding a

The LX1993 produces high slew-rate voltage and

current waveforms hence; the designer should take this into

consideration when laying out the circuit. Minimizing

trace lengths from the IC to the inductor, diode, input and

output capacitors, and feedback connection (i.e., pin 3) are

typical considerations. Moreover, the designer should

maximize the DC input and output trace widths to

accommodate peak current levels associated with this

topology.

The LXE1993 evaluation board is available from

Microsemi for assessing overall circuit performance.

The

evaluation board, shown in Figure 3, is 3 by 3 inches (i.e.,

7.6 by 7.6cm) square and programmed to drive 2 to 4 LEDs

(provided). Designers can easily modify circuit parameters

Moreover, the inductor, FET, and switching diode are easily

swapped out to promote design verification of a circuit that

maximizes efficiency and minimizes cost for a specific

application.

The evaluation board input and output

connections are described in Table 1.

The DC input voltage is applied to VBAT (not VCC)

however the LX1993 IC may be driven from a separate DC

source via the VCC input. The output current (i.e., LED

brightness) is controlled by adjusting the on-board

potentiometer.

The designer may elect to drive the

brightness adjustment circuit from VBAT or via a separate

voltage source by selecting the appropriate jumper position

(see Table 2). Optional external adjustment of the output

LED current is achieved by disengaging the potentiometer

and applying either a DC voltage or a PWM-type signal to

the VADJ input. The PWM signal frequency should be

higher than 150KHz and contain a DC component less than

350mV.

typ) during shutdown mode. The SHDN pin is used to

exercise the shutdown function on the evaluation board.

This pin is pulled-up to VCC via a 10KΩ resistor.

Grounding the SHDN pin shuts down the IC (not the circuit

output). The output voltage (i.e., voltage across the LED

string) is readily measured at the VOUT terminal and LED

current is derived from measuring the voltage at the VFDBK

pin and dividing this value by 15Ω (i.e., R4). The factory

installed component list for this must-have design tool is

provided in Table 3 and the schematic is shown in Figure 4.

Efficiency Measurement Hint: When doing an efficiency

evaluation using the LX1993 Evaluation Board, VPOT

should be driven by a separate voltage supply to account for

losses associated with the onboard reference (i.e., the 1.25V

shunt regulator and 1KΩ resistor). This circuit will have

VBAT - 1.25V across it and at the higher input voltages the

1KΩ resistor could have as much as 4mA through it. This

shunt regulator circuitry will adversely effect the overall

efficiency measurement.

It is not normally used in an

application; hence, it should not be considered when

measuring efficiency.