NSC LM3401MMX

LM3401
Hysteretic PFET Controller for High Power LED Drive
General Description
Features
The LM3401 is a switching controller designed to provide
constant current to high power LEDs. The LM3401 drives an
external P-MOSFET switch for step-down (Buck) regulators.
The LM3401 delivers constant current within ±6% accuracy
to a wide variety and number of series connected LEDs. Output current is adjusted with an external current sensing resistor to drive high power LEDs in excess of 1A.
For improved accuracy and efficiency, the LM3401 features
dual-side hysteresis, very low reference voltage, and short
propagation delay. A cycle by cycle current limit provides protection against over current and short circuit failures. Additional features include adjustable hysteresis and a CMOS
compatible input pin for PWM dimming.
■
■
■
■
■
■
■
■
■
■
Hysteretic Control for Speed and Simplicity
Input Operating Range of 4.5V-35V
1.5MHz maximum switching frequency
Low 200mV reference voltage
Programmable current limit
High speed CMOS compatible enable/dimming
Adjustable hysteresis
Input UVLO
No output capacitor required
MSOP-8 package
Applications
■ LED Driver
■ Battery Charger
Typical Application Circuit
30021401
© 2007 National Semiconductor Corporation
300214
www.national.com
LM3401 Hysteretic PFET Controller for High Power LED Drive
August 2007
LM3401
Connection Diagram
30021402
Top View
8-Lead Plastic MSOP
Ordering Information
Part Number
Package Type
NS Package Drawing
MSOP-8
MUA08A
LM3401MM
LM3401MMX
Supplied As
1000 Units on Tape and Reel
3500 Units on Tape and Reel
Pin Descriptions
Pin #
Pin Name
1
CS
Current sense pin. Connect to the PFET drain
2
DIM
Dimming input pin. When DIM is low, the HG drive is off. Can be connected to a logic level PWM signal
3
SNS
Current feedback pin – to LED cathode. Connect a resistor from this pin to ground to set the DC LED
current
4
HYS
Hysteresis adjust pin. Connect a resistor from this pin to GND to set the hysteresis window
5
GND
Ground pin
6
HG
Gate drive output. Connect to the PFET gate
7
VIN
Power supply input
8
ILIM
Current limit adjust pin. Connect a resistor from this pin to the PFET source to set the current limit threshold
www.national.com
Description
2
If Military/Aerospace specified devices are required,
please contact the National Semiconductor Sales Office/
Distributors for availability and specifications.
VIN
CS
SNS
ILIM
DIM
HYS
Storage Temperature
-0.3V to 36V
-2.0V to 36V
-0.3V to 8V
-0.3V to 36V
-0.3V to 36V
-0.3V to 4V
-65°C to +150°C
215°C
220C
2.5kV
Operating Ratings
(Note 1)
VIN
Junction Temperature Range
4.5V to 35V
-40°C to +125°C
Thermal Resistance θJA (Note 3)
Power Dissispation (Note 3)
151°C/W
0.66W
Electrical Characteristics
Specifications in standard type are for TJ = 25°C only, and limits in boldface type
apply over the junction temperature (TJ) range of -40°C to +125°C. Unless otherwise stated, VIN = 24V. Minimum and Maximum
limits are guaranteed through test, design, or statistical correlation. Typical values represent the most likely parametric norm at
TJ = 25°C, and are provided for reference purposes only (Note 4).
Symbol
Parameter
Conditions
Min
Typ
Max
Units
188
200
212
mV
SYSTEM
VREF
ΔVREF / ΔVIN
Reference Voltage
Line regulation
5V < VIN < 35V
IQ
Operating VIN Current (Note 5)
IHYS
Hysteresis Pin Source Current
HYS pin = 50 mV to 500 mV
SNS Comparator Hysteresis Multiplier
HYS pin = 250 mV
TDLY
SNS Comparator to HG Delay
SNS rising
TDIM
DIM to HG Delay
DIM rising
IILIM
ILIM Pin Sink Current
SNSHYS_MU
VCL_OFF
Current Limit Comparator Offset
VZC
Zero Cross Comparator Threshold
VDIM
DIM Threshold Voltage
Measured at CS pin
UVLO
mV/V
1.05
mA
15
20
25
µA
0.168
0.20
0.224
-
46
80
ns
69
120
ns
4
5.5
8
µA
-10
0
+10
mV
-70
-130
-200
mV
1.85
2.0
2.25
V
Hysteresis
ISNS
0.002
286
mV
SNS pin Bias Current
VSNS = 200 mV
300
780
nA
UVLO threshold
Vin rising
4.3
4.48
V
Hysteresis
0.5
V
Minimum on-time
150
ns
Source Current = 100 mA
5.3
Ω
Sink Current = 100 mA
10.5
Source, HG = VIN -2.5V
0.41
Ω
A
DRIVER
Ton_min
RHG
Gate Drive Resistance
IHG
Driver Output Current
VHG
HG on voltage
Sink, HG = VIN -2.5V
Referenced to VIN, steady state voltage
0.33
-4.2
-4.7
A
-5.5
V
Note 1: Absolute Maximum Ratings indicate limits beyond which damage may occur to the device. Operating Ratings indicate conditions for which the device is
intended to be functional, but do not guarantee specific performance limits. For guaranteed specifications, see Electrical Characteristics.
Note 2: The human body model is a 100 pF capacitor discharged through a 1.5 kΩ resistor into each pin.
Note 3: The maximum allowable power dissipation is a function of the maximum junction temperature, TJ_MAX, the junction-to-ambient thermal resistance, θJA,
and the ambient temperature, TA. The maximum allowable power dissipation at any ambient temperature is calculated using: PD_MAX = (TJ_MAX - TA)/θJA. The
maximum power dissipation of 0.66W is determined using TA = 25°C, θJA = 151°C/W, and TJ_MAX = 125°C. θJA will vary with board size and copper area. The
θJA spec is based on a JEDEC standard 4-layer board.
Note 4: All room temperature limits are 100% production tested. All limits at temperature extremes are guaranteed through correlation using standard Statistical
Quality Control (SQC) methods. All limits are used to calculate Average Outgoing Quality Level (AOQL).
Note 5: IQ specifies the current into the VIN pin and applies to non-switching operation.
3
www.national.com
LM3401
Lead Temperature
Vapor Phase (60sec)
Infrared (15sec)
ESD Rating (Note 2)
Human Body Model
Absolute Maximum Ratings (Note 1)
LM3401
Typical Performance Characteristics
Unless otherwise specified, VIN = 24V, TA = 25°C.
VREF vs Temperature
HYS Current vs VIN
30021403
30021404
HYS Multiplier vs Temperature
Startup Waveforms
30021456
30021406
ILIM Sink Current vs Temperature
HYS Current vs Temperature
30021408
www.national.com
30021409
4
LM3401
UVLO Threshold vs Temperature
SNS to HG Propagation Delay vs VIN
30021410
30021455
Line Transient Response
Initial HG Voltage vs Input Voltage
30021457
30021414
350 mA Efficiency
700 mA and 1A Efficiency
30021458
30021459
5
www.national.com
LM3401
Block Diagram
30021415
FIGURE 1. Block Diagram
Operational Description
HYSTERETIC CONTROL
The LM3401 is a step-down DC-DC controller designed to
provide a constant current source for driving a high power
LED string.
The LM3401 uses comparator-based voltage mode hysteretic
control for a simple and stable design. Hysteretic control does
not utilize an internal oscillator, but relies on output conditions
to directly control switching. The LM3401 controls LED current within the adjustable hysteresis window by monitoring
peak and valley voltage at the SNS pin. A dual sided hysteresis window is used to optimize accuracy.
Regulated LED current flows to ground through a sense resistor at the SNS pin. The voltage generated at the SNS pin
is compared to the 200 mV internal reference. When the SNS
voltage falls below the reference voltage minus hysteresis,
the output of the hysteretic comparator goes low. This results
in the driver output, HG, pulling the gate of the PFET low and
turning on the PFET.
With the PFET on, LED current ramps up through the PFET
and the inductor. As the LED current increases, the SNS voltage reaches its upper threshold (reference voltage plus hysteresis). This forces the output of the comparator and HG to
go high, which turns the PFET off. When the PFET turns off,
current flows through the catch diode, and LED and inductor
current ramp down. When the SNS voltage falls to its lower
threshold, the cycle repeats. The resulting LED current, SNS,
and switch node waveforms are shown in Figure 2.
www.national.com
30021416
FIGURE 2. Hysteretic Switching Waveforms
OUTPUT CURRENT SETTING
The LED average current is programmed using a resistor between SNS and GND, shown as R1 in the typical application
schematic. The SNS resistor (RSNS) can be calculated as follows:
Where VSNS is 200 mV typically, and ILED is the DC average
LED current.
The sense resistor power rating must be higher than its power
dissipation. The required power rating can be calculated (in
Watts) as follows:
6
more precise equation should be used to determine the actual
ripple current and LED peak current. Larger hysteresis values
will result in lower switching frequency and higher ripple current for a given inductor. Typical examples are shown in
Figure 3 below.
The LED ripple current, ILED_RIP, is described below in the
Hysteresis Adjust section.
HYSTERESIS ADJUST
Adjustable hysteresis (via the HYS pin) provides direct control
over the LED ripple current. The HYS pin also gives some
control over the switching frequency. Although the hysteresis
value can be set after the inductor is selected, a preliminary
value must be set in order to begin the frequency calculation.
The hysteresis window must be set small enough to keep the
peak LED current below its maximum rating, ILED_max.
The maximum SNS pin hysteresis can be calculated as
shown:
30021422
FIGURE 3. Switching Frequency and Ripple Current vs
Hysteresis
SNSHYS_MAX = (ILED_max - ILED) x RSNS
SWITCHING FREQUENCY
Although hysteretic control is a simple control method, the
switching frequency depends on application conditions and
components. If the inductance, input voltage, or LED forward
voltage is changed, there will be a change in the switching
frequency. Therefore, care must be taken to select components which will provide the desired frequency range. Usually,
the best approach is to determine a nominal switching frequency for the application and then select the inductor accordingly. Once the inductor is selected, VHYS can be adjusted
to set the frequency range more precisely. This design process usually involves a few iterations to select appropriate
standard values that will result in the desired frequency and
ripple current.
Switching frequency can be approximately calculated using
the formula:
Any SNSHYS value between 10mV and this maximum is acceptable.
The SNSHYS value is set with a single resistor from the HYS
pin to GND, shown as R2 in the typical application schematic.
The HYS pin voltage, VHYS, is internally multiplied by
SNSHYS_MU (0.2 typical) to generate the hysteresis at the SNS
pin, SNSHYS. The hysteresis setting resistor can be determined from the following equation:
Where 20 µA is the typical HYS source current, and
SNSHYS is the resulting SNS pin hysteresis voltage. The hysteresis voltage can be set within a range of 10 mV to 100 mV
(50 mV to 500 mV at the HYS pin). The SNSHYS value defines
both the upper and lower threshold of the SNS pin. For example, with a VHYS setting of 100 mV, SNSHYS will be 20 mV.
Therefore, the total hysteretic window will be 40 mV, or 20 mV
above and 20mV below the 200 mV reference voltage. This
directly correlates to peak-to-peak inductor and LED ripple
current, approximated by the following equation:
Where D is the duty cycle, defined as (VOUT + VDIODE) / VIN,
VANODE is 200 mV plus the sum of LED forward voltages, and
delay is the sum of the LM3401 propagation delay time and
the PFET delay time. The LM3401 propagation delay is 46 ns
typically (See the Propagation Delay curve). Alternately, the
inductor value can be calculated from a known frequency by
re-arranging the same equation:
If LED ripple current is a design priority, the preliminary R2
value can be determined using a target LED ripple current as
shown:
A more precise equation for ripple current is given in the Inductor Selection section. Once an inductor is selected the
Switching frequency for a typical application is shown in Figure 4 along with the calculated frequency.
7
www.national.com
LM3401
WRSNS = VSNS x ILED
When selecting a sense resistor, thermal de-rating must also
be taken into consideration.
While RSNS sets the DC LED current, the AC peak LED current will be higher than the DC setting. This peak current must
not be greater than the maximum peak current rating of the
LED, ILED_max. Peak LED current can be calculated from the
equation below:
LM3401
RIPPLE REDUCTION CAPACITOR
The typical application schematic shown on the front page is
the simplest application of the LM3401. In this schematic, inductor current is equal to LED current. Therefore, the equations for ripple current apply to both LED ripple and inductor
ripple.
However, LED ripple current can be reduced without altering
the inductor current by using a bypass capacitor placed in
parallel with the LED string (shown as C2 in the Figure 11
schematic).
This allows larger hysteresis values to be used while significantly reducing ripple current in the LED string. Figure 5 below
shows this effect: inductor ripple current is unaffected while
LED ripple current is dramatically reduced.
30021425
FIGURE 4. Frequency vs Input Voltage
Maximum switching frequency will typically occur around the
input voltage which corresponds to 25% duty cycle.
If the input voltage falls below the forward voltage of the LED
string, the LM3401 will operate at 100% duty cycle. In this
state, the anode voltage will be equal to the input voltage and
LED current is determined by the v-i curve of the LED. At
100% duty cycle, LED current will be continuous with a maximum value equal to the ILED_PK level set at the HYS pin.
In some applications, maximum operating frequency will be
limited by the minimum on-time as shown in the following
equation:
30021428
FIGURE 5. LED Ripple Current Reduction with a 1 µF
Ripple Reduction Capacitor
If a ripple reduction capacitor is used, the equation for
SNSHYS_MAX only applies for 100% duty operation.
LED average DC current and peak inductor current are not
affected by the ripple reduction capacitor. However, LED
peak current is reduced and switching frequency may shift
slightly.
Any type of capacitor can be used, provided the working voltage rating is sufficient. In general, higher capacitance and
lower ESR will provide more ripple reduction; a typical value
greater than 100 nF is recommended. Smaller capacitance
values will be less effective, and large ESR values may actually increase LED ripple current.
Despite its effectiveness to smooth LED ripple current, there
are two notable disadvantages to using a ripple reduction capacitor.
First, when used, care must be taken to avoid shorting the
LED anode to ground. If this occurs, the capacitor will force a
large negative voltage spike at the SNS pin which could damage the IC.
Second, this capacitor will limit the maximum PWM dimming
frequency because it takes some additional time to charge
and discharge. Additionally, ceramic capacitors can generate
audible noise due to fast voltage changes during dimming. To
reduce audible noise, use the smallest possible case size,
use dimming frequencies below 500 Hz, or use a non-ceramic
ripple reduction capacitor.
A small bypass capacitor, in the range of 100 pF to 200 pF
can also be used to reduce high frequency switching noise.
This is recommended in higher current applications, where
When the on-time reaches minimum (150 ns typical) due to
increasing input voltage, the frequency will be reduced in order to maintain the proper duty cycle.
INDUCTOR SELECTION
The most important parameters for the inductor are the inductance and the current rating. The LM3401 operates over
a wide frequency range and can use a wide range of inductance values.
Once an inductance value is determined from the frequency
equation, the maximum operating current must be verified.
Although peak-to-peak ripple current is controlled by the hysteresis value, there is some variation due to propagation
delay. This means that the inductance has a direct effect on
LED current line regulation.
In general, a larger inductor will result in lower frequency and
better line regulation. The actual peak-to-peak inductor current can be calculated as follows:
Use the maximum value of Vin to determine the worst case
ILED_RIP value. This value should be used to determine the
peak current, ILED_PK, shown in the Output Current Setting
section.
Since the LM3401 can operate at 100% duty cycle, the inductor must be rated to handle ILED_PK continuously.
www.national.com
8
value at 25°C. The Gate capacitance of the PFET has a direct
impact on both PFET transition time and the power dissipation
in the LM3401. Most of the power dissipated in the LM3401
is used to drive the PFET switch. This power can be calculated as follows:
The average amount of gate driver current required during
switching (IG) is:
HG AND PFET SELECTION
When switching, the HG pin swings from VIN (off state) to 4.7V
below VIN (typical). As long as the DIM pin is high and the SNS
pin is below the upper threshold, HG will stay low, driving the
PFET on.
The PFET should be selected based on the maximum DrainSource voltage (VDS), Drain current rating (Id), maximum
Gate-Source voltage (VGS), on-resistance (RDS(on)), and Gate
capacitance.
The voltage across the PFET in the off state is equal to the
sum of the input voltage and the diode forward voltage. The
VDS must therefore be selected to provide some margin beyond this voltage.
Since the peak current through the PFET is equal to the peak
current through the inductor, Id must be rated higher than the
maximum ILED_PK. The LM3401 is capable of 100% duty cycle, therefore, the PFET drain current should be rated to
handle ILED_PK continuously. In this case there is no ripple, so
IPK = IAVE.
Although the typical HG voltage is VIN - 4.7V, this voltage can
go much lower during the initial PFET turn-on time. How far
HG swings at turn-on depends on several factors including
the gate capacitance, on-time, and input voltage. As shown
in the Typical Performance Characteristics, the initial HG voltage swing increases with decreasing PFET gate capacitance.
Therefore, A PFET must be selected with a maximum VGS
rating larger than the initial HG voltage. Conversely, when
driving PFETs with larger gate capacitance, the initial HG
voltage will be lower. In some cases, a low VGS threshold
PFET may be required to ensure complete turn-on. Use the
Typical Performance curve as a guideline to selecting a proper PFET.
Note that HG will eventually settle around the typical voltage
of VIN - 4.7V regardless of the PFET gate capacitance.
HG has an absolute minimum voltage of 1.2V typically. When
the input voltage is below approximately 6V, this minimum
limit causes a reduction in drive voltage. At 5V input, for example, HG will swing to 1.2V (or a gate drive voltage of -3.8V).
This may not be sufficient to drive some PFETs, and at this
reduced HG voltage, RDS(on) is likely to increase and trigger
current limit. Therefore, a low VGS threshold PFET is also
recommended for lower input voltage applications.
The power loss in the PFET consists of switching losses and
conducting losses. Although switching losses are difficult to
precisely calculate, the equations below can be used to estimate total power dissipation, which is the sum of PDCOND and
PDSW.
IG = Qg x fSW
Where Qg is the PFET gate charge.
And the total power dissipated in the IC is:
PD = (Iq x VIN) + (IG x 4.7V)
Where Iq is typically 1.05 mA and 4.7V is the typical HG voltage.
Maximum power dissipation within the LM3401 is limited by
ambient temperature. Use the following equation to determine maximum allowable power dissipation, or maximum
allowable ambient temperature:
Where θJA is the typical thermal resistance of 151°C/W. In
general, keeping the gate capacitance below 2000 pF is recommended to keep propagation delay, switching losses, and
power losses low. PFETs with very fast rise times may cause
excessive ringing at the HG node when combined with the
inductance of a long HG trace. To reduce this ringing, a small
resistor can be added between HG and the PFET gate. A
typical value of 10Ω is usually sufficient.
CURRENT LIMIT OPERATION
The LM3401 current limit monitors inductor current at each
switching cycle. Current is sensed across the RDS(on) of the
PFET at the CS pin. When the PFET current exceeds the
current limit threshold, HG is turned off and the current limit
latch is set. In current limit mode, the PFET is held off until the
inductor current falls to near zero.
PDFET_COND = RDS(on) x ILED2 x D
30021431
FIGURE 6. Typical Current Limit Operation
Where Pon = PFET turn-on time, Poff = PFET turn-off time, and
D is the duty cycle. A value of 10 ns to 50 ns is typical for ton
and toff. Longer PFET on and off times will degrade both efficiency and accuracy.
Increasing RDS(on) will increase power losses and degrade efficiency. FET RDS(on) has a positive temperature coefficient. At
125°C, the RDS(on) may be as much as 150% higher than the
The current limit threshold is adjusted with a setting resistor,
shown as R3 in the typical application schematic, connected
from ILIM to the VIN node of the PFET.
An internal 5.5 µA (typical) current sink at the ILIM pin creates
a voltage across the setting resistor. This voltage is compared
9
www.national.com
LM3401
switching noise can adversely affect the SNS or DIM pins. A
small capacitor for noise reduction will have little to no effect
on the LED ripple current or dimming but may help solve potential EMI problems.
LM3401
to the sensed voltage at the CS pin. Current limit is activated
and latched when the voltage at the CS pin drops below the
voltage at the ILIM pin.
The current limit setting resistor, R3, can be calculated from
the equation below. The minimum current limit occurs at maximum RDS(on) and minimum IILIM value.
limited by the inductor size and input voltage to anode voltage
ratio. Less inductance and higher VIN/VANODE ratios will allow
the inductor and LED current to increase faster, thus allowing
for a faster PWM frequency, or lower dimming duty cycle.
Where 4 µA is the minimum IILIM value and ILIM_PK is the peak
inductor current limit threshold. ILIM_PK should be set somewhat higher than the maximum LED current, ILED_PK, to avoid
false current limit triggering. The temperature variation of the
PFET RDS(on) will result in an equivalent variation in current
limit. To ensure that current limit is not falsely triggered, use
the highest RDS(on) value over the temperature range to set
the R3 value.
When current limit is activated, the HG driver remains off until
the CS voltage rises to -130 mV (typical). This ensures that
inductor current is close to 0A when the current limit latch is
released. The actual minimum inductor current will depend on
the catch diode forward voltage characteristic, which determines the CS pin negative voltage.
Although the LM3401 monitors voltage at the CS pin to reset
the current limit, there is also a minimum off time of typically
3 µs. When current limit is triggered, HG will be turned off for
at least this amount of time, regardless of the inductor current.
The current limit comparator imposes typically 150 ns of
blanking time at the beginning of each switching cycle. This
ensures that the PFET is fully on and any switch node ringing
has dissipated when the current is sensed. However a slower
PFET may not fully turn on within the blanking time. In this
case, the current limit threshold must be increased or a faster
PFET must be used.
Because the current limit comparator has a limited differential
voltage capability, a maximum of 1MΩ is recommended for
R3.
30021433
FIGURE 7. Typical PWM DIM Signal and LED Current
L = 22 µH
DIM is a high impedance pin, which is somewhat sensitive to
noise. If there is excessive switching noise at the DIM pin, a
small bypass filter capacitor can be used. See the Ripple Reduction Capacitor section. VIN can also be used for PWM
dimming when a logic signal is not available. In this mode of
operation DIM should be connected to VIN through a 10 kΩ
resistor. There is typically 10 us of startup delay time when
using VIN for dimming. Depending on the application, this delay limits the maximum dimming frequency to typically several
hundred Hz.
Higher dimming frequency and lower dimming duty cycle can
be achieved by using a FET switch in parallel with the LED
string. This is shown in Figure 8 below.
PWM DIMMING
The DIM pin is a CMOS compatible input for a PWM (Pulse
Width Modulation) dimming signal. PWM dimming adjusts
LED brightness by varying the duty cycle, which varies the
average LED current. This type of dimming is recommended,
because LED peak current remains constant regardless of
brightness, which results in more predictable LED color and
performance as compared to analog dimming. Figure 7
shows a typical PWM dimming waveform.
When DIM is high (above 2V typically) the LM3401 operates
normally and the LED string will be driven at the set current.
When pulled low, DIM will disable HG and switching will stop.
The PFET will remain off as long as DIM is low. When the
LM3401 is powered up or enabled with the DIM pin, the LED
current will very rapidly increase to its set point.
There is minimal delay time between a DIM logic change and
HG switching. Also, because the LM3401 requires no output
capacitor, minimal time is required to ramp-up the LED current. This allows for low duty cycle, high frequency PWM
dimming signals to be used.
A dimming frequency greater than 100 Hz is recommended
to avoid visible flicker. The LM3401 is capable of PWM dimming frequencies up to 10 kHz with a duty cycle between 1
and 100%. Any DIM signal pulse width longer than 100 ns can
be used. In most cases, the maximum dimming frequency is
www.national.com
30021434
FIGURE 8. Parallel FET Dimming
When the FET switches on, inductor current flows through the
FET and the regulated average inductor current is unchanged. Using this method, inductor current rise time does
not limit the dimming frequency. A ripple reduction capacitor
should not be used with the parallel FET dimming method
since it significantly slows the LED current rise time. However,
a small noise filter capacitor can be used.
INPUT CAPACITOR SELECTION
An input bypass capacitor is required between VIN and
ground. The input capacitor prevents large voltage transients
10
LM3401
at the input and provides the instantaneous current when the
PFET turns on. The important parameters for the input capacitor are the voltage rating and the RMS current rating.
Follow the manufacturer’s recommended voltage de-rating.
RMS current can be calculated with the equation below. The
highest RMS current will occur around 50% duty cycle.
A ceramic input capacitor must be placed close to the drain
of the PFET. This minimizes the trace inductance between
VIN and the PFET, which is a source of switching noise. If the
input capacitor is not properly located, switching noise can
cause current limit and stability problems.
CATCH DIODE SELECTION
The catch diode provides the current path to the LED string
during the PFET off-time and must be rated higher than the
average current through the diode, which can be calculated
as shown:
30021437
FIGURE 9. LED DC current line regulation LED Vf = 7.0V
IDIODE = ILED x (1-D)
For most applications, the average LED current will be the
highest at the maximum input voltage and lowest at a duty
cycle somewhat greater than 50%. The maximum LED current variation can be estimated as:
The peak reverse voltage across the catch diode is approximately equal to the input voltage. Therefore, the diode’s peak
reverse voltage rating should be larger than the maximum input voltage, plus some safety margin.
A Schottky diode is recommended because its low forward
voltage maximizes efficiency. For high temperature applications, diode leakage current may become significant and
require a higher reverse voltage rating or a low leakage diode
to achieve acceptable performance.
Where VIN_60% is the input voltage corresponding to a duty
cycle of 60%. Since the actual input voltage where minimum
LED current occurs varies with the application, this is an approximation. As the duty cycle approaches 100%, the average LED current will approach ILED_PK. The average LED
current will be the highest at the point that 100% duty cycle is
reached. In the case that 100% duty cycle can occur, maximum LED current variation is calculated as:
LED CURRENT ACCURACY
The total accuracy of average LED current is affected by several factors, both internal and external to the LM3401. Total
static accuracy is the part-to-part variation and can be calculated from the equation below:
Where the worst case VSNS% is ±6%, and RSNS% is the sense
resistor accuracy. Because these factors are not correlated,
the RSS (root-sum-square) method of calculation is used.
The LED current will also show some variation with input voltage. This is primarily due to propagation delay and the dynamic resistance of the LED. In longer on-time operation, the
error due to dynamic resistance tends to dominate, while at
shorter on-time, the propagation delay will dominate. These
two effects counteract each other, resulting in typical regulation curves similar to those shown in Figure 9. A larger inductor will reduce the error due to propagation delay and will
result in better overall line regulation.
PCB LAYOUT
PCB layout is very important in all switching regulator designs. Poor layout can cause EMI problems, excess switching
noise, and improper device operation. The following key
points should be followed to ensure a quality layout.
Traces carrying large AC currents should be as wide and
short as possible to minimize trace inductance. These areas,
shown as darker regions in Figure 10, are:
- VIN between the input capacitor and PFET
- GND between the input capacitor and catch diode
- The switch node
As shown in Figure 10, place the input capacitor ground as
close as possible to the anode of the catch diode. The VIN
side of the input capacitor should be placed close to the top
of the PFET.
The CS node (the node connecting the catch diode cathode,
inductor, and PFET source) should be kept as small as possible. This node is one of the main sources for radiated EMI.
The SNS and HYS pins are sensitive to noise. Be sure to route
the SNS trace away from the inductor and the switch node,
which are sources of noise.
11
www.national.com
LM3401
The SNS and HYS resistors should be placed close to their
respective pins and grounded close to the GND pin. An isolated ground area shown as SGND in Figure 10 is recommended for the SNS, HYS, and GND pin connections. The
two ground areas, GND and SGND, should be connected on
an inner or bottom layer. This connection is shown as two vias
in Figure 10.
A large, continuous ground plane can also be used, as long
as the input capacitor and catch diode ground area is somewhat isolated.
The HG trace should be kept as short as possible to minimize
inductance and gate ringing (See HG and PFET selection
section).
Finally, for accurate current limit sensing, the CS pin and ILIM
resistor connections should be made at the PFET pads, via
separate traces.
www.national.com
30021440
FIGURE 10. Example PCB Layout
12
LM3401
Design Example
30021441
FIGURE 11. Example Circuit
The following design example is intended to illustrate the
step-by-step design process described in the previous sections. The example refers to the circuit in Figure 11, and the
results are summaized in Table 1. The resulting circuit will
drive a string of 2 Luxeon V Star LEDs at 700 mA from an
input voltage between 18V and 35V.
The example LEDs have a maximum DC current rating of 700
mA, a forward voltage of 5.4V to 8.3V, and a maximum peak
current rating of 1.0A.
First, set the LED DC current with R1:
For 1 MHz switching frequency and 25 mV hysteresis, inductance can be calculated. Because frequency varies with input
voltage and LED forward voltage, for this calculation, assume
typical values of 24V and 13.6V respectively, and a PFET
delay time of 15 ns.
Select a value of 33 µH and the hysteresis can be adjusted
downward by re-arranging the same frequency equation:
And the required wattage is:
WRSNS = 700 mA2 x 0.286 = 140 mW
Select a standard value of 290 mΩ, 1/4W resistor, which will
result in a 690 mA LED DC current.
To keep the peak LED current below ILED_MAX, the maximum
hysteresis is determined by:
SNSHYS_MAX = (1.0A - .690A) x 0.29Ω = 90 mV
This gives a new R2 value of 5.6k. This will result in a typical
operating frequency of 1 MHz at 24VIN. Next, it must be verified that the peak LED current is within the maximum allowed.
For now, the design is created without using a ripple reduction
capacitor. Therefore, LED ripple current is equal to inductor
ripple current. Maximum LED ripple current is calculated as:
Which gives a maximum R2 value of:
Next, a preferred switching frequency of 1 MHz is selected for
this example.
Since this is a relatively high switching frequency, a low starting point of 25 mV is selected for the comparator hysteresis
to maintain good line regulation. This will allow a larger inductor at the same operating frequency and is well below the
calculated maximum. Set a preliminary hysteresis value with
R2:
Note that the maximum input voltage and minimum anode
voltage were used for this worst case calculation.
Now peak LED current can be determined:
13
www.national.com
LM3401
With the selected components, the maximum ambient temperature is above 100°C, sufficient for most applications. Note
that this limit applies to the IC only and depends on the pcb
type and size. Lower ambient temperature limits may apply to
the PFET and other components.
Now the current limit threshold is set with R3 at 0.95A, which
is 120% of the maximum peak current. The worst case RDS
(on) value at 125°C is used, which is 150% of nominal, and the
worst case ILIM pin sink current.
This confirms that the component selections will keep LED
peak current below the maximum LED rating. Notice if a ripple
reduction capacitor is chosen, the peak inductor current is still
804mA, but the LED peak current is reduced. Therefore, the
inductor must be rated for a DC current greater than 804 mA.
Now that the inductor value has been selected and verified,
the operating frequency range can be determined. Lowest
operating frequency occurs at minimum input voltage and
maximum anode voltage. For this example the values are 18V
minimum input and 16.8V maximum anode voltage (200 mV
SNS voltage plus the maximum LED forward voltages) and
calculate:
The typical current limit threshold will be higher than 1A and
can be determined by using typical values for RDS(on) and
ILIM sink current. The PFET, inductor, and catch diode must
be able to handle this current for short periods of time.
The next component is the input capacitor, C1. A low ESR
ceramic capacitor must be used and properly located on the
PCB. For this design, the capacitor working voltage must be
rated to at least 40V, and 50V is recommended. A 2.2 µF input
capacitor should be sufficient, assuming a good PCB layout.
The worst case input RMS current is calculated below 50% at
duty cycle:
At duty cycles close to 100% (96% in this case) the frequency
equation becomes less accurate. Actual switching frequency
will typically be lower than the calculated value.
To estimate maximum operating frequency, calculate using a
Vin which corresponds to a duty cycle of 25%. In this example,
25% duty cycle would occur above 35VIN, therefore maximum
frequency will occur at the maximum input voltage:
It must be verified that the selected input capacitor can tolerate this current. An additional bulk capacitor placed at the
input voltage connection is also recommended.
Next, select D1, the catch diode. A Schottky diode should be
used. The reverse voltage rating must be greater than 35V
and the average forward current rating must be greater than:
Using the equation in the Switching Frequency section, it can
be verified that this maximum frequency is within the minimum
on-time limited frequency (and below the maximum operating
frequency).
The maximum frequency calculation is only an estimate, the
actual maximum should be verified on the bench.
The next step is to select a PFET. The critical PFET parameters must meet the minimum circuit requirements of 35V
input, 804 mA DC current, and adequate gate drive voltage
rating.
Therefore, select a PFET with the following ratings:
40V maximum VDS
-20V maximum VGS
1.8A continuous Id
130mohm maximum RDS(on)
Typically the PFET may be only sourcing 690 mA for about
50% duty cycle. However, at minimum input voltage the duty
cycle will increase close to 100%. Therefore, the PFET Id rating should be based on its continuous, not pulsed, current
capability.
Now the power dissipation should be verified. Assume the
selected PFET has a gate capacitance of 200 pF, which is
within recommendation, and a gate charge of 15 nC. Maximum frequency and input voltage are used for a worst case
calculation:
IDIODE = 690 mA x (1 - 0.31) = 480 mA
This calculation assumes the minimum duty cycle, which is
maximum input voltage and minimum anode voltage. The
diode must also be able to handle peak currents as high as
the current limit threshold for short periods. We select a 1A
diode to ensure adequate capability over temperature.
If desired, a ripple reduction capacitor can be added at C2 to
reduce the LED ripple current. A minimum starting value of
100 nF is recommended for C2, and a value of 1 µF will work
well in most applications. In case of an open LED failure, the
ripple reduction capacitor must be rated to the maximum input
voltage of 35V. If C2 is used, LED ripple current is reduced
and the calculated maximum R2 value no longer applies as a
limit.
Finally, check the accuracy. The static accuracy is calculated
below using a 1% sense resistor.
To estimate line regulation, maximum input voltage and 60%
duty cycle input voltage is used. For this example 60% duty
cycle occurs at (13.6V + 0.2V)/0.60 or 23V input.
IG = 15 nC x 1.1 MHz = 16.5 mA
PD = (1.05 mA x 35V) + (16.5 mA x 4.7V) = 0.114W
Ta_max = 125°C - (151°C/W x 0.114W) = 108°C
www.national.com
14
17V, the LED current variation would be calculated using the
100% duty cycle equation.
TABLE 1. Design Example Summary
Parameter
Value
Result
Comment
R1
290 mΩ
690 mA DC
1%
R2
5.6 kΩ
±22.4 mV hysteresis
VHYS = 112 mV(adjustable)
L1
33 µH, >804 mA
fsw
-
1MHz typical
219 kHz min
Iripple
-
227 mA p-p
worst case
ILED_PK
-
804 mA
worst case
PFET
40V, 1.8A, 130 mΩ
R3
46 kΩ
0.95A minimum peak current limit
adjustable
C1
2.2 µF, 50V ceramic
>345 mA rms
D1
40V, 1A
Accuracy
±6.1%
±42 mA max variation
part-to-part
Regulation
1.6%
11 mA variation
vs VIN
Ta_max
108°C
C2
1.0 µF, 50V
Schottky
worst case
LED ripple reduction
15
optional
www.national.com
LM3401
This is the estimated amount of LED current variation over the
input voltage range. If the minimum input voltage was below
LM3401
Physical Dimensions inches (millimeters) unless otherwise noted
8-Lead MSOP Package
NS Package Number MUA08A
www.national.com
16
LM3401
Notes
17
www.national.com
LM3401 Hysteretic PFET Controller for High Power LED Drive
Notes
THE CONTENTS OF THIS DOCUMENT ARE PROVIDED IN CONNECTION WITH NATIONAL SEMICONDUCTOR CORPORATION
(“NATIONAL”) PRODUCTS. NATIONAL MAKES NO REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY
OR COMPLETENESS OF THE CONTENTS OF THIS PUBLICATION AND RESERVES THE RIGHT TO MAKE CHANGES TO
SPECIFICATIONS AND PRODUCT DESCRIPTIONS AT ANY TIME WITHOUT NOTICE. NO LICENSE, WHETHER EXPRESS,
IMPLIED, ARISING BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS
DOCUMENT.
TESTING AND OTHER QUALITY CONTROLS ARE USED TO THE EXTENT NATIONAL DEEMS NECESSARY TO SUPPORT
NATIONAL’S PRODUCT WARRANTY. EXCEPT WHERE MANDATED BY GOVERNMENT REQUIREMENTS, TESTING OF ALL
PARAMETERS OF EACH PRODUCT IS NOT NECESSARILY PERFORMED. NATIONAL ASSUMES NO LIABILITY FOR
APPLICATIONS ASSISTANCE OR BUYER PRODUCT DESIGN. BUYERS ARE RESPONSIBLE FOR THEIR PRODUCTS AND
APPLICATIONS USING NATIONAL COMPONENTS. PRIOR TO USING OR DISTRIBUTING ANY PRODUCTS THAT INCLUDE
NATIONAL COMPONENTS, BUYERS SHOULD PROVIDE ADEQUATE DESIGN, TESTING AND OPERATING SAFEGUARDS.
EXCEPT AS PROVIDED IN NATIONAL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, NATIONAL ASSUMES NO
LIABILITY WHATSOEVER, AND NATIONAL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY RELATING TO THE SALE
AND/OR USE OF NATIONAL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR
PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY
RIGHT.
LIFE SUPPORT POLICY
NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR
SYSTEMS WITHOUT THE EXPRESS PRIOR WRITTEN APPROVAL OF THE CHIEF EXECUTIVE OFFICER AND GENERAL
COUNSEL OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein:
Life support devices or systems are devices which (a) are intended for surgical implant into the body, or (b) support or sustain life and
whose failure to perform when properly used in accordance with instructions for use provided in the labeling can be reasonably expected
to result in a significant injury to the user. A critical component is any component in a life support device or system whose failure to perform
can be reasonably expected to cause the failure of the life support device or system or to affect its safety or effectiveness.
National Semiconductor and the National Semiconductor logo are registered trademarks of National Semiconductor Corporation. All other
brand or product names may be trademarks or registered trademarks of their respective holders.
Copyright© 2007 National Semiconductor Corporation
For the most current product information visit us at www.national.com
National Semiconductor
Americas Customer
Support Center
Email:
[email protected]
Tel: 1-800-272-9959
www.national.com
National Semiconductor Europe
Customer Support Center
Fax: +49 (0) 180-530-85-86
Email: [email protected]
Deutsch Tel: +49 (0) 69 9508 6208
English Tel: +49 (0) 870 24 0 2171
Français Tel: +33 (0) 1 41 91 8790
National Semiconductor Asia
Pacific Customer Support Center
Email: [email protected]
National Semiconductor Japan
Customer Support Center
Fax: 81-3-5639-7507
Email: [email protected]
Tel: 81-3-5639-7560