找货询价

一对一服务 找料无忧

专属客服

服务时间

周一 - 周六 9:00-18:00

QQ咨询

一对一服务 找料无忧

专属客服

服务时间

周一 - 周六 9:00-18:00

技术支持

一对一服务 找料无忧

专属客服

服务时间

周一 - 周六 9:00-18:00

售后咨询

一对一服务 找料无忧

专属客服

服务时间

周一 - 周六 9:00-18:00

LM3000ASQ

型号:

LM3000ASQ

描述:

双同步仿电流模式控制器[ Dual Synchronous Emulated Current-Mode Controller ]

品牌:

NSC[ National Semiconductor ]

页数:

26 页

PDF大小:

788 K

July 2, 2009  
LM3000  
Dual Synchronous Emulated Current-Mode Controller  
General Description  
Features  
The LM3000 is a dual output synchronous buck controller  
which is designed to convert input voltages ranging from 3.3V  
to 18.5V down to output voltages as low as 0.6V. The two  
outputs switch at a constant programmable frequency of 200  
kHz to 1.5 MHz, with the second output 180 degrees out of  
phase from the first to minimize the input filter requirements.  
The switching frequency can also be phase locked to an ex-  
ternal frequency. A CLKOUT provides an external clock 90  
degrees out of phase with the main clock so that a second  
chip can be run out of phase with the main chip. The emulated  
current-mode control utilizes bottom side FET sensing to pro-  
vide fast transient response and current limit without the need  
for external current sense resistors or RC networks. Separate  
Enable, Soft-Start and Track pins allow each output to be  
controlled independently to provide maximum flexibility in de-  
signing system power sequencing.  
VIN range from 3.3V to 18.5V  
Output voltage from 0.6V to 80% of VIN  
Remote differential output voltage sensing  
1% accuracy at FB pin  
Interleaved operation reduces input capacitors  
Frequency sync/adjust from 200 kHz to 1.5 MHz  
Startup with pre-bias load  
Independent power good, enable, soft-start and track  
Programmable current limit without external sense resistor  
Hiccup mode short circuit protection  
Applications  
DC Power Distribution Systems  
Graphic Cards - GPU and Memory ICs  
The LM3000 has a full range of protection features which in-  
clude input under-voltage lock-out (UVLO), power good  
(PGOOD) signals for each output, over-voltage crowbar and  
hiccup mode during short circuit events.  
FPGA, CPLD, and ASICs  
Embedded Processor  
1.8V and 2.5V I/O Supplies  
Networking Equipment (Routers, Hubs)  
Simplified Application  
300905a1  
© 2009 National Semiconductor Corporation  
300905  
www.national.com  
Connection Diagram  
30090502  
Top View  
32-Lead LLP  
Ordering Information  
Order Number Package Marking  
Package Type  
32-Lead LLP  
32-Lead LLP  
32-Lead LLP  
32-Lead LLP  
NSC Package Drawing  
Supplied As  
LM3000ASQ  
LM3000ASQX  
LM3000SQ  
3000A  
3000A  
3000  
SQA32A  
SQA32A  
SQA32A  
SQA32A  
1000 Units Tape and Reel  
4500 Units Tape and Reel  
1000 Units Tape and Reel  
4500 Units Tape and Reel  
LM3000SQX  
3000  
www.national.com  
2
Pin Descriptions  
Pin #  
1
Name  
VSW2  
PGND2  
LG2  
Description  
Switch node sense for channel 2.  
2
Power ground for channel 2 low-side drivers.*  
Channel 2 low-side gate drive for external MOSFET.  
Chip supply voltage, input to the VDD and VDR regulators. (3.3V to 18.5V)  
Supply for low-side gate drivers.  
3
4
VIN  
5
VDR  
6
LG1  
Channel 1 low-side gate drive for external MOSFET.  
Power ground for channel 1 low-side drivers.*  
Switch node sense for channel 1.  
7
PGND1  
VSW1  
ILIM1  
8
9
Current limit setting input for channel 1.  
10  
11  
12  
13  
14  
15  
16  
17  
18  
19  
20  
21  
22  
23  
24  
25  
26  
27  
28  
29  
30  
31  
32  
HG1  
Channel 1 high-side gate drive for external MOSFET.  
Boost voltage for channel 1 high-side driver.  
Supply for control circuitry.  
VCB1  
VDD  
EA1_GND  
FB1  
Error amplifier ground sense for channel 1.*  
Error amplifier input for channel 1.  
COMP1  
PGOOD1  
FREQ/SYNC  
EN1  
Error amplifier output for channel 1.  
Power good signal for channel 1 under-voltage and over-voltage.  
Frequency set / synchronization input for internal PLL.  
Channel 1 enable input. Used to set the emulated current slope for channel 1.  
Channel 1 track input.  
TRK1  
SS1  
Channel 1 soft-start.  
TRK2  
Channel 2 track input.  
SS2  
Channel 2 soft-start.  
EN2  
Channel 2 enable input. Used to set the emulated current slope for channel 2.  
Power good signal for channel 2 under-voltage and over-voltage.  
Error amplifier output for channel 2.  
PGOOD2  
COMP2  
FB2  
Error amplifier input for channel 2.  
EA2_GND  
CLKOUT  
SGND  
VCB2  
HG2  
Error amplifier ground sense for channel 2.*  
Output clock. CLKOUT is shifted 90 degrees from SYNC input.  
Local signal ground.*  
Boost voltage for channel 2 high-side driver.  
Channel 2 high-side gate drive for external MOSFET.  
Current limit setting input for channel 2.  
ILIM2  
DAP  
Exposed die attach pad. Connect the DAP directly to SGND.*  
*The LM3000 offers true remote ground sensing to achieve very tight line and load regulation. For best layout practice, the EA1_GND, and EA2_GND should be  
tied to the ground end of the output capacitor (or output terminal) for VOUT1 and VOUT2 respectively. Inside the LM3000, the two power ground nodes PGND1 and  
PGND2 are physically isolated from each other and also isolated from the internal signal ground SGND. In order to achieve the best cross-channel noise rejection,  
it is advised to keep these three grounds isolated from each other for the most part in the board layout and only tie them together at the ground terminals.  
3
www.national.com  
Junction Temperature (TJ-MAX  
Storage Temperature Range  
Maximum Lead Temperature  
Soldering, 5 seconds  
ESD Rating  
)
150°C  
Absolute Maximum Ratings (Note 1)  
-65°C to +150°C  
If Military/Aerospace specified devices are required,  
please contact the National Semiconductor Sales Office/  
Distributors for availability and specifications.  
260°C  
2000V  
VIN to SGND, PGND  
-0.3V to 20V  
-3V to 20V  
-0.3V to 5.5V  
24V  
HBM (Note 2)  
VSW1, VSW2 to SGND, PGND  
VDD, VDR to SGND, PGND (Note 3)  
VCB1, VCB2 to SGND ,PGND  
VCB1 to VSW1, VCB2 to VSW2  
FB1, FB2 to SGND, PGND  
All other input pins to SGND, PGND  
(Note 4)  
Operating Ratings (Note 1)  
Input Voltage Range  
VDD = VDR = VIN (Note 3)  
VIN  
5.5V  
3.3V to 5.5V  
3.3V to 18.5V  
-0.3V to 3.0V  
Junction Temperature (TJ) Range  
−40°C to +125°C  
-0.3V to 5.5V  
Electrical Characteristics Limits in standard type are for TJ = 25°C only; limits in boldface type apply over the  
junction temperature (TJ) range of -40°C to +125°C. 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.  
Unless otherwise noted, VIN = 12.0V, IEN1 = IEN2 = 40 µA.  
Symbol  
Parameter  
Condition  
Min  
Typ  
0.6  
0.6  
0.6  
0.6  
0.15  
0.3  
0.1  
5
Max  
Units  
VFB  
FB Pin Voltage FB1, FB2 (LM3000A)  
-20°C to +85°C  
0.594  
0.591  
0.591  
0.588  
0.606  
0.609  
0.609  
0.612  
V
VFB  
FB Pin Voltage FB1, FB2 (LM3000)  
-20°C to +85°C  
V
Line Regulation VDD = VIN = VDR  
Line Regulation VIN > 6V  
Load Regulation  
3.3V < VIN < 5.5, COMP = 1.5V  
6V < VIN < 18.5V, COMP = 1.5V  
VIN = 12.0V, 1.0V < COMP < 1.4V  
%
%
ΔVFB/VFB  
%
Iq  
VIN Operating Current  
mA  
µA  
µA  
ISD  
IEN  
VIN Shutdown Current  
IEN1 , IEN2 < 5 µA  
IEN Rising  
50  
EN Input Threshold Current  
15  
35  
Hysteresis  
10  
ILIM  
ISS  
Source Current ILIM1, ILIM2  
Soft-Start Pull-Up Current  
COMP Pin Hiccup Thresholds  
VILIM1, VILIM2 = 0V  
VSS = 0.5V  
17  
20  
23  
µA  
µA  
5.5  
8.5  
2.85  
50  
11.5  
VHICCUP  
COMP Threshold High  
Hysteresis  
V
mV  
tDELAY  
tCOOL  
VOVP  
Hiccup Delay  
16  
Cycles  
Cycles  
%
Cool-Down Time Until Restart  
Over-Voltage Protection Threshold  
4096  
115  
3
As a % of Nominal Output Voltage  
Hysteresis  
110  
120  
VUVP  
Under-Voltage Protection Threshold  
As a % of REF1, REF2 (see Block  
Diagram)  
85  
%
GATE DRIVE  
ICB  
VCB Pin Leakage Current  
VCB - VSW = 5.5V  
250  
3
nA  
RDS1  
Top FET Drive Pull-Up On-Resistance VCB - VSW = 4.5V, VCB - HG = 100  
mV  
RDS2  
RDS3  
RDS4  
Top FET Drive Pull-Down On-Resistance VCB - VSW = 4.5V, HG - VSW = 100  
mV  
2
2
1
Bottom FET Drive Pull-Up On-  
Resistance  
VDR - PGND = 5V, VDR - LG = 100  
mV  
Bottom FET Drive Pull-Down On-  
Resistance  
VDR - PGND = 5V, LG - PGND = 100  
mV  
www.national.com  
4
Symbol  
OSCILLATOR  
fSW  
Parameter  
Condition  
Min  
Typ  
Max  
575  
Units  
Switching Frequency  
230  
500  
kHz  
kHz  
kHz  
V
RFRQ = 100 kΩ  
RFRQ = 42.2 kΩ  
425  
1550  
RFRQ = 10 kΩ  
Rising  
VSYNC  
Threshold for Synchronization at the  
FREQ/SYNC Pin  
2.2  
Falling  
0.6  
fSYNC  
tSYNC  
tSYNC-TRS  
DMAX  
ERROR AMPLIFIER  
SYNC Range  
200  
100  
1500  
kHz  
ns  
SYNC Pulse Width  
SYNC Rise/Fall Time  
Maximum Duty cycle  
10  
ns  
85  
%
IFB  
ISOURCE  
ISINK  
FB Pin Bias Current  
FB = 0.6V  
20  
80  
nA  
µA  
µA  
V
COMP Pin Source Current  
COMP Pin Sink Current  
FB = 0.5V, COMP = 1.0V  
FB = 0.7V, COMP = 0.7V  
80  
VCOMP-HI  
VCOMP-LO  
VOS-TRK  
gm  
COMP Pin Voltage High Clamp  
COMP Pin Voltage Low Clamp  
Offset Using TRK Pin  
2.80  
-9.0  
3.0  
0.48  
0
3.2  
9.0  
V
TRK = 0.45V  
mV  
µS  
MHz  
Transconductance  
1400  
10  
fBW  
Unity Gain Bandwidth Frequency  
INTERNAL VOLTAGE REGULATOR  
VVDD  
Internal Core Regulator Voltage  
UVLO Thresholds  
No External Load  
VDD Rising  
5.15  
2.12  
0.14  
1.1  
V
V
VVDD-ON  
Hysteresis  
VVDD-DO  
IVDD-ILIM  
Internal Core Regulator Dropout Voltage No External Load  
Internal Core Regulator Current Limit VDD Short to Ground  
Regulator for External MOSFET Drivers IVDR = 100 mA  
V
mA  
V
80  
VVDR  
5.2  
VVDR-DO  
Driver Regulator Dropout Voltage  
Driver Regulator Current Limit  
IVDR = 100 mA  
1.0  
V
IVDR-ILIM  
VDR Short to Ground  
450  
mA  
PGOOD OUTPUT  
RPG-ON  
PGOOD On-Resistance  
FB1 = FB2 = 0.47V  
VPGOOD = 5V  
250  
100  
nA  
IOH  
PGOOD High Leakage Current  
THERMAL RESISTANCE  
Junction-to-Ambient Thermal  
Resistance  
LLP-32 Package (Note 5)  
26.4  
°C/W  
θJA  
Note 1: Absolute Maximum Ratings indicate limits beyond which damage to the device may occur, including inoperability and degradation of device reliability  
and/or performance. Functional operation of the device and/or non-degradation at the Absolute Maximum Ratings or other conditions beyond those indicated in  
the Recommended Operating Conditions is not implied. Operating Range conditions indicate the conditions at which the device is functional and the device should  
not be operated beyond such conditions. For guaranteed specifications and conditions, see the Electrical Characteristics table.  
Note 2: Human Body Model (HBM) is 100 pF capacitor discharged through a 1.5k resistor into each pin. Applicable standard is JESD22-A114C.  
Note 3: VDD and VDR are outputs of the internal linear regulator. Under normal operating conditions where VIN > 5.5V, they must not be tied to any external  
voltage source. In an application where VIN is between 3.3V to 5.5V, it is recommended to tie the VDD, VDR and VIN pins together, especially when VIN may  
drop below 4.5V. In order to have better noise rejection under these conditions, a 10Ω, 1μF input filter may be used for the VDD pin.  
Note 4: HG1, HG2, LG1, LG2 and CLKOUT are all output pins and should not be tied to any external power supply. COMP1 and COMP2 are also outputs and  
should not be tied to any lower output impedance power source. PGOOD1 and PGOOD2 are open drain outputs, with a pull-down resistance of about 250.  
Each of them may be tied to an external voltage source less than 5.5V through an external resister greater than 3k, although 10kand above are preferred to  
reduce the necessary signal ground current.  
Note 5: Tested on a four layer JEDEC board. Four vias provided under the exposed pad. See JEDEC standards JESD51-5 and JESD51-7.  
5
www.national.com  
Typical Performance Characteristics  
3.3V Output Efficiency at 500 kHz  
1.2V Output Efficiency at 500 kHz  
30090517  
30090519  
3.3V Output Load and Line Regulation  
1.2V Output Load and Line Regulation  
30090518  
30090520  
FB1, FB2 Reference vs Temperature  
VDD Voltage vs Temperature  
30090503  
30090505  
www.national.com  
6
Soft-Start without Load  
No Load Soft-Start with Pre-Bias  
Soft-Start with Load  
Pulse Skipping during Over-Current Condition  
30090509  
30090510  
Output Short Circuit Hiccup  
30090511  
30090512  
Switch Node Short Circuit Hiccup  
30090513  
30090514  
7
www.national.com  
External Clock Synchronization  
External Tracking  
30090516  
30090515  
Error Amplifier Transconductance vs Temperature  
Enable Current Threshold vs Temperature  
30090507  
30090508  
Switching Frequency vs Temperature  
RFRQ vs Switching Frequency  
30090504  
30090506  
www.national.com  
8
Block Diagram  
30090521  
9
www.national.com  
tSS should be set to 75% of the minimum expected rise time  
of the controlling supply. In the event that the LM3000 is en-  
abled with a pre-biased master supply controlling track, the  
soft-start capacitor will control the tracking output voltage rise  
time. Pulling TRK down after a normal startup will cause the  
output voltage to follow the track signal.  
Functional Description  
THEORY OF OPERATION  
The LM3000 is a dual emulated current-mode PWM syn-  
chronous controller. Unlike traditional peak current-mode  
controllers which sense the current while the high-side FET  
is on, the LM3000 senses current while the low-side FET is  
on. It then emulates the peak current waveform and uses that  
information to regulate the output voltage. The blanking time  
when the high-side FET first turns on that is normally associ-  
ated with high-side sensing is not needed, allowing high-side  
ON pulses as low as 50 ns. The LM3000 therefore has both  
excellent line transient response and the ability to regulate low  
output voltages from high input voltages.  
STARTUP  
After the EN1 or EN2 current exceeds the enable ON thresh-  
old and the voltage at the VDD pin reaches 2.2V, an internal  
8.5 µA current source charges the soft-start capacitor of the  
enabled channel. Once soft-start is complete the converter  
enters steady state operation. Current limit is enabled during  
soft-start in case of a short circuit at the output. The soft-start  
time is calculated as:  
30090522  
FIGURE 1. Tracking with VOUT1 Controlling VOUT2  
Figure 1 shows a tracking example with the highest output  
voltage at VOUT1 controlling VOUT2. Tracking may be set so  
that VOUT1 and VOUT2 both rise together. For this case, the  
equation governing the values of the tracking divider resistors  
RT1 and RT2 is:  
To avoid current limit during startup, the soft-start time tSS  
should be substantially longer than the time required to  
charge COUT to VOUT at the maximum output current. To meet  
this requirement:  
STARTUP INTO OUTPUT PRE-BIAS  
A value of 10 k1% is recommended for RT1 as a good com-  
promise between high precision and low quiescent current  
through the divider. Using an example of VOUT1 = 3.3V and  
VOUT2 = 1.2V, the value of RT2 is 34.4 k1%. A timing diagram  
for VOUT1 controlling VOUT2 is shown in Figure 2. Note that the  
TRK pin must finish at least 100 mV higher than the 0.6V ref-  
erence to achieve the full accuracy of the LM3000 regulation.  
To meet this requirement the tracking voltage is offset by 150  
mV. The tracking output voltage will reach its final value at  
80% of the controlling output voltage.  
If the output capacitor of the LM3000 has been charged up to  
some pre-bias level before the converter is enabled, the chip  
will force the soft-start capacitor to the same voltage as the  
FB pin. This will cause the output to ramp up from the existing  
output voltage without discharging it. During the soft-start  
ramp, the low-side FET is disabled whenever the COMP volt-  
age is below the active regulation voltage range.  
LOW INPUT VOLTAGE  
The LM3000 includes an internal 5.2V linear regulator con-  
nected from the VIN pin to the VDD pin. This linear regulator  
feeds the logic and FET drive circuitry. For input voltages less  
than 5.5V, the VIN, VDD and VDR pins can be tied together  
externally. This allows the full input voltage to be used for  
driving the power FETs and also minimizes conduction loss  
in the LM3000.  
TRACKING  
The LM3000 has individual tracking inputs which control each  
output during soft-start. This allows the output voltage slew  
rates to be controlled for loads that require precise sequenc-  
ing. When the tracking function is not being used the TRK1  
or TRK2 pins should be connected directly to the VDD pin.  
30090524  
FIGURE 2. Tracking with VOUT1 Controlling VOUT2  
During start-up, the error amplifier will follow the lower of the  
SS or TRK voltages. For design margin, the soft-start time  
www.national.com  
10  
Alternatively, the tracking feature can be used to create equal  
slew rates for the output voltages. In order to track properly,  
use the highest output voltage to control the slew rate. In this  
case, the tracking resistors are found from:  
Again, a value of 10 k1% is recommended for RT1. For the  
example case of VOUT1 = 5V and VOUT2 = 1.8V, RT2 is 17.8  
k1%. A timing diagram for the case of equal slew rates is  
shown in Figure 3.  
Either method ensures that the output voltage of the tracking  
supply always reaches regulation before the output voltage of  
the controlling supply.  
30090594  
FIGURE 5. Tracking a Master Supply with Equal Start  
Time  
For equal slew rates, the circuit of Figure 6 is used. The re-  
lationship for the tracking divider is set by:  
30090526  
FIGURE 3. Tracking with Equal Slew Rates  
The LM3000 can track the output of a master power supply  
by connecting a resistor divider to the TRK pins as shown in  
Figure 4. For equal start times, the tracking resistors are de-  
termined by:  
30090595  
FIGURE 6. Tracking a Master Supply with Equal Slew  
Rates  
30090591  
FIGURE 4. Tracking a Master Supply with Equal Start  
Time  
11  
www.national.com  
sufficient amplitude of the signal at the FREQ/SYNC. It is  
possible to drive this pin directly from a 0 to 2.2V logic output,  
though not recommended for the typical application.  
Circuits that use an external clock should still have a resistor  
RFRQ connected from the FREQ/SYNC pin to ground. RFRQ is  
selected using the equation from the Frequency Setting sec-  
tion to match the external clock frequency. This allows the  
controller to continue operating at approximately the same  
switching frequency if the external clock fails and the coupling  
capacitor on the clock side is grounded or pulled to logic high.  
In the case of no external clock edges at startup, the internal  
oscillator will be controlled by the external set resistor until the  
first clock edge is detected. After the first edge, the PLL will  
lock within a few clock cycles, after which any missing edges  
will cause the oscillator to be programmed by RFRQ. If RFRQ  
is chosen to program the oscillator very close to the external  
clock frequency, the PLL will lock very quickly and there will  
be very little disturbance in the switching frequency.  
30090596  
FIGURE 7. Tracking a Master Supply with Equal Slew  
Rates  
Care must be taken to prevent errant pulses from triggering  
the synchronization circuitry. In circuits that will not synchro-  
nize to an external clock, CSYNC should be connected from the  
FREQ/SYNC pin to SGND as a noise filter. When a clock  
pulse is first detected, the LM3000 begins switching at the  
external clock frequency. Noise or a short burst of clock puls-  
es may result in variations of the switching frequency due to  
loss of lock by the PLL.  
Continuous Conduction Mode  
The LM3000 controls the output voltage by adjusting the duty  
cycle of the power MOSFETs with trailing edge pulse width  
modulation. The output inductor and capacitor filter the  
square wave produced as the power MOSFETs switch the  
input voltage, thereby creating a regulated output voltage.  
The dc level of the output voltage is determined by feedback  
resistors using the following equation:  
The output inductor current can flow from the drain to the  
source of the low-side MOSFET, which keeps the converter  
in continuous-conduction-mode (CCM). CCM has the advan-  
tage of constant frequency and nearly constant duty cycle (D  
= VOUT / VIN) over all load conditions, and also allows the  
converter to sink current at the output if needed.  
30090529  
FIGURE 8. Clock Synchronization Circuit  
FREQUENCY SETTING  
In the case where two LM3000 controllers are used, the CLK-  
OUT of the first controller can be used as a synchronization  
input for the second controller. Note that the CLKOUT is 90  
degrees out of phase with the main controller clock, so that  
the four phases of the two controllers are separated for min-  
imum input ripple current.  
The switching frequency of the internal oscillator is set by a  
resistor, RFRQ, connected from the FREQ/SYNC pin to  
SGND. The proper resistor for a desired switching frequency  
fSW can be selected from the curves in the Typical Perfor-  
mance Characteristics section labeled “RFRQ vs Switching  
Frequency” or by using the following equation:  
MOSFET GATE DRIVE  
The LM3000 has two sets of gate drivers designed for driving  
N-channel MOSFETs in a synchronous mode. Power for the  
high-side driver is supplied through the VCB pin. For the high-  
side gate HG to turn on the top FET, the VCB voltage must  
be at least one VGS(th) greater than VIN. This voltage is sup-  
plied from a local charge pump which consists of a Schottky  
diode and bootstrap capacitor, shown in Figure 9. For the  
Schottky, a rating of at least 250 mA and 30V is recommend-  
ed. A dual package may be used to supply both VCB1 and  
VCB2.  
Where fSW is the switching frequency in Hz.  
FREQUENCY SYNCHRONIZATION  
The switching frequency of the LM3000 can be synchronized  
by an external clock or other fixed frequency signal in the  
range of 200 kHz to 1.5 MHz. The external clock should be  
applied through a 100 pF coupling capacitor as shown in Fig-  
ure 8. In order for the oscillator to synchronize properly, the  
minimum amplitude of the SYNC signal is 2.2V and the max-  
imum amplitude is VDD. The minimum pulse width both pos-  
itive and negative is 100 ns. The nominal dc voltage at the  
FREQ/SYNC pin is 0.6V, which is also the clamp voltage level  
for the falling edge of the SYNC pulse. Depending on the  
pulse width and frequency, CSYNC may be adjusted to provide  
Both the bootstrap and the low-side FET driver are fed from  
VDR, which is the output of a 5V internal linear regulator. This  
regulator has a dropout voltage of approximately 1V. The  
drive voltage for the top FET driver is about VDR - 0.5 at light  
load condition and about VDR at normal to full load condition.  
This information is needed to select the type of MOSFETs  
used, as well as calculate the losses in driving them.  
www.national.com  
12  
HICCUP MODE  
During hiccup mode the LM3000 disables both the high-side  
and low-side MOSFETs, and remains in this state for 4096  
switching cycles. After this cool down period the circuit  
restarts again through the normal soft-start sequence. If the  
shorted fault condition persists, hiccup will retrigger once the  
soft-start has finished. This occurs when the SS voltage is  
greater than 0.7V and switching has reached the continuous  
conduction mode state.  
There is a coarse high-side current limit which senses the  
voltage across the high-side MOSFET. The threshold is ap-  
proximately 0.5V, which may provide some level of protection  
for a catastrophic fault. Hiccup will immediately trigger after  
two consecutive high-side current limit fault events.  
30090530  
POWER GOOD  
FIGURE 9. Bootstrap Circuit  
Power good pins PGOOD1 and PGOOD2 are available to  
monitor the output status of the two channels independently.  
The PGOOD1 pin connects to the output of an open drain  
MOSFET, which will remain open while Channel 1 is within  
the normal operating range. PGOOD1 goes low (low  
impedance to ground) under the following three conditions:  
UVLO  
For the case where VIN is > VDD, the VIN UVLO thresholds  
are determined by the VDD UVLO comparator and the VDD  
dropout voltage. This sets the rising threshold for VIN at ap-  
proximately 3V, with 30 mV of hysteresis.  
1. Channel 1 is turned off.  
2. OVP on Channel 1.  
3. UVP on Channel 1.  
For the case where VIN is < 5.5V and tied to VDD and VDR,  
the UVLO trip point is 2.12V rising. UVLO consists of turning  
off the top and bottom FETs and remaining in that condition  
until VDD rises above 2.12V. The falling trip point is 140 mV  
below the rising trip point.  
PGOOD2 functions in a similar manner. UVP tracks REF1,  
REF2 as shown in the block diagram. OVP sets a fault which  
turns off the high gate and turns on the low gate. This dis-  
charges the output voltage until it has fallen 3% below the  
OVP threshold.  
CURRENT LIMIT  
The current limit of the LM3000 is realized by sensing the  
current in the low-side FET while the output current circulates  
through it. This voltage (IOUT x RDS(on)_LO) is compared against  
the voltage of a fixed, internal 20 µA current source and a  
user-selected resistor, RLIM, connected between the switch  
node and the ILIM pin. Once a current limit event is sensed,  
the high-side switch is disabled for the following cycle and the  
low-side FET is kept on during this time. If sixteen consecutive  
current limit cycles occur, the part enters hiccup mode.  
PGOOD may be pulled up through a resistor to any voltage  
which is < 5.5V. When using VDD for the pull-up voltage, a  
typical value of 100 kis used to minimize loading on VDD.  
ENABLE  
A fixed external voltage source and resistors to EN1 and EN2  
are used to independently enable each output. The LM3000  
can be put into a low power shutdown mode by pulling the  
EN1 and EN2 pins to ground, or by applying 0V to the enable  
resistors. During shutdown both the high-side and low-side  
FETs are disabled. The quiescent current during shutdown is  
approximately 30 µA.  
The value of RLIM for a desired current limit IILIMIT can be se-  
lected by the following equation:  
The enable pins also control the emulated current ramp am-  
plitude by programming the current into EN1 and EN2. The  
recommended range for IEN is 40 μA to 160 μA. See the Ap-  
plications Information section under Control Loop Compen-  
sation for the complete design method.  
13  
www.national.com  
Application Information  
The most common circuit controlled by the LM3000 is a non-  
isolated, synchronous buck regulator. The buck regulator  
steps down the input voltage and has a duty ratio D of:  
Where η is the estimated converter efficiency.  
The following is a design example selecting components for  
the Typical Application Schematic of Figure 24. The circuit is  
designed for two outputs of 3.3V at 8A and 1.2V at 15A from  
an input voltage of 6V to 18V. This circuit is typical of a ‘brick’  
module and has a height requirement of 6.5mm or less. Other  
assumptions used to aid in circuit design are that the expected  
load is a small microprocessor or ASIC with fast load tran-  
sients, and that the type of MOSFETs used are in SO-8 or its  
equivalent packages such as PowerPAK®, PQFN and LFPAK  
(LFPAK-i).  
Where QGD is the high-side FET Miller charge with a VDS  
swing between 0 to VIN; CISS is the input capacitance of the  
high-side MOSFET in its off state with VDS = VIN. α and β are  
fitting coefficient numbers, which are usually between 0.5 to  
1, depending on the board level parasitic inductances and re-  
verse recovery of the low-side power MOSFET body diode.  
Under ideal condition, setting α = β = 0.5 is a good starting  
point. Other variables are defined as:  
IL_VL = IOUT - 0.5 x ΔIL  
IL_PK = IOUT + 0.5 x ΔIL  
SWITCHING FREQUENCY  
The selection of switching frequency is based on the tradeoff  
between size, cost and efficiency. In general, a lower fre-  
quency means larger, more expensive inductors and capac-  
itors. A higher switching frequency generally results in a  
smaller but less efficient solution, because the power MOS-  
FET gate capacitances must be charged and discharged  
more often in a given amount of time. For this application a  
frequency of 500 kHz is selected. 500 kHz is a good compro-  
mise between the size of the inductor and MOSFETs, tran-  
sient response and efficiency. Following the equation given  
for RFRQ in the Frequency Setting section, for 500 kHz oper-  
ation a 42.2 k1% resistor is used.  
RG_ON = 8.5 + RG_INT + RG_EXT  
RG_OFF = 2.8 + RG_INT + RG_EXT  
Switching loss is calculated for the high-side FET only. 8.5  
and 2.8 represent the LM3000 high-side driver resistance in  
the transient region. RG_INT is the gate resistance of the high-  
side FET, and RG_EXT is the external gate resistance if appli-  
cable. RG_EXT may be used to damp out excessive parasitic  
ringing at the switch node.  
MOSFETS  
Selection of the power MOSFETs is governed by a tradeoff  
between size, cost and efficiency. Buck regulators that use a  
controller IC and discrete MOSFETs tend to be most efficient  
for output currents of 4A to 20A.  
For this example, the maximum drain-to-source voltage ap-  
plied to either MOSFET is 18V. The maximum drive voltage  
at the gate of the high-side MOSFET is 5V, and the maximum  
drive voltage for the low-side MOSFET is 5V. The selected  
MOSFET must be able to withstand 18V plus any ringing from  
drain to source, and be able to handle at least 5V plus ringing  
from gate to source. If the duty cycle of the converter is small,  
then the high-side MOSFET should be selected with a low  
gate charge in order to minimize switching loss whereas the  
bottom MOSFET should have a low RDSONto minimize con-  
duction loss.  
Losses in the high-side FET can be broken down into con-  
duction loss, gate charge loss and switching loss. Conduc-  
tion, or I2R loss is approximately:  
PCOND_HI = D x (IOUT2 x RDS(on)_HI x 1.3)  
(High-side FET)  
PCOND_LO = D x (IOUT2 x RDS(on)_LO x 1.3)  
(Low-side FET)  
In the above equations the factor 1.3 accounts for the in-  
crease in MOSFET RDS(on) due to self heating. Alternatively,  
the 1.3 can be ignored and the RDS(on) of the MOSFET esti-  
mated using the RDS(on) vs. Temperature curves in the MOS-  
FET datasheets.  
For a typical input voltage of 12V and output currents of 8A  
and 12A, the MOSFET selections for the design example are  
HAT2168 for the high-side MOSFET and RJK0330DPB for  
the low-side MOSFET.  
The gate charge loss results from the current driving the gate  
capacitance of the power MOSFETs, and is approximated as:  
A 3resistor for RCBT is added in series with the VDR regu-  
lator output, as shown in Figure 24. This helps to control the  
MOSFET turn-on and ringing at the switch node, without af-  
fecting the MOSFET turn-off.  
PDR = VIN x (QG_HI + QG_LO) x fSW  
Where QG_HI and QG_LO are the total gate charge of the high-  
side and low-side FETs respectively at the typical 5V driver  
voltage. Gate charge loss differs from conduction and switch-  
ing losses in that the majority of dissipation occurs in the  
LM3000.  
To improve efficiency, 3A, 40V Schottky diodes are placed  
across the low-side MOSFETs. The external Schottky diodes  
have a much lower forward voltage than the MOSFET body  
diode, and help to minimize the loss due to the body diode  
recovery characteristic.  
The switching loss occurs during the brief transition period as  
the FET turns on and off, during which both current and volt-  
age are present in the channel of the FET. This can be  
approximated as the following:  
www.national.com  
14  
OUTPUT INDUCTORS  
Where ΔVO (V) is the peak to peak output voltage ripple, ΔIL  
(A) is the peak to peak inductor ripple current, RC (Ω) is the  
equivalent series resistance or ESR of the output capacitor,  
fSW (Hz) is the switching frequency, and CO (F) is the output  
capacitance. The amount of output ripple that can be tolerated  
is application specific. A general recommendation is to keep  
the output ripple less than 1% of the rated output voltage. The  
output capacitor selection will also affect the output voltage  
droop and overshoot during a load transient. The peak tran-  
sient of the output voltage during a load current step is de-  
pendent on many factors. Given sufficient control loop  
bandwidth an approximation of the transient voltage can be  
obtained from:  
The first criterion for selecting an output inductor is the induc-  
tance itself. In most buck converters, this value is based on  
the desired peak-to-peak ripple current, ΔIL that flows in the  
inductor along with the load current. As with switching fre-  
quency, the selection of the inductor is a tradeoff between size  
and cost. Higher inductance means lower ripple current and  
hence lower output voltage ripple. Lower inductance results  
in smaller, less expensive devices. An inductance that gives  
a ripple current of 1/6 to 1/3 of the maximum output current is  
a good starting point. (ΔIL = (1/6 to 1/3) x IOUT). Minimum in-  
ductance is calculated from this value, using the maximum  
input voltage as:  
By calculating in terms of amperes, volts, and megahertz, the  
inductance value will come out in micro henries.  
Where VP (V) is the output voltage transient and ΔIO (A) is the  
load current step change. CO (F) is the output capacitance, L  
(H) is the value of the inductor and RC (Ω) is the series resis-  
tance of the output capacitor. VL (V) is the minimum inductor  
voltage, which is duty cycle dependent.  
The inductor ripple current is found from the minimum induc-  
tance equation:  
For D < 0.5, VL = VOUT  
For D > 0.5, VL = VIN - VOUT  
This shows that as the input voltage approaches VOUT, the  
transient droop will get worse. The recovery overshoot re-  
mains fairly constant.  
The second criterion is inductor saturation current rating. The  
LM3000 has an accurately programmed valley current limit.  
During an instantaneous short, the peak inductor current can  
be very high due to a momentary increase in duty cycle. Since  
this is limited by the coarse high-side switch current limit, it is  
advised to select an inductor with a larger core saturation  
margin and preferably a softer roll off of the inductance value  
over load current.  
The loss associated with the output capacitor series resis-  
tance can be estimated as:  
For the design example, standard values of 1.2 μH for the  
1.2V, 15A output and 2.7 μH for the 3.3V, 8A output are cho-  
sen to fall within the ΔIL = (1/6 to 1/3) x IOUT range.  
Output Capacitor Design Procedure  
For the design example VIN = 12V, VOUT = 3.3V, D = VOUT  
/
The dc loss in the inductor is determined by its series resis-  
tance RL. The dc power dissipation is found from:  
VIN = 0.275, L = 2.7 μH, ΔIL = 1.8A, ΔIO = 8A and VP = 0.15V.  
To meet the transient voltage specification, the maximum  
RC is:  
PDC = IOUT2 x RL  
The ac loss can be estimated from the inductor  
manufacturer’s data, if available. The ac loss is set by the  
peak-to-peak ripple current ΔIL and the switching frequency  
fSW  
.
For the design example, the maximum RC is 18.75 m.  
Choose RC = 15 mas the design limit.  
From the equation for VP, the minimum value of CO is:  
OUTPUT CAPACITORS  
The output capacitors filter the inductor ripple current and  
provide a source of charge for transient load conditions. A  
wide range of output capacitors may be used with the LM3000  
that provide excellent performance. The best performance is  
typically obtained using aluminum electrolytic, tantalum, poly-  
mer, solid aluminum, organic or niobium type chemistries in  
parallel with a ceramic capacitor. The ceramic capacitor pro-  
vides extremely low impedance to reduce the output ripple  
voltage and noise spikes, while the aluminum or other capac-  
itors provide a larger bulk capacitance for transient loading  
and series resistance for stability.  
For D < 0.5, VL = VOUT  
For D > 0.5, VL = VIN - VOUT  
With RC = VP / ΔIO this reduces to:  
When selecting the value for the output capacitor the two per-  
formance characteristics to consider are the output voltage  
ripple and transient response. The output voltage ripple can  
be approximated as:  
15  
www.national.com  
With RC = 0 this reduces to:  
For the dual output design operating 180° out of phase, the  
general equation for the input capacitor rms current is ap-  
proximated as:  
Since D < 0.5, VL = VOUT. With RC = 15 m, the minimum  
value for CO is 218 μF.  
The minimum control loop bandwidth fC is given by:  
Where the output currents are I1, I2 and the duty cycles are  
D1, D2 respectively. D3 represents the overlapping effective  
duty cycle, which adds to the RMS current.  
For the design example, the minimum value for fC is 39 kHz.  
A 220 μF, 15 mpolymer capacitor in parallel with a 22 μF,  
3 mceramic will meet the target output voltage ripple and  
transient specification.  
If D > 0.5 for both or D < 0.5 for both, the worst case rms  
current occurs with one output at full load and the other at no  
load. The maximum rms current can be approximated as:  
For the 1.2V, 15A output, two 220 μF, 15 mpolymer capac-  
itors in parallel with a 22 μF, 3 mceramic are chosen to meet  
the target design specifications.  
INPUT CAPACITORS  
If D > 0.5 for one and D < 0.5 for the other, the worst case rms  
current becomes:  
The input capacitors for a buck regulator are used to smooth  
the large current pulses drawn by the inductor and load when  
the high-side MOSFET is on. Due to this large ac stress, input  
capacitors are usually selected on the basis of their ac rms  
current rating rather than bulk capacitance. Low ESR is ben-  
eficial because it reduces the power dissipation in the capac-  
itors. Although any of the capacitor types mentioned in the  
Output Capacitor section can be used, ceramic capacitors are  
common because of their low series resistance. In general the  
input to a buck converter does not require as much bulk ca-  
pacitance as the output.  
In most applications for point-of-load power supplies, the in-  
put voltage is the output of another switching converter. This  
output often has a lot of bulk capacitance, which may provide  
adequate damping.  
When the converter is connected to a remote input power  
source through a wiring harness, a resonant circuit is formed  
by the line impedance and the input capacitors. If step input  
voltage transients are expected near the maximum rating of  
the LM3000, a careful evaluation of the ringing and possible  
overshoot at the device VIN pin should be completed. To  
minimize overshoot make CIN > 10 x LIN. The characteristic  
source impedance and resonant frequency are:  
The input capacitors should be selected for rms current rating  
and minimum ripple voltage. The equation for the rms current  
and power loss of the input capacitor in a single phase can  
be estimated as:  
Where IO (A) is the output load current and RCIN (Ω) is the  
series resistance of the input capacitor. Since the maximum  
values occur at D = 0.5, a good estimate of the input capacitor  
rms current rating in a single phase is one-half of the maxi-  
mum output current.  
The converter exhibits a negative input impedance which is  
lowest at the minimum input voltage:  
Neglecting the series inductance of the input capacitance, the  
input voltage ripple for a single phase can be estimated as:  
The damping factor for the input filter is given by:  
By defining the maximum input voltage ripple, the minimum  
requirement for the input capacitance can be calculated as:  
Where RLIN is the input wiring resistance and RCIN is the series  
resistance of the input capacitors. The term ZS / ZIN will always  
be negative due to ZIN.  
When δ = 1, the input filter is critically damped. This may be  
difficult to achieve with practical component values. With δ <  
0.2, the input filter will exhibit significant ringing. If δ is zero or  
www.national.com  
16  
negative, there is not enough resistance in the circuit and the  
input filter will sustain an oscillation.  
50V, 0.18, 670 mA capacitor in a 10 mm x 10.2 mm package  
is chosen for each input. Calculated rms current for the 3.3V  
phase is 322 mA, with 242 mA calculated for the 1.2V phase.  
When operating near the minimum input voltage, an alu-  
minum electrolytic capacitor across CIN may be needed to  
damp the input for a typical bench test setup. Any parallel  
capacitor should be evaluated for its rms current rating. The  
current will split between the ceramic and aluminum capaci-  
tors based on the relative impedance at the switching fre-  
quency. Using a square wave approximation, the rms current  
in each capacitor is found from:  
CURRENT LIMIT  
For the design example, the desired current limit set point is  
chosen to be 150% of the maximum load current. To account  
for the tolerance of the internal current source and allowing  
RDS(on) = 4 mfor the low-side MOSFET at elevated temper-  
ature, a target of 23A is used for the 1.2V output, with 13A for  
the 3.3V output. Following the equation from the Current Limit  
section the values for RLIM are 4.64 k, 1% for the 1.2V output  
and 2.67 k, 1% for the 3.3V output.  
TRACK  
Tracking for the design example is configured such that  
VOUT1 is controlling VOUT2. The divider values are set so that  
both outputs will rise together, with VOUT2 reaching its final  
value just before VOUT1. Following the method in the Tracking  
section and allowing for a 120 mV offset between FB and  
TRK, standard 1% values are selected for RT1 = 10 kand  
RT2 = 35.7 kΩ.  
SOFT START  
To prevent over-shoot, the soft start time is set to be longer  
than the time it would take to charge the output voltage at  
current limit. Following the equations in the Startup section for  
VOUT1 and VOUT2  
:
Input Capacitor Design Procedure  
tSS1(MIN) = (3.3V x 242 μF) / (13A - 8A) = 160 μs  
tSS2(MIN) = (1.2V x 462 μF) / (23A - 15A) = 69 μs  
Choosing a value of CSS1 = 27 nF, the soft start time is:  
tSS1 = (27 nF x 0.6V) / 8.5 μA = 1.9 ms  
Ceramic capacitors are sized to support the required rms cur-  
rent. Aluminum electrolytic capacitors are used for damping.  
Treating each phase separately, find the minimum value for  
the ceramic capacitor from:  
To ensure that VOUT2 tracks VOUT1, tSS2 is set at two-thirds of  
tSS1 by making CSS2 = 18 nF.  
VDD, VDR and VCB CAPACITORS  
For the design example allowing 0.25V input voltage ripple,  
the worst case occurs for the 3.3V, 8A output at D = 0.5. The  
minimum value is CIN = 16 μF. For the 1.2V, 15A output, the  
worst case D = 1.2V / 6V = 0.2. Then CIN = 4.8 μF. Find the  
rms current rating for each from:  
VDD is used as the supply for the internal control and logic  
circuitry. A 1 μF ceramic capacitor provides sufficient filtering  
for VDD.  
VDR provides power for both the high-side and low-side  
MOSGET gate drives, and is sized to meet the total gate drive  
current. Allowing for ΔVVDR = 100 mV of ripple, the minimum  
value for CVDR is found from:  
Using the same criteria, results are 4A rms for the 3.3V phase  
and 3A rms for the 1.2V phase. Manufacturer data for 10 μF,  
25V, X5R capacitors in a 1206 package allows for 3A rms with  
a 20°C temperature rise. For the design example, using two  
ceramic capacitors for each phase will meet both the input  
voltage ripple and rms current target. Since the series resis-  
tance is so low at about 5 mper capacitor, a parallel alu-  
minum electrolytic is used for damping. A good general rule  
is to make the damping capacitor at least five times the value  
of the ceramic. By sizing the aluminum such that it is primarily  
resistive at the switching frequency, the design is greatly sim-  
plified since the ceramic is primarily reactive. In this case the  
approximation for the rms current in the damping capacitor is:  
Using QG_HI = 15 nC and QG_LO = 30 nC with a 5V gate drive,  
the minimum value for CVDR = 0.45 μF.  
VCB provides power for the high-side gate drive, and is sized  
to meet the required gate drive current. Allowing for ΔVVCB  
=
100 mV of ripple, the minimum value for CBOOT is found from:  
To use the minimum number of different components, CVDR  
and CBOOT are also selected as 1 μF ceramic for the design  
example.  
Where CIN2 is the damping capacitance, RCIN2 is its series  
resistance and CIN1 is the ceramic capacitance. A 150 μF,  
17  
www.national.com  
CONTROL LOOP COMPENSATION  
bilizes the modulator gain from variations in MOSFET resis-  
tance over temperature, providing a robust design solution.  
The LM3000 uses emulated peak current-mode PWM control  
to correct changes in output voltage due to line and load tran-  
sients. This unique architecture combines the fast line tran-  
sient response of peak current-mode control with the ability  
to regulate at very low duty cycles. In order to facilitate the  
use of MOSFET RDS(on) sensing, the control ramp is set by  
the enable voltage and a resistor to the enable pin. This sta-  
The control loop is comprised of two parts. The first is the  
power stage, which consists of the duty cycle modulator, out-  
put filter and load. The second part is the error amplifier, which  
is a transconductance amplifier with a typical gm of 1400  
μmho (or 1400 μS). Figure 10 shows the power stage and  
error amplifier components.  
30090546  
FIGURE 10. Power Stage and Error Amplifier  
The power stage transfer function (also called the control-to-  
output transfer function) in a buck converter can be written as:  
For the emulated peak current-mode control, Km is the dc  
modulator gain and Ri is the current-sense gain. KSL is the  
proportional slope compensation, which is set by the enable  
resistor REN and enable voltage VEN  
.
Figure 11 shows a more detailed view of the current sense  
amplifier, which includes a three stage filter for increased  
noise immunity. The effective gain and phase are shown in  
Figure 12 and Figure 13. The equivalent current sense gain  
A = 7.  
Where:  
With:  
30090550  
FIGURE 11. Current Sense Amplifier and Filter  
www.national.com  
18  
on above the threshold. A minimum enable voltage of 3V is  
recommended to keep the temperature coefficient of the  
0.75V internal VBE from becoming a significant error term.  
30090553  
FIGURE 12. Current Sense Amplifier Gain  
30090566  
FIGURE 14. Maximum Enable Current vs. Input Voltage  
Typical frequency response of the gain and the phase for the  
power stage are shown in Figure 15 and Figure 16. It is de-  
signed for VIN = 12V, VOUT = 3.3V, IOUT = 8A, VEN = 5V and a  
switching frequency of 500 kHz. The power stage component  
values are:  
L = 2.7 μH, RL = 3.4 m, CO1 = 220 μF, RC1 = 15 m, CO2  
22 μF, RC2 = 3 m, RO = VOUT / IOUT = 0.41Ω, RS = RDS(on)  
4 mand REN = 43 kΩ.  
=
=
30090554  
FIGURE 13. Current Sense Amplifier Phase  
A relatively high value of slope compensating ramp is used to  
stabilize the gain. This minimizes the effect of the current  
sense filter on the control loop and swamps out the need for  
a sampling-gain term. When designing within the recom-  
mended operating range, there is no tendency toward sub-  
harmonic oscillation. The proportional slope compensation is  
defined as:  
30090567  
FIGURE 15. Power Stage Gain  
ISL is the internal current source scale factor, KSW is the  
switching frequency correction factor and IEN is the external  
enable current. The recommended range for IEN is 40 μA to  
160 μA. With VEN = 5V, this corresponds to a range for REN  
of 25 kto 100 k. For operation below 4.2V input, the max-  
imum enable current is limited, as shown in Figure 14. At the  
minimum input of 3.3V, a value of 80 μA maximum corre-  
sponds to REN = 50 kwith VEN = 5V. The minimum enable  
current is set by the enable bias circuit to ensure proper turn-  
19  
www.national.com  
30090569  
30090568  
FIGURE 18. Transconductance Amplifier Open Loop  
Gain  
FIGURE 16. Power Stage Phase  
The effective total PWM ramp height is controlled by REN  
.
Higher REN creates a higher ramp voltage, providing more  
noise immunity and less variation in the modulator gain over  
temperature. Lower REN requires less RC (output capacitor  
ESR) for the desired phase margin and a more ideal current-  
mode behavior.  
Figure 17 shows the transconductance amplifier network,  
which takes the output impedance of the amplifier and the  
internal filter into account. To simplify the analysis, the 12.75  
kand 10 pF internal filter is absorbed into the transconduc-  
tance amplifier. This produces an equivalent REA = 15 Mand  
CBW = 22 pF for an effective 10 MHz unity gain bandwidth.  
30090575  
FIGURE 19. Transconductance Amplifier Open Loop  
Phase  
Assuming a pole at the origin, the simplified equation for the  
error amplifier transfer function can be written in terms of the  
mid-band gain as:  
30090555  
FIGURE 17. Equivalent Transconductance Amplifier and  
COMP Filter  
www.national.com  
20  
Where:  
Calculate the parallel equivalent CO and RC at the target  
crossover frequency:  
In general, the goal of the compensation circuit is to give high  
dc gain, a bandwidth that is between one-fifth and one-tenth  
of the switching frequency, and at least 45° of phase margin.  
Control Loop Design Procedure  
Once the power stage design is complete, the power stage  
components are used to determine the proper frequency  
compensation. By equating the power stage transfer function  
to the error amplifier transfer function term by term, the control  
loop design procedure targets an ideal single-pole system re-  
sponse.  
For the design example X1 = 0.00723, X2 = 0.0723, Z =  
0.01478 and A = 0.6304. The parallel equivalent CO = 183  
μF and RC = 11.9 mΩ.  
Find the optimal value of the enable current:  
The compensation components will scale from the feedback  
divider ratio and selection of the bottom feedback divider re-  
sistor. A maximum value for the divider current is typically set  
at 1 mA. Using a divider current of 200 μA will allow for a  
reasonable range of values. For the bottom feedback resistor  
RFBB = VREF / 200 μA = 3 k. Choosing a standard 1% value  
of 2.94 k, the top feedback resistor is found from:  
If IEN is not within the range of 40μA to 160μA use either the  
minimum or maximum limit. Find REN from:  
For VOUT = 3.3V and VREF = 0.6V, RFBT = 13.2 kΩ.  
Based on the previously defined power stage values, calcu-  
late general terms:  
For the design example IEN = 95.5 μA and REN = 44.7 k.  
Choosing a standard value of 43 k, IEN = 94.4 μA.  
Calculate other general terms:  
For the design example D = 0.275, Ri = 0.028Ω, T = 2 μs,  
KSW = 1.147 and KFB = 0.1818.  
Choose a target crossover frequency fC greater than the min-  
imum control loop bandwidth from the Output Capacitors  
section. This is typically set between 1/10 and 1/5 of the  
switching frequency.  
For the design example KSL = 0.0978, Km = 10.7 and KD  
1.73.  
=
If the enable resistor has been adjusted from the nominal val-  
ue to provide more noise immunity or to meet the minimum  
input voltage limit, calculate the optimal value of RC. The min-  
imum value of RC to maintain adequate phase margin for  
stability is about half this value.  
Choosing fC = 100 kHz for the design example ωC = 628 krad/  
sec. The switching frequency ωSW = 3.14 Mrad/sec and the  
error amplifier bandwidth ωBW = 62.8 Mrad/sec.  
Checking for the design example RC = 9.1 mΩ.  
21  
www.national.com  
Calculate the compensation components:  
The complete control loop transfer function is equal to the  
product of the power stage transfer function and error ampli-  
fier transfer function. For the Bode plots, the overall loop gain  
is the equal to the sum in dB and the overall phase is equal  
to the sum in degrees. Results are shown in Figure 22 and  
Figure 23. The crossover frequency is 100 kHz with a phase  
margin of 75°.  
For the design example, the calculated values are CBW = 22  
pF, CFF = 904 pF, CHF = 11 pF, CCOMP = 2505 pF and  
RCOMP = 9523Ω.  
Using standard values of CFF = 820 pF, CHF = 10 pF, CCOMP  
= 2200 pF and RCOMP = 10 k, the error amplifier plots of gain  
and phase are shown in Figure 20 and Figure 21.  
30090578  
FIGURE 22. Control Loop Gain  
30090576  
FIGURE 20. Error Amplifier Gain  
30090579  
FIGURE 23. Control Loop Phase  
Compensator design for the 1.2V output is similar. With  
VREF = 0.6V, the feedback divider resistors are chosen as  
RFBB = RFBT = 22.6 k. This results in a divider current of  
about 25 μA, which is considered to be the minimum accept-  
able level. With VEN = 5V, the nearest standard value to meet  
the optimal enable current is REN = 62 k. For a target  
crossover frequency of 100 kHz, standard values are CFF  
=
220 pF, CHF = 10 pF, CCOMP = 2200 pF and RCOMP = 10 kΩ.  
For the small-signal analysis, it is assumed that the control  
voltage at the COMP pin is dc. In practice, the output ripple  
voltage is amplified by the error amplifier gain at the switching  
frequency, which appears at the COMP pin adding to the  
control ramp. This tends to reduce the modulator gain, which  
may lower the actual control loop crossover frequency.  
30090577  
FIGURE 21. Error Amplifier Phase  
www.national.com  
22  
The EA_GND pins should also be connected with a separate  
Kelvin trace, running from the output ground sense point. The  
sense output, which is connecting to the top of the feedback  
resistor divider, should also run with a dedicated Kelvin trace  
together with the EA_GND. Keep these lines away from the  
switch node and output inductor to avoid stray coupling. If  
possible, the FB and EA_GND traces should be shielded from  
the switch node by ground planes. If necessary, the feedback  
divider impedance may be lowered to improve noise immu-  
nity.  
Efficiency and Thermal  
Considerations  
The total power dissipated in the power components can be  
obtained by adding together the loss as mentioned in the  
MOSFET, input capacitor, output capacitor and output induc-  
tor sections.  
The efficiency is defined as:  
SEPARATE PGND AND SGND  
Good layout techniques include a dedicated signal ground  
plane, usually on an internal layer adjacent to the LM3000 and  
signal component side of the board. Signal level components  
like the compensation and feedback resistors should be con-  
nected to this internal plane. The SGND pin should connect  
directly to the DAP, with vias from the DAP to the signal  
ground plane. Separate power ground plane areas for each  
phase should be made on the power component side of the  
board, as well as other layers. This allows separate lines for  
each PGND pin to connect to its respective power ground  
plane area at each low-side MOSFET source. The signal  
ground plane is then connected to a quiet point on each power  
ground plane area. These connections are typically made at  
the common input/output power terminals or capacitor re-  
turns. An equivalent schematic representation is shown in the  
Typical Application Schematic of Figure 24.  
The highest power dissipating components are the power  
MOSFETs. The easiest way to determine the power dissipat-  
ed in the MOSFETs is to measure the total conversion loss  
(PIN - POUT), then subtract the power loss in the capacitors,  
inductors and LM3000. The resulting power loss is primarily  
in the switching MOSFETs. Selecting MOSFETs with ex-  
posed pads will aid the power dissipation of these devices.  
Careful attention to RDS(on) at high temperature should be ob-  
served.  
LM3000 OPERATING LOSS  
This term accounts for the current drawn at the VIN pin, used  
for driving the logic circuitry and the power MOSFETs. For the  
LM3000, this current is equal to the steady state operating  
current Iq plus the MOSFET gate charge current IGC, which is  
defined as:  
MINIMIZE THE SWITCH NODE  
IGC = (QG_HI + QG_LO) x fSW  
The copper area that connects the power MOSFETs and out-  
put inductor together radiates more EMI as it gets larger. Use  
just enough copper to give low impedance for the switching  
currents and provide adequate heat spreading for the MOS-  
FETs.  
PD = VIN x (Iq + IGC  
)
Where PD represents the total power dissipated in the  
LM3000. Iq is about 5 mA from the Electrical Characteristics  
table. The LM3000 has an exposed thermal pad to aid power  
dissipation.  
LOW IMPEDANCE POWER PATH  
In a buck regulator the primary switching loop consists of the  
input capacitor connection to the MOSFETs. Minimizing the  
area of this loop reduces the stray inductance, which mini-  
mizes noise and possible erratic operation. The ceramic input  
capacitors should be placed as close as possible to the MOS-  
FETs, with the VIN side of the capacitors connected directly  
to the high-side MOSFET drain, and the PGND side of the  
capacitors connected as close as possible to the low-side  
source. The complete power path includes the input capaci-  
tors, power MOSFETs, output inductor, and output capaci-  
tors. Keep these components on the same side of the board  
and connect them with thick traces or copper planes. Avoid  
connecting these components through vias whenever possi-  
ble, as vias add inductance and resistance. In general, the  
power components should be kept close together, minimizing  
the circuit board losses.  
Layout Considerations  
To produce an optimal power solution with a switching con-  
verter, as much care must be taken with the layout and design  
of the printed circuit board as with the component selection.  
The following are several guidelines to aid in creating a good  
layout.  
KELVIN TRACES FOR GATE DRIVE AND SENSE LINES  
The HG and SW pins provide the gate drive and return for the  
high-side MOSFET. Likewise the LG and PGND pins provide  
the gate drive and return for the low-side MOSFET. These  
lines should run as parallel pairs to each MOSFET, being  
connected as close as possible to the respective MOSFET  
gate and source. Although it may be difficult in a compact de-  
sign, these lines should stay away from the output inductor if  
possible, to avoid stray coupling.  
23  
www.national.com  
Typical Application  
30090501  
FIGURE 24. Typical Application Schematic  
www.national.com  
24  
Physical Dimensions inches (millimeters) unless otherwise noted  
32-Lead LLP Package  
NS Package Number SQA32A  
25  
www.national.com  
Notes  
For more National Semiconductor product information and proven design tools, visit the following Web sites at:  
Products  
www.national.com/amplifiers  
Design Support  
www.national.com/webench  
Amplifiers  
WEBENCH® Tools  
App Notes  
Audio  
www.national.com/audio  
www.national.com/timing  
www.national.com/adc  
www.national.com/interface  
www.national.com/lvds  
www.national.com/power  
www.national.com/appnotes  
www.national.com/refdesigns  
www.national.com/samples  
www.national.com/evalboards  
www.national.com/packaging  
www.national.com/quality/green  
www.national.com/contacts  
www.national.com/quality  
www.national.com/feedback  
www.national.com/easy  
Clock and Timing  
Data Converters  
Interface  
Reference Designs  
Samples  
Eval Boards  
LVDS  
Packaging  
Power Management  
Green Compliance  
Distributors  
Switching Regulators www.national.com/switchers  
LDOs  
www.national.com/ldo  
www.national.com/led  
www.national.com/vref  
www.national.com/powerwise  
Quality and Reliability  
Feedback/Support  
Design Made Easy  
Solutions  
LED Lighting  
Voltage Reference  
PowerWise® Solutions  
www.national.com/solutions  
www.national.com/milaero  
www.national.com/solarmagic  
www.national.com/training  
Serial Digital Interface (SDI) www.national.com/sdi  
Mil/Aero  
Temperature Sensors  
Wireless (PLL/VCO)  
www.national.com/tempsensors SolarMagic™  
www.national.com/wireless  
PowerWise® Design  
University  
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© 2009 National Semiconductor Corporation  
For the most current product information visit us at www.national.com  
National Semiconductor  
Americas Technical  
Support Center  
National Semiconductor Europe  
Technical Support Center  
Email: europe.support@nsc.com  
National Semiconductor Asia  
Pacific Technical Support Center  
Email: ap.support@nsc.com  
National Semiconductor Japan  
Technical Support Center  
Email: jpn.feedback@nsc.com  
Email: support@nsc.com  
Tel: 1-800-272-9959  
www.national.com  
厂商 型号 描述 页数 下载

CREE

LM3-EWH1-01-N2-00001 [ Visible LED ] 6 页

CREE

LM3-EWN1-01-N2 SMD LED[ SMD LED ] 6 页

CREE

LM3-FWH1-01-N5-00001 [ Visible LED ] 6 页

CREE

LM3-PBG1-01-N1 SMD LED[ SMD LED ] 5 页

CREE

LM3-PBL1-01-N1 SMD LED[ SMD LED ] 5 页

CREE

LM3-PPG1-01-N1 SMD LED[ SMD LED ] 5 页

MORNSUN

LM30-00J0512-03E [ 30W, AC/DC converter ] 5 页

MORNSUN

LM30-00J0512-03E_15 [ 30W, AC/DC converter ] 5 页

INTERSIL

LM300 N沟道JFET[ N-CHANNEL JFET ] 122 页

NSC

LM3000 双同步仿电流模式控制器[ Dual Synchronous Emulated Current-Mode Controller ] 26 页

PDF索引:

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

Q

R

S

T

U

V

W

X

Y

Z

0

1

2

3

4

5

6

7

8

9

IC型号索引:

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

Q

R

S

T

U

V

W

X

Y

Z

0

1

2

3

4

5

6

7

8

9

Copyright 2024 gkzhan.com Al Rights Reserved 京ICP备06008810号-21 京

0.246728s