Design guide for LED Lighting Controller XC9401(6/9)

6. Selection of the external components of a non-isolated circuit

Selection of the external components of a non-isolated circuit is explained below using the non-isolated circuit shown in Fig. 23 as an example. This circuit uses the XC9401 series B type at 100VAC.

Fig.23 100VAC Non-isolated Buck, B type Typical Application Circuit

6-1. Number of LED Series

First, the criteria for selecting the number of LED series in this application is described.

The LED connection method, number of LED series, and LED current play an important role in efficient LED illumination. The general relation between the number of LED series and the LED current at a fixed LED output power is shown in Fig. 24.

It can be seen that increasing the number of LED series reduces LED current. When LED current decreases in a non-isolated circuit, loss in peripheral components of the power circuit decreases, efficiency improves, and smaller components can be used. This makes it possible to reduce mounting area and cost. It is actually possible to hold down the total cost of LEDs and peripheral components by selecting an optimum value for the number of LED series.

Fig.24 The general relation between the number of LED series and the LED current at a fixed LED output power

When the input voltage is large and the LED voltage is small in a non-isolated circuit, the on time may in some cases become shorter than the minimum on time tONMIN. When the on time is shorter than the minimum on time, control of the LED current is not possible and the LED current becomes higher than the set value.
For this reason, select a LED voltage that satisfies equation (6) to keep the on time from becoming shorter than the minimum on time.

tONMIN Minimum on time
VLED LED voltage
VF Forward voltage of rectification diode
Vrms_max Maximum input RMS voltage
tOFF Off time 6.0μs(typ.)

In this example, external components will be selected based on 20 LED series and a LED current of 110mA.

6-2. Bridge Diode (BR)

This is a bridge diode for full wave rectification of the AC input. Select a bridge diode with a peak inverse voltage and average rectification current that are more than sufficient for the input voltage and current.

In this example, the peak value of the input current is about 500mA and the maximum voltage applied to the bridge diode is about 282V, and thus a product with a rated current of 0.8A and a rated voltage of 400V is selected.

6-3. Input Filter (L1,C1,C2)

C1 and L1 form a filter circuit that reduces noise from the AC input and noise that returns to the AC input. In the typical circuit example (Fig. 23), a filter is formed that attenuates 20kHz and higher noise to remove switching frequency (50kHz to 150kHz) and higher noise. The capacitance value of C1 must be kept small to limit rush current from the AC input, so select a capacitor that is about 0.1μF.

It will be necessary to adjust the input filter constants and filter circuit to meet the regulations and standards that will actually apply.

The voltage after full wave rectification is smoothed by C2. LED flickering is reduced by using a higher capacitance for C2. When the smoothed voltage Vrec after full wave rectification drops lower than the LED voltage, switching stops and the LED current falls (Fig. 25). The longer switching stops, the more the LED current falls, and when it falls below 5% of its peak value, flickering occurs. (The PSE definition is used for the definition of flickering.)

To prevent flickering, the LED voltage and C2 capacitance value must be selected to satisfy Equation (7). Note, however, that the power factor decreases as the capacitance value is increased.

Fig.25 Various waveforms in flickering

PIN Input Power
f Utility frequency 50Hz / 60Hz
Vrms_min Minimum input RMS voltage

An example calculation is given below.
VLED =60V, ILED=0.11A, f=50Hz, Vrms_min=90V, the minimum value of the C2 capacitance is

and flickering can be prevented by using a capacitance of 7.15μF or higher.

The result of the above calculation is an ideal value.
The actual capacitance value to be used can be calculated from the separate calculation file.

6-4. Power Supply to VDD pin (R5,R6,C3,ZD1)

This circuit supplies power to the power pin (VDD pin) of the IC. There are two power supply methods: a method that uses a Zener diode and a method that uses a transformer auxiliary coil. The method that uses a transformer supplies power to the VDD pin through an auxiliary coil. This reduces loss in RVDD and enables a higher efficiency than the Zener diode method to be obtained.

This example uses the Zener diode method, but the transformer auxiliary coil method is also explained. Selection of components for each method is described below.

Method using a Zener diode

A VDD power supply circuit using a Zener diode is shown in Fig. 26.

Fig.26 VDD power supply circuit using a Zener diode

ZD1

This is a Zener diode that determines the voltage applied to the VDD pin.
Use a Zener diode that satisfies
VDD minimum voltage (9V) < Zener voltage < VDD maximum voltage (15V)
In this example, a product with a Zener voltage of 12V has been selected.

CVDD

This capacitor stabilizes the VDD pin voltage. Use a capacitor with a capacitance of 10μF or higher.
If a ceramic capacitor will be used, select a product in which the electrostatic capacitance falls minimally when a B type (JIS Standards) or X7R/X5R (EIA Standards) DC bias is applied.

RVDD

This resistance determines the current to the VDD pin and ZD1 from the smoothed voltage after full wave rectification. The current that flows through RVDD is the steady IC supply current plus the current for charging the external power MOSFET gate for switching. Setting too high a value for this resistance lowers the VDD pin voltage and may cause unstable operation. Setting too low a value increases the loss in RVDD and reduces efficiency. Therefore, it is important to set an appropriate value.

In this example, the total value of the IC supply current and the current for charging the external power MOSFET gate is assumed to be 1 mA, and 66kΩ is selected for RVDD (the total value of R5 and R6 in Fig. 23).

The optimum resistance value depends on the input voltage, gate capacitance of the external power MOSFET, coil inductance value, and other parameters. To calculate the actual resistance value to be used, refer to the separate calculation file.

Method using a transformer

A VDD power supply circuit using a transformer is shown in Fig. 27.

Fig.27 VDD power supply circuit using a transformer

LT1

Current is supplied to the VDD pin using the LT1 auxiliary coil.
For selection of the transformer, refer to section 6-6.

DVDD

This is a rectifying diode that supplies power voltage from LT1. A reverse bias voltage VDvdd that depends on the LED voltage and transformer turn ratio as shown in Equation (8) is applied to DVDD. Select a diode with a rated voltage appropriate for this reverse bias voltage.

N1 Number of windings of transformer primary coil
NAUX Number of windings of transformer auxiliary coil
VDD VDD pin voltage
Vrms_max Maximum input RMS voltage
Vspike Spike voltage that accompanies switching (to 50 V)

A calculation example is shown below.
N1=150, NAUX=30, VDD=12V, Vrms_max=110V, Vspike=50V, the reverse bias voltage VDvdd is

The same calculation is made in the separate calculation file. Please make use of this file.

CVDD

This capacitor stabilizes the VDD pin voltage. Use a capacitor with a capacitance of 10μF or higher. If a ceramic capacitor will be used, select a product in which the electrostatic capacitance falls minimally when a B type (JIS Standards) or X7R/X5R (EIA Standards) DC bias is applied.

RVDD

This resistance is used to supply current to the VDD pin at startup. When the input voltage is applied and the VDD pin voltage rises above the UVLO release voltage, GATE output starts and normal operation takes place. After startup, power is mainly supplied to the VDD pin through the auxiliary coil of the transformer.

When RVDD is large and the current through RVDD is smaller than the current consumed in the IC, the VDD pin voltage does not rise higher than the UVLO release voltage and startup is not possible. For this reason, select a resistance value for RVDD that satisfies Equation (9). (Fig. 27)

ISTB Stand-by Current 225μA (typ.)
VUVLOR UVLO Release Voltage 7.5V (typ.)
Vrms_min Minimum input RMS voltage

A calculation example is shown below.
ISTB=225µA, VUVLOR=7.5V, Vrms_min=90V, RVDD is

and the IC can be started normally by using a resistance lower than 532kΩ.
The same calculation is made in the separate calculation file. Please make use of this.

RVDD1

To supply current to the VDD pin, LX_VDD is made to oscillate and supply voltage to the VDD pin (refer to Fig. 28). However, in actuality a spike voltage sometimes occurs in LX_VDD and causes the VDD pin voltage to rise higher than the VDD target voltage (= VLED × N3/N2). A countermeasure for this VDD pin voltage rise is to insert a resistance in RVDD1 to reduce the current supplied to the VDD pin.

Fig.28 Operation waveforms (VDD power supply circuit using a transformer)

6-5. Coil (L2)

In the XC9401 series, the external power MOSFET off time is fixed at 6.0μs (typ.) and the peak current of the coil is controlled. For this reason, the operation mode, continuous mode or discontinuous mode, is determined by the smoothed voltage after full wave rectification and the inductance value of the coil.

In control continuous mode, which has a fixed off time, the LED current ideally does not fluctuate due to fluctuations in the input voltage. However, in discontinuous mode, the LED current fluctuates with fluctuations in the input voltage. For this reason, select a coil with an inductance value suitable for operation in continuous mode. The detailed method is described below.

First, calculate from Equation (10) the minimum inductance value required to enter continuous mode. In continuous mode, deviations in the LED current due to deviations in the inductance are smaller when the inductance value is larger, so choose an inductance value that is as large as possible. Using a product with good inductance accuracy can also reduce LED current fluctuation.

If the inductance value is too large, the switching frequency may enter the audible range (20 to 20kHz), so make sure the inductance satisfies the equations below to prevent entry into the audible range.

Once you have selected an inductance value, select a coil taking peak coil current and heat generation into consideration.

VLED LED voltage
VF Forward voltage of rectification diode
ILED LED current
tOFF Off time 6.0μs(typ.)
L Coil inductance value
ΔIL Coil current amplitude
Vrec_min_ave Average value of voltage smoothed after full wave rectification at minimum input voltage
(The calculation is complex, so please check the calculation file.)

A calculation example is shown below.
When VLED=60V, VF=1.0V, ILED=0.15A, tOFF=6.0µs the minimum value of the inductance is

and an indactance of 1.66 mH or higher should be selected. Because it is desired to minimize deviations in the LED current, a 3.3 mH coil is selected here.
Next, we check if the switching frequency is within the audible range when the selected inductance is used.
When L=3.3mH, Vrec_min_ave=120V, ΔIL=0.11A, therefore equation (11) is

and it can be seen that the switching frequency is not within the audible range.
To select the coil that will actually be mounted, refer to the separate calculation file.

6-6. Flywheel Diode (D1)

Flywheel diode for discharge of the energy that is stored in the inductance when MOSFET Q1 is in the off state. Use a flywheel diode with a short reverse recovery time. A diode with a long reverse recovery time will adversely affect efficiency.
Because the peak current reaches 180mA, a product with a rated current of 0.7A is selected in this example.

6-7. MOSFET, Gate Resistor (R7)

Power MOSFET for switching and gate resistor for switching time adjustment.
Inserting a gate resistance makes it possible to slow the MOSFET switching time and reduce the high-frequency EMI level. However, a larger gate resistance and a slower switching speed increases MOSFET switching loss, resulting in lower efficiency. The optimum value depends on the MOSFET that is used, but in general a gate resistance of about 5 to 50Ω should be selected.

The MOSFET selection method varies depending on the VDD power supply method. The selection methods are explained below.

Power Supply to VDD pin:Method using a Zener diode

When a MOSFET with a large gate capacitance is selected, the current for gate charging supplied to the VDD pin is larger, resulting in increased loss in R5 and R6 and decreased efficiency. A larger loss in R5 and R6 means that resistors with a larger allowable loss must be selected, which increases the mounting area and results in higher cost.
For this reason, it is important to select a MOSFET with a small gate capacitance and increase the efficiency of the overall circuit. In this example, the IPD60R3K3C6 (Gate charge total: 4.6nC @10V) is selected as a MOSFET with a small gate capacitance.

Power Supply to VDD pin:Method using a transformer

Unlike the Zener diode method, power is supplied with high efficiency through the transformer to the VDD pin when the transformer method is used, and thus using a MOSFET with a small on-resistance to reduce MOSFET loss even when the gate capacitance is large results in high efficiency.
For this reason, select a MOSFET with a small on-resistance.

6-8. LED current adjustment (R3,R4)

Sensing resistor that adjusts the external power MOSFET current in order to adjust the LED current. The LED current is set by adjusting the sensing resistance.
In the B type used in this example, the ISEN voltage is compared to the internal reference voltage, and the peak value of the MOSFET current is determined by the sensing resistances R3 and R4 as given in Equation (12). (Refer to Fig. 29.)

Ip Peak value of MOSFET current (same as peak value of coil current described above)
VISEN ISEN Voltage 0.343V (typ.)

The MOSFET current, coil current, and LED current in continuous mode in a non-isolated circuit are shown in Fig. 30. The LED current is the average value of the coil current, and thus by using the resistance values calculated in Equation (13) for the sensing resistors R3 and R4, the LED current can be adjusted to the target value.

VISEN ISEN Voltage 0.343V (typ.)
ILED Target value of LED current
VLED LED voltage
VF Forward voltage of flywheel diode
L Coil inductance value
tOFF Off time 6.0μs(typ.)

Fig.29 MOSFET Current and ISEN Voltage

Fig.30 MOSFET Current, Coil , LED Current

A calculation example is shown below.
When VISEN=0.3400V , ILED=0.15A, VLED=60V, VF=1.0V, L=3.3mH, tOFF=6.0µs, the LED current can be set to 0.11A by using resistance values for sensing resistors R3 and R4 that satisfy.

The actual resistance values that are used must be calculated by a formula that includes parameters such as the circuit delay, so calculate these using the separate calculation file.

6-9. Output Capacitor (C4)

Capacitor that limits LED ripple current and ripple voltage.
As in this example, if the smoothed voltage after full wave rectification Vrec never falls below the LED voltage, flickering does not occur and a smaller capacitance value can be used for the output capacitance C4. For this reason, a ceramic capacitor can be used for the output capacitance instead of an electrolytic capacitor, and this enables improvement of the reliability of the LED lighting.

The capacitance value of the output capacitance is determined by the ripple current ratio of the LED current. If the ripple current ratio is to be kept under 0.8 (ripple current: 110mA ×0.8 = 88mA) for ILED = 110mA, first calculate the allowed Vripple from the LED IV characteristic that is used. Here this is 0.35V×20=7.0V from Fig. 31.

If a ceramic capacitor is used, select a capacitor for the output capacitance with a larger capacitance value than that given in Equation (14) to attain Vripple = 7.0V. A DC bias, temperature changes, and other conditions will cause the capacitance of a ceramic capacitor to drop lower than the nominal value, so select a product whose effective capacitance satisfies Equation (14), taking into consideration conditions such as the DC bias and temperature changes.

Fig.31 LED IV characteristic

C Minimum effective capacitance value of output capacitance C4
Vripple Ripple voltage allowed in LED voltage
tON On time
tOFF Off time 6.0μs(typ.)
ΔIL Coil current amplitude

A calculation example is shown below.
When Vripple=9.0V, tON=6.05µs, tOFF=6.0µs, ΔIL=0.11A, the minimum effective capacitance value of the output capacitor C4 is

By selecting a capacitance of 0.024μF or higher for the effective capacitance during operation, the ripple current ratio can be held to 0.8 or less.

The same calculation is made in the separate calculation file. Please make use of this.
The actual capacitor ESR effects and LED IV characteristics are non-linear, and thus the value may vary in the actual equipment. Test in the actual equipment before selecting the capacitance value.

6-10. Line regulation improvement circuit

In the XC9401 series, the LED current may sometimes fluctuate due to input voltage fluctuations caused by delay times inside the circuit and other factors. If input voltage fluctuations of the LED current are observed, line regulation can be improved using the circuit shown in Fig. 32.

For the resistors, select resistance values such that the voltage applied to both ends of RL2 is 0.1V or less. Using this circuit as a countermeasure causes a lower LED current than normal. For this reason, lower resistance values must be used for the sensing resistances than those calculated in Equation (13).

The effectiveness of the improvement varies depending on the input voltage, coil inductance value, and sensing resistance value, so check using the calculation sheet and in the actual equipment.

Fig.32 Line regulation improvement circuit

Torex Products in this article

Close