10.3.1 Design

When designing for thermal performance, many variables must be considered, such as ambient temperature, airflow, external components, and PCB design.

The surrounding maximum air temperature is fairly explanatory. As the temperature increases, the junction temperature increases. This may not be linear though. As the surrounding air temperature increases, resistances of semiconductors, wires and traces increase. This decreases the efficiency of the application, and more power is converted into heat, and increases the silicon junction temperatures further.

Forced air can drastically reduce the device junction temperature. Air flow reduces the hot spots within a design. Warm airflow is often much better than a lower ambient temperature with no airflow.

Choose components that are efficient, and the mutual heating between devices can be reduced.

The PCB design is a very important step in the thermal design procedure. The LM3410 and LM3410-Q1 are available in three package options (6-pin WSON, 8-pin MSOP, and 5-pin SOT-23). The options are electrically the same, but there are differences between the package sizes and thermal performances. The WSON and MSOP have thermal die attach pads (DAP) attached to the bottom of the packages, and are therefore capable of dissipating more heat than the SOT-23 package. It is important that the customer choose the correct package for the application. A detailed thermal design procedure has been included in this data sheet. This procedure helps determine which package is correct, and common applications are analyzed.

There is one significant thermal PCB layout design consideration that contradicts a proper electrical PCB layout design consideration. This contradiction is the placement of external components that dissipate heat. The greatest external heat contributor is the external Schottky diode. Increasing the distance between the LM3410 or LM3410-Q1 and the Schottky diode may reduce the mutual heating effect. This, however, creates electrical performance issues. It is important to keep the device, the output capacitor, and Schottky diode physically close to each other (see Layout Guidelines). The electrical design considerations outweigh the thermal considerations. Other factors that influence thermal performance are thermal vias, copper weight, and number of board layers.

Heat energy is transferred from regions of high temperature to regions of low temperature via three basic mechanisms: radiation, conduction and convection. Conduction and convection are the dominant heat transfer mechanism in most applications.

The data sheet values for each packages thermal impedances are given to allow comparison of the thermal performance of one package against another. To achieve a comparison between packages, all other variables must be held constant in the comparison (PCB size, copper weight, thermal vias, power dissipation, V_IN, V_OUT, load current, and others). This provides indication of package performance, but it would be a mistake to use these values to calculate the actual junction temperature in an application.

10.3.2 LM3410 and LM3410-Q1 Thermal Models

Heat is dissipated from the LM3410, LM3410-Q1, and other devices. The external loss elements include the Schottky diode, inductor, and loads. All loss elements mutually increase the heat on the PCB, and therefore increase each other’s temperatures.

Figure 35. Thermal Schematic

Figure 36. Associated Thermal Model

10.3.3 Calculating Efficiency and Junction Temperature

Use Equation 31 to calculate R_θJA.

Equation 31.

A common error when calculating R_θJA is to assume that the package is the only variable to consider.

Other variables are:

• Input voltage, output voltage, output current, R_DS(ON)
• Ambient temperature and air flow
• Internal and external components' power dissipation
• Package thermal limitations
• PCB variables (copper weight, thermal vias, and component placement)

Another common error when calculating junction temperature is to assume that the top case temperature is the proper temperature when calculating R_θJC. R_θJC represents the thermal impedance of all six sides of a package, not just the top side. This document refers to a thermal impedance called R_ΨJC. R_ΨJC represents a thermal impedance associated with just the top case temperature. This allows for the calculation of the junction temperature with a thermal sensor connected to the top case.

The complete LM3410 and LM3410-Q1 boost converter efficiency can be calculated using Equation 32.

Equation 32.

where

• P_LOSS is the sum of two types of losses in the converter, switching and conduction

Conduction losses usually dominate at higher output loads, where as switching losses remain relatively fixed and dominate at lower output loads.

To calculate losses in the LM3410 or LM3410-Q1 device, use Equation 33.

Conversion ratio of the boost converter with conduction loss elements inserted is calculated with Equation 34.

Equation 34.

where

• R_DCR is the Inductor series resistance

Equation 35.

If the loss elements are reduced to zero, the conversion ratio simplifies to Equation 36.

Equation 36.

Equation 37.

Therefore:

Equation 38.

Only calculations for determining the most significant power losses are discussed. Other losses totaling less than 2% are not discussed.

A simple efficiency calculation that takes into account the conduction losses is Equation 39.

Equation 39.

The diode, NMOS switch, and inductor (DCR) losses are included in this calculation. Setting any loss element to zero simplifies the equation.

V_D is the forward voltage drop across the Schottky diode. It can be obtained from Electrical Characteristics.

Conduction losses in the diode are calculated with Equation 40.

Depending on the duty cycle, this can be the single most significant power loss in the circuit. Choose a diode that has a low forward voltage drop. Another concern with diode selection is reverse leakage current. Depending on the ambient temperature and the reverse voltage across the diode, the current being drawn from the output to the NMOS switch during time (D) could be significant, this may increase losses internal to the LM3410 or LM3410-Q1 and reduce the overall efficiency of the application. See the Schottky diode manufacturer’s data sheets for reverse leakage specifications.

Another significant external power loss is the conduction loss in the input inductor. The power loss within the inductor can be simplified to Equation 41,

Or Equation 42.

Equation 42.

The LM3410 and LM3410-Q1 conduction loss is mainly associated with the internal power switch.

Figure 37. LM3410 and LM3410-Q1 Switch Current

Equation 44.

(small ripple approximation)

Or

Equation 46.

The value for R_DS(ON) must be equal to the resistance at the desired junction temperature for analyzation. As an example, at 125°C and R_DS(ON) = 250 mΩ (See Typical Characteristics for value).

Switching losses are also associated with the internal power switch. They occur during the switch ON and OFF transition periods, where voltages and currents overlap resulting in power loss.

The simplest means to determine this loss is empirically measuring the rise and fall times (10% to 90%) of the switch at the switch node.

10.3.4 Example Efficiency Calculation

Operating Conditions:5 × 3.3-V LEDs + 190 mV_REF ≊ 16.7 V

Quiescent Power Loss:

Switching Power Loss:

Internal NFET Power Loss:

Diode Loss:

Inductor Power Loss:

10.3.5 Calculating R_θJA and R_ΨJC

Equation 65.

We now know the internal power dissipation, and we are trying to keep the junction temperature at or below 125°C. The next step is to calculate the value for R_θJA or R_ΨJC. This is actually very simple to accomplish, and necessary for determining the correct package option for a given application.

The LM3410 and LM3410-Q1 have a thermal shutdown comparator. When the silicon reaches a temperature of 165°C, the device shuts down until the temperature drops to 150°C. From this, it is possible calculate the R_θJA or the R_ΨJC of a specific application. Because the junction to top case thermal impedance is much lower than the thermal impedance of junction to ambient air, the error in calculating R_ΨJC is lower than for R_θJA . However, a small thermocouple needs to be attached onto the top case of the device to obtain the R_ΨJC value.

Knowing the temperature of the silicon when the device shuts down provides three of the four variables. After calculating the thermal impedance, working backwards with the junction temperature set to 125°C, the maximum ambient air temperature to keep the silicon below 125°C can be calculated.

Procedure:

Place the application into a thermal chamber. Dissipate enough power in the device to obtain an accurate thermal impedance value.

Raise the ambient air temperature until the device goes into thermal shutdown. Record the temperatures of the ambient air and the top case temperature of the device. Calculate the thermal impedances.

Example from previous calculations (SOT-23 Package):

Equation 69.

Typical WSON and MSOP typical applications produces R_θJA numbers from 53.7°C/W to 55.3°C/W, and R_θJC varies from 61.4°C/W to 65.9°C/W. These values are for PCBs with two and four layer boards with 0.5 oz copper, and four to six thermal vias to bottom side ground plane under the DAP. The thermal impedances calculated above are higher due to the small amount of power being dissipated within the device.

Figure 38. R_θJA vs Internal Dissipation

For 5-pin SOT-23 package typical applications, R_θJA numbers range from 164.2°C/W, and R_θJC varies from 115.3°C/W. These values are for PCBs with two and four layer boards with 0.5 oz copper, with two to four thermal vias from GND pin to bottom layer.

Using typical thermal impedances and an ambient temperature maximum of 75°C, if the design requires more dissipation than 400 mW internal to the device, or there is 750 mW of total power loss in the application, TI recommends using the 6-pin WSON or the 8-pin MSOP-PowerPad package with the exposed DAP.