MOSFET power consumption calculation and thermal design details of the key points
1. Correctly understand the thermal resistance values in the MOSFET data sheet
MOSFET thermal resistance values Rth(j-a) and Rth(j-mb) are usually listed in the data sheet
Rth(j-a): Refers to the thermal resistance of the device junction (die) to the surrounding environment. It can be understood that it is an inherent property of the MOSFET component itself, which cannot be improved by external measures;
Rth(j-mb): Refers to the thermal resistance from the device junction to the solder substrate. A solder substrate is generally defined as a point soldered to a PCB and is the single primary path of heat conduction.
However, it should be noted that the values given in the table are subject to test conditions, and if they are not the same test conditions, the thermal resistance values will be different. As clearly mentioned in the note below the table, soldered to an FR4 type PCB with only one layer of copper foil, the surface of the copper foil is tinned, and it is in a standard pad package.
However, in the actual PCB layout, basically there is not only a layer of copper foil, but also a PCB made of OSP material without tin plating, so the data in the data sheet must not be directly applied to the temperature calculation of the actual product, but according to the actual circuit consumption and PCB layout through simulation or measurement to obtain real and credible temperature Tj data.
2. Correctly design the area of copper foil for heat dissipation, and understand the relationship between the area of copper foil and the component Tj through the simulation model
It is not that the larger the area of the copper foil, the better the heat dissipation effect, but the relationship between the area of the copper foil and the element Tj is seen through the following simulation model.
The simulation model below shows a MOSFET device soldered to a 40 x 40 mm FR 4 PCB, with a square x mm side of copper foil directly in contact under the component, and an ambient temperature of 20°C.
By increasing the edge length of the copper foil of the pad, the following curve is drawn by continuously simulating Tj. It can be seen that:
The junction temperature Tj is highly dependent on the edge length x, or the area of a single layer of copper foil.
However, with the increase of the copper foil area, the decline of Tj will slow down, and after increasing to a certain area, Tj will no longer be affected by the copper foil area. This also demonstrates the principle of the "law of diminishing effect".
So it's not that the larger the copper foil area of the pad, the better the heat dissipation of the component.
3. Set up a positive relationship between the copper foil without electrical connection and the heat dissipation directly below the element
A layer of PCB is shown above, and after increasing the heat dissipation copper foil area to a certain size, the heat dissipation effect will no longer be obvious. Then we can turn the PCB into a two-layer board and add the corresponding copper foil directly below the component to help dissipate heat, even if there are no vias for electrical connections.
Based on the simulation model in the second part, a copper foil with an area of 25 mm x 25 mm was installed directly below the component, and by changing the area of the copper foil on the top pad, it was found that the copper foil was helpful for heat dissipation even if it did not have electrical connectivity. The reason can be understood as the conduction of heat between the upper and lower layers by means of thermal radiation.
From the above simulation results diagram, when the density of the top layer element is high, you can consider reducing the top layer copper foil area from 25 x 25 mm to about 15 x 15mm, and then adding 25mm x 25mm copper foil to the bottom layer, so that the same thermal performance can be guaranteed (Tj is about 56°C).
4. The design of PCB heat dissipation holes
4.1 What are PCB heat dissipation holes Heat dissipation holes are used to conduct heat to the back side by using a channel (via) through the PCB board to dissipate heat, and are placed directly below or as close to the heating element as possible. Heat dissipation hole is a method of using PCB board to improve the heat dissipation effect of surface mount parts, and structurally through holes are set on the PCB board. In the case of a single-layer, double-sided PCB, the copper foil on the surface of the PCB board and the back of the PCB are connected to increase the area and volume used for heat dissipation, that is, to reduce the thermal resistance. In the case of multilayer PCBs, it is possible to connect the faces between the layers or the layers that are partially connected by the limit, and the purpose is the same.
4.2 How to set up PCB heat vents The placement and size of PCB heat vents vary widely, depending on the type of component, different rules, and expertise.
But one main rule is to use heat vents, as close as possible to the heat source directly below the heating element. Then, in cases where the heat source is not ideal, the hot vias can also be placed on the periphery of the component, regardless of the component pad placement. In this case, the rule also remains the same, i.e., place the heat vents as close to the periphery of the group as possible. In order to use the heat dissipation holes effectively, it is important to place the heat dissipation holes close to the heating element, such as directly below the component. As you can see in the figure below, it is a good way to connect locations with large temperature differences by using the heat balance effect.
Example of PCB Heat Sink Placement Example: Configuration of Heat Sink HolesThe following is an example of the layout and size of the heat vents of the HTSOP-J8 in the exposed type package with the rear heat sink.
4.3 What is the general size of the PCB heat dissipation hole In order to improve the thermal conductivity of the heat dissipation hole, it is recommended to use a small hole with an inner diameter of about 0.3mm that can be filled by electroplating. It is important to note that if the pore size is too large, solder creeping may occur during the reflow process. The heat dissipation holes are spaced about 1.2 mm apart and are placed directly below the heat sink on the back of the package. If just below the back heatsink is not enough to dissipate heat, heat dissipation holes can also be placed around the IC. In this case, the key point is to place the configuration as close to the IC as possible.
4.4 Thermal conductivity of different materials Thermal conductivity is a key factor that determines how much heat a material can absorb. The table below gives an idea of the thermal conductivity of different materials. With the help of this table, you can have a certain reference See the table below:
Therefore, as can be seen from the above table, aluminum has a worse thermal conductivity than copper. However, due to the larger area of the aluminum radiator, it produces a more efficient cooling effect on the heated equipment. However, as we can see, if copper is used effectively, it can emit more heat than aluminum in the same area. Effective hot via placement is when the vias are used appropriately in an IC or a heating element pad using conduction, as a heat transfer method where heat is distributed between multiple layers of copper and then passed through free air, heat dissipation begins to be transferred in the air using the convection method. It is recommended that the inner diameter of the vire need to be smaller, e.g. about 0.35 mm. If the pore size is large, incorrect soldering may occur during the reflow soldering process, so extra care is required. However, if a larger diameter is required, hot fill may help compensate for this.
4.5 Balancing the number, cost, and performance of heat dissipation holes through thermal simulation Generally speaking, adding vias under the device can improve thermal performance, but it is difficult to know how many vias are needed to optimize the solution. When adding vias, consider both EMC and PCB cost to achieve a balance with heat dissipation. EMC: Considering the path of EMC outward radiation and signal crosstalk, it is generally necessary to have a complete ground plane in the PCB for shielding, but the increase of vias will inevitably destroy the integrity of the ground plane. PCB cost: Every vias are drilled by a drill bit, so it will increase the cost of PCB fabrication.
The above simulation results plot shows a significant decrease in the Tj of the device from no vias underneath the device to 20 vias underneath the device. This is a clear indication that the heat is conducted from the MOSFET heatsink through the via to the fourth layer, and the results are exactly as expected. However, it can also be found that although the number of vias under the device is gradually increased (from 20 to 77), it does not lead to additional cooling of Tj. This is because we added more vias, which increased the heat transfer between the layers of the PCB, but also reduced the area of the first layer of PCB copper foil that can be exposed to air to contact with the device. So we didn't see a significant improvement in thermal performance. Therefore, the conclusion is that the heat dissipation performance can be improved by adding vias, but if too many vias are added, the heat dissipation performance will not be significantly improved.
4.6 Other precautions for PCB heat dissipation hole design In the process of hot vivia design, there are few matters that need to be paid attention to, and there are the following 6 suggestions:
4.6.1. The design of the exposed pad is to transfer heat directly from the shell to the copper area. The effect of solder as a heat sink is not noticeable because it is thin and the conductivity of the solder is poor.
Hot Through-Through Via on U1 Exposed Pad The image above shows a hot through via on U1 exposed pad.
4.6.2, For exposed pad packages, the greatest heat dissipation occurs through the vias to the bottom layer of the PCB and then dissipated into the air. As a result, the large underlying area will also reduce heat dissipation in the component package.
4.6.3. Separate the heated components and use the heat dissipation holes to dissipate heat, which helps to evenly distribute the heat to other packages.
4.6.4, Thermal vias are the only source of heat dissipation on DFN and QFN packages because the top layer of copper does not have the maximum space due to pin allocation. Therefore, to use the underlying copper, the only way to increase the thermal conductivity is to use heat dissipation holes.
U5 and IC2 using heat vents. IC2 uses a QFN flat package, where hot vias are the only possible because this does not include a larger copper area on the solder layer due to the distribution of the component pads.
4.6.5. The effective copper area of the RTVIA connection equipment will be the maximum copper length (independent of the solder layer) that is directly connected to the component package using the RTA.
4.6.6. The thickness of the copper plane also affects the thermal conductivity, and 2Oz copper has better heat resistance than 1.0 Oz or 0.5Oz copper.
5. Analysis of the influence of temperature on MOSFET components
The MOSFET bonding point temperature Tj is composed of the temperature generated by the self-heating and the ambient temperature. While trying to better dissipate heat, we should also pay attention to the temperature outside the components, such as: The ambient temperature where the product is located: generally according to customer needs, it cannot be changed. Whether there are other high-power, high-heat devices around the component: This will increase the temperature around the component by means of thermal radiation, resulting in failures that cannot be considered in theoretical design calculations. Increase air convection heat dissipation mode: Whether the product can add a fan to speed up the air flow and improve the heat dissipation performance. The heat dissipation of MOSFETs is not only by soldering the substrate, but also by the pins, so increasing the copper foil area of the pin pads also has a positive effect. In extreme cases, you can also consider adding a heat sink to increase the heat dissipation area.
6. Select advanced packaging process products such as top heat dissipation, double-sided heat dissipation, and integrated sealing (take Chongqing Pingwei TCOP10 top heat dissipation as an example)
Since the lead frame of the device (including the exposed drain pad) is directly soldered to the copper clad area, this causes heat to be propagated mainly through the PCB. The rest of the device is enclosed in a molding compound that dissipates heat only through air convection. Therefore, the heat transfer efficiency is highly dependent on the characteristics of the board: the area size, the number of layers, the thickness, and the layout of the copper pour. This can happen regardless of whether the board is mounted to a heat sink or not. Often, the maximum power capability of a device is not optimal because the PCB generally does not have high thermal conductivity and thermal mass. To solve this problem and further reduce the size of the application, the world's power device giants are aggressively developing a new MOSFET package that exposes the MOSFET's lead frame (drain) at the top of the package (as shown in the figure below).
Why is there a top-side thermal package design? For a considerable period of time in the development of the semiconductor industry, whether it is power semiconductors, analog semiconductors, or digital semiconductors, the size of chips is constantly shrinking, and the process is constantly shrinking. Specific to power semiconductors, the evolution of chips in the past decade or so has been mainly in the wafer part, such as using smaller chip sizes, achieving lower on-resistance, etc., and over time, experts have gradually found that packaging technology has become a key way to break through the bottleneck. The FOM value of silicon power devices has basically reached the physical limit, and in this case, if you want to continue to reduce the on-impedance or achieve higher energy efficiency, packaging technology is the only way to continue to maximize the power of silicon. Not only silicon-based semiconductors, but also the popular wide bandgap semiconductor SiC/GaN also need to rely on the new packaging technology, and the benefits that TCOP10 top-side thermal packaging brings to customers.
The biggest benefit of TCOP10's top-side thermal packaging technology is that it highly optimizes the production process, reduces the number of steps in the entire assembly process, and makes the automated manufacturing process simpler, and finally achieves a reduction in the number of PCBs, layers, and inter-board connectors at the downstream vendor's end, resulting in a significant reduction in assembly and overall system costs. Optimizing MOSFET applications requires minimizing the system thermal resistance (Rthja) while achieving the highest junction temperature (Tj). This maximizes the amount of heat flowing into the heatsink and minimizes the amount of heat flowing into the PCB. Readers familiar with the power semiconductor industry should still have the impression that 10 years ago, high-power applications of kilowatts and above were basically dominated by plug-in packaging (THD) technology, such as the well-known TO247 and TO220 packages. The advantage of this type of plug-in packaging technology is that under the assembly and packaging process at the time, engineers can maximize the use of the external heat sink, and the heat generated inside the chip is very efficient to dissipate outside the chip, so that the chip can work in a high-power application scenario. However, as equipment such as data centers, 5G wireless communication macro base stations, and industrial or automotive drive motors require higher power density, the size of the equipment is getting smaller and smaller. There is a need for less, or no separate heat sinks, in board designs for power or motor drive applications, while dissipating more heat evenly across the device. After a long time of discussion with the industry's leading customers and engineers in the downstream of the industrial chain, we finally reached a consensus in the industry that top heat dissipation is the fundamental way to solve this contradiction. Placement is the first step from plug-in packaging with separate heatsinks to higher power heat dissipation. Generally, the heat dissipation of SMD packaging mainly depends on the contact between the bottom of the chip and the PCB (printed circuit board), and the heat generated by the chip is dissipated by the PCB copper foil. However, one of the obvious drawbacks of this is that it requires a relatively large area of PCB copper foil to effectively dissipate heat. If it is not possible to use a large enough area of PCB copper foil during this period, then a hot spot will form at the bottom of the chip, and thisAnd this hot spot will put a lot of pressure on the PCB. At present, the commonly used PCB in the industry is FR4 material, and the maximum temperature limit of this material is about 110°C. In higher power designs, the bottom heat dissipation package cannot evenly dissipate more heat through the bond between the chip and the PCB, resulting in a bottleneck in this heat dissipation method.
The top heat dissipation only needs to use a thin heat sink at the top, instead of relying on the bottom heat sink to dissipate heat, which can dissipate heat more efficiently and evenly under the same PCB material. At the same time, the benefit it brings to users is that it can increase the external output power of the entire device under the same heat dissipation area. The main advantages of top heat dissipation are; ●Meet the demand for higher power: optimize the use of circuit board space and save PCB costs; ●Improved power density: top heat dissipation can achieve the highest board utilization; ●Improved efficiency: Optimized structure with low resistance and ultra-low parasitic inductance for higher efficiency; ●Weight reduction: Comprehensively optimize heat dissipation and heat generation, which helps to create a smaller shell, thereby reducing materials and weight;
The TCOP10 top package is most suitable for data centers, communication base stations, automobiles, and other application scenarios that require products with small size, light weight, high power density, and high efficiency. In fact, several international power manufacturers have invested in the research and development of top heat dissipation packaging technology relatively early. As friends discover the development trend of packaging technology, there will be more and more friends who will do top heat dissipation packaging technology in the future. Pingwei Industry keeps up with the trend of international power advanced technology, and has continuously invested R&D resources in the past three years to develop and mass-produce DFN copper clip process, DFN5X6 double-sided heat dissipation, TOLL/STOLL high-power packaging, to the current TCOP10 top heat dissipation, technology iteration and innovation, and constantly break through the limit, open up new routes for domestic counterparts, and fill the technical gap.
The advanced packaging of Pingwei Industrial MOSFET is still iterative and innovative, and after the top heat dissipation and double-sided heat dissipation, it is still doing innovative research on integrated sealing,
Any MOS selection, design, heat dissipation simulation, etc., welcome to leave a message or contact the regional responsible sales manager, Starrymicro provides you with super added value and serves customers wholeheartedly.
