A power supply unit is an electric device that provides power to one or more electric loads. Its main function is adapting electrical energy to the load by converting the energy from the electrical source to the corresponding current, voltage and frequency required by the load. Power supplies also provide protection to the supplied load by limiting the load current and shutting it off in case of a fault. These devices also modify the output power to prevent electronic noise and provide power-factor correction.

In this article, we'll explain the different types of power supplies and examine why fanless power supplies can provide the best performance.

Linear Regulators and SMPSs

Power supplies can be divided into two main types: linear regulated and switch-mode power supplies (SMPS).

Linear regulators convert alternating current (AC) into a steady direct current (DC). They use a transformer to provide galvanic insolation between the high voltage AC input and low voltage DC output, to filter out noise from the input, and to reduce the input AC voltage to be rectified to output DC voltage. Linear regulated power supplies have a low efficiency (30-60 percent), which means more power losses and heat to dissipate.

Switch mode power supplies have a high efficiency (more than 80 percent) and advanced performance. SMPSs convert the incoming grid frequency voltage into a high-frequency AC voltage, then step it down to a lower voltage using a small transformer. The voltage is then rectified, filtered and regulated to be outputted as steady DC voltage.

SMPSs have a more complex design and higher cost compared to linear regulators. However, they have numerous advantages compared to linear regulators, including small cases and very high efficiency without dissipating excessive amounts of heat.

The Importance of Power Supply Efficiency

Power supply efficiency is the percentage ratio of total output power to input power, normally specified at full load and nominal input voltage:

For example, if the power supply unit has 50 percent efficiency and the requested load power is 100W, the total input power needs to be 200W. The other 50 percent of input power gets wasted, mostly as heat, and has to be dissipated properly to avoid damaging the unit. This indicates how important high efficiency power supplies are. They have significantly fewer losses, meaning they stay cooler and use less power from the network to provide the same output power.

Power supply devices do not have constant efficiency, as it is dependent on the environmental and load conditions. Generally, power supplies are at their maximum efficiency when operated at 50 percent of the nominal load. Power supply units have higher efficiency when supplied by 230V than 110V AC. Nowadays, the market offers advanced models that can operate with constantly high efficiency even in the range of 50-100 percent of nominal load. For example, the diagram in Figure 1 shows the efficiency curve of the MEAN WELL UHP-2500 series against the supplying voltage and load. 

It is not possible to achieve a power supply efficiency of 100 percent. However, with an advanced design and proper components, high efficiencies are possible (e.g. the MEAN WELL UHP-2500 in Figure 1 has efficiency up to 96 percent). Losses in power supplies are caused by passive and active components and occur more in linear supplies than in SMPSs. SMPS power losses occur mostly in switching devices (MOSFETS and diodes). Other losses are caused by the capacitors and inductors (this can be reduced if advanced low resistance components are used, which are usually costly). Passive component losses include resistor losses (I²R), capacitor losses and inductor losses due to core and winding induction. Active component losses are mostly caused by MOSFETs and diodes as conduction and switching losses.

Although SMPSs can be more expensive, their high efficiency saves electricity costs. They are also more reliable, produce less noise and require less cooling. They use high-quality components such as switching devices, heavy-duty capacitors and chokes with good characteristics. They produce better output voltage with low ripples, less noise and less heat, and they provide better voltage regulation.

Power Supply Cooling Methods

Power supplies generate heat which could damage the components in the circuit. Because of this, power supplies need to have a cooling system to dissipate excess heat. These are the most commonly used heat-conducting or cooling methods:

1. Natural airflow is a simple convection cooling method where air naturally flows and transfers heat away from a device. Natural airflow is a simple, low-cost cooling option that does not include additional mechanical components. Because this cooling type has low heat dissipation capability, power supplies using it are limited in their safe temperature operating range.
2. Forced air cooling involves one or more fans to improve natural airflow, allowing higher operating temperatures. Since they consume additional power, fans decrease the power supply efficiency. They also require regular maintenance and replacements. This cooling type is not practical for applications with limited space.
3. Heat sinks transfer heat through direct contact with a cooler component, which can be heat sinks (with fins) and cold plates (without fins). A heat sink with fins has a bigger surface area for contact with air and thus allows faster heat dissipation. It is often combined with air forced cooling for better dissipation. However, heat sinks require more space which limits their use in some applications. Cold plates are thick metal plates that are used to conduct heat away from the device. They are used in applications where simplicity and space are crucial parameters.
4. Liquid cooling consists of a cold plate with an embedded tube running a closed-loop of cool liquid. The liquid helps the base plate remain at a very steady temperature, making the heat transfer as efficient as possible. Since it eliminates the need for fans, liquid cooling does not take too much space. Liquid cooling has great heat dissipation, does not generate noise and vibration, and increases the power supply's lifetime. However, liquid cooling is more costly as it requires more complex design. Liquid cooling is commonly used in applications that produce high heat such as industrial lasers, or charging stations.

Liquid-cooled power supplies can be designed with very high output power. For example, MEAN WELL's liquid-cooled power supplies, the PHP-3500 series, are capable of outputting up to 3,500W. They also have an active current sharing function, allowing up to four parallel units to deliver up to 14,000W.

The video below contrasts the temperature difference between liquid and forced-air cooling of a MOSFET and capacitor. The temperature is 6-18°C lower with liquid cooling compared to forced-air cooling.

Although costly, liquid-cooled power supplies are the most effective solution for applications where forced-air cooling is not suitable. In many applications, forced-air cooling should be avoided since adding a fan to a power supply decreases its reliability, adds cost, creates noise and introduces a maintenance headache.

A fanless power supply (such as the MEAN WELL UHP-2500 in Figure 2 below) avoids issues with fan failures and reduces the maintenance costs. Fanless supplies are also suitable for environments that requires absolute quiet.

Power Supplies and Power Management

Power management is an important feature in every electronic system because it controls, regulates and distributes DC power to the system. Power management affects reliability, performance and cost.

Advanced power supplies are equipped with programmable voltage and current (PV/PC) functions, which allows the trimming of the output voltage between 50-120 percent and current between 20-100 percent by an external DC control signal. Additionally, they can have communication protocols (e.g. PMBus or CANBus) to achieve an intelligent power supply. Generally, there are two types of power supply output regulation: external analog signal interface and digital control interface.

External analog signal interface control technology uses the external voltage to control the power supply output voltage. Analog power supplies are stable and reliable, but they have limited flexibility. While status can be monitored, control is usually limited to one possible reaction per parameter. External analog voltage input is generally in the range of 1-5V. The output current can be trimmed to 20-100 percent of the rated current by applying external voltage. Figure 3 shows the diagram of output current regulation using external DC voltage.

Digital control provides access to a high level of information and enables more flexible output. The table below shows the advantages of digitalized power supply design:

The most widely-used communications protocol for digital control of power supplies is PMBus (Power Management Bus). PMBus is an open standard power-management protocol that provides simpler design and reduces the complexity of digitally controlled power systems. CANBus (Controller Area Network) is another communication protocol. It can achieve most of the commands and functions of the PMBus protocol, with the major difference between them being the signaling method. The choice between CANBus and PMBus depends on the applications’ specific needs. MEAN WELL offers optional CANBus models for products already integrated with PMBus.

Power supplies with built-in PMBus protocol allow the end-system to monitor and control the power supply's various operating parameters such as voltage, current and temperature. Fanless power supplies combined with the PMBus protocol greatly improve lifetime reliability and installation flexibility, enabling more compact and intelligent end-systems. The PHP-3500, for example, supports PMBus with a maximum 100kHz bus speed, allowing information reading, status monitoring and output trimming.

OnlineComponents.com has sponsored this post.  All opinions are mine.  --Edis Osmanbasic