Previous in Forum: Analog and Digital Next in Forum: SENSOR TO COMPUTER INTERFACE
Participant

Join Date: Mar 2008
Posts: 1

# Switching Power Supplies

09/06/2008 12:41 PM

Hi, I would like to know What is a " SWITCHING POWER SUPPLY" ? I know what is a power supply but why : SWITCHING?

Thank you very much to all

Tony Varone

Interested in this topic? By joining CR4 you can "subscribe" to

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!
Power-User

Join Date: Jan 2008
Posts: 141
#1

### Re: Power supply

09/06/2008 2:48 PM

Wikipedia is your friend when it comes to questions like this.

http://en.wikipedia.org/wiki/Switching_power_supply

__________________
Perfection is achieved, not when there is nothing left to add, but when there is nothing else to take away
3
Guru

Join Date: Oct 2007
Location: Sour Lake, TX 30°08'59.68"N 94°19'42.81"W
Posts: 671
#2

### Re: Power supply

09/06/2008 5:10 PM

I agree that the Wikipedia answers are very good. This time, I feel that they haven't stressed the difference between the switching and linear power supplies.

A linear power supply has the main part a device that has the input connected to the unregulated DC voltage and the output to the load. The output is regulated to a low or higher level of regulation. A few millivolts are common for most of them. So, if the input voltage is 20V, the output (regulated) voltage is 10V and the consumption is 1A, the difference (20-10)x1=10W must be dissipated by the series element. This is lost power. You might say that the efficiency is 50%.

A switching power supply will have, mostly, an inductive element that accumulates and delivers the accumulated energy at a high speed. The frequency is several hundreds KHz up to GHz area.

We know that an inductor accumulates energy when a current is passing through it. There are some losses (the resistance of the wire of the inductor (some will add the Foucault losses, too). If you have a SPST switch to one end of the inductor and to the ground ; if the other end of the inductor is connected to the unregulated dc voltage, every time you turn on the switch the current flows from the unregulated voltage, through the inductor, to ground. When you turn off the switch, the energy accumulated in the inductor can be collected into a load. The advantage of a switching PS is that the switch does not dissipates too much power, the inductor neither, so the efficiency of a switching power supply is 80-90-96 %. My two bits.

__________________
Bridge rule #1: Nobody is as good as he thinks about himself nor as dumb, as his partner thinks...
Guru

Join Date: Nov 2007
Location: Christchurch, (The Garden City), South Island, New Zealand
Posts: 4395
#3

### Re: Power supply

09/07/2008 3:06 AM

Hello indel

from me

Kind Regards....

__________________
"The number of inventions increases faster than the need for them at the time" - SparkY
Off Topic (Score 5)
Guru

Join Date: Oct 2007
Location: Sour Lake, TX 30°08'59.68"N 94°19'42.81"W
Posts: 671
#8

### Re: Power supply

09/09/2008 10:05 PM

I always wanted to go to New Zealand. They say that the places are fabulous. I would dare say the people too.

__________________
Bridge rule #1: Nobody is as good as he thinks about himself nor as dumb, as his partner thinks...
Off Topic (Score 5)
Guru

Join Date: Apr 2007
Posts: 596
#4

### Re: Switching Power Supplies

09/08/2008 3:30 AM

visit: www.smps.us/ to know everything.

__________________
Subramanyam
Guru

Join Date: Nov 2007
Location: Christchurch, (The Garden City), South Island, New Zealand
Posts: 4395
#5

### Re: Switching Power Supplies

09/08/2008 7:48 AM

Hello kvsubramanyam

from me,

Excellent website reference, thanks.

Kind Regards....

__________________
"The number of inventions increases faster than the need for them at the time" - SparkY
Off Topic (Score 5)
Anonymous Poster
#6

### Re: Switching Power Supplies

09/08/2008 8:21 AM

In short a linear power supply is basically a variable resistor in series with the input power source that drops some of the input voltage to maintain the proper output voltage. The value or the resistor changes automatically to maintain the correct output voltage as load current changes and or input voltage changes. In practice the variable resistor is actually a transistor ( usually bipolar or MOSFET). The best possible efficiency for an ideal linear regulator is Vout/Vin and power dissipation is (Vin - Vout)*Iload. Figure a 12V to 3.3V at 2A - pretty inefficient even for an ideal case.

A switching regulator in short switches between the input source voltage and 0Volts with an appropriate duty cycle so the average voltage is the desired output voltage. An ideal buck type regulator duty cycle is 100% * Vout / Vin, the 12V to 3.3V regulator mentioned above would require the switch to be on for 27.5% of the time. The switching speed ranges from a few KHz to a few MHZ - most of my work is in the 300KHz - 500KHZ range. The squared up wave form is averaged using an LC filter to reduce it to a DC level with some output ripple. The Ideal switching regulator is 100% efficient but in practice is suffers switching losses and control losses that can be considerably large in some instances.

The switching regulator is much larger and produces much more EMI than a linear regulator but it is much more efficient in many instances. Switching regulators can also be used to provide full galvanic isolation between the input and output by using a transformer and isolated feedback. They are often referred to as DC to DC converter when they are isolated and switching regulators when non-isolated but the either term can be correctly applied to either type. We use the convention mentioned because it seems to be common within the industry.

Normally I would have said google it for an initial definition but that was fun. Now Google it for the details. These things are incredibly complex and their hundreds of way they are built. Good Luck

Guru

Join Date: Nov 2007
Location: Christchurch, (The Garden City), South Island, New Zealand
Posts: 4395
#7

### Re: Switching Power Supplies

09/08/2008 5:57 PM

Hello "Guest",

from me,

Kind Regards....

__________________
"The number of inventions increases faster than the need for them at the time" - SparkY
Off Topic (Score 5)
Power-User

Join Date: Jan 2008
Location: since 20 Jan 09, the USSA
Posts: 375
#9

### Re: Switching Power Supplies

09/21/2008 12:00 AM

There are bits and pieces of the truth in the other answers, but some misleading information as well. They tend to focus on the regulation, when the OP question was about switching vs. linear supplies.

Linear power supplies can only operate from an ac bus. Linear regulation assumes dc power, but that regulation in a linear supply occurs after the ac has been converted to dc via diode rectification and filtering via storage capacitors. Also, if the desired dc output potential is less than the peak of the rectified ac mains (170 Volts in the USA), you need a step-down transformer to convert the mains potential to the desired potential.

Linear supplies require large transformers and capacitors due to the low frequency of the ac mains, either 50/60 Hz. That requires a lot of iron in a transformer to get the inductance necessary, and a lot of capacitance to provide holdup between adjacent peaks of the rectified waveforms (10 or 8 ms apart for full wave rectified 50/60 Hz mains).

A switched mode power supply (SMPS) rectifies the ac mains with no intervening transformer, then provides a rough dc filter in the form of a single electrolytic cap of a few hundred microfarads. The resulting dc potential is applied though an inductor or transformer to the drain or collector of a transistor, which switches at a rate typically between 20 and 500 kHz these days, lower frequenies for higher power, higher frequencies at low power. Because the switching frequency is so high relative to the mains 50/60 Hz, the inductor or transformer is proportionally smaller, and so is the output filter capacitor. It is the smaller size of these two components that is the driving reason for using switchers instead of linear supplies. Also, the regulation in an SMPS is very efficient; it simply draws less current from the mains source when less power is required. Whereas regulation in a linear supply burns up unneeded power in the regulator.

Because a SMPS is really switching dc power, it is also the only kind of power supply you will find on a dc bus, such as in an automobile, a small airplane or the dc bus of a large aircraft. The SMPS is the only efficient electronic means of converting one dc potential to another, and such SMPS are also termed "dc/dc converters."

The main drawback of an SMPS is that because it switches at a high frequency, it causes rfi. SMPS have to be filtered to meet rfi type requirements. Whole books have been written about how to do this.

Another drawback is stability. In a linear supply, if the bus potential sags, the power supply output will either stay the same, if there is enough headroom in the regulator section, or it will sag with the bus. With a switcher, when the bus sags, the switcher acts like a constant power load. It either increases the switching duty cycle, or increases the switching frequency, depending on design, but ends up drawing more current from the bus to make up for the lost bus potential. The I*V product remains the same. This is termed "negative resistance" since when the potential sags, current increases, just the opposite of what happens with a resistor. If the extra SMPS current draw then sags the bus a little more, the SMPS must draw yet a little more current to make up for it, and then the bus sags some more, and... you can see where this is heading. Special stability criteria have been developed so that SMPS are designed to operate with stability.

If you want a feel for the difference in size and mass of linear and switching power supplies of the same power output, consider that your PC power supply is typically about 200 Watts. Now go find a 100 watt per channel stereo amplifier and look at the power supply section (transformer, rectifiers and electrolytic filter caps). Along with the above explanation, that will answer all your questions.

Participant

Join Date: Apr 2010
Location: Shenzhen,China
Posts: 2
#10

### Re: Switching Power Supplies

04/28/2010 10:37 PM

Hello,Tony.

I am Emma from Wearnes Global Co.Ltd( www.wgl.com.tw ),weare a manufacturer of switching power supply.

Switching power supply is different from linear power supply which is heavily and different making materials.This technology origned from 1970's,now it plays an important role in our daily power supply.

__________________
Professional Manufacturer of Power Supply--emma.yan@wgl.com.tw
Interested in this topic? By joining CR4 you can "subscribe" to

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!

Anonymous Poster (1); emc_c (1); emma.wearnes (1); indel (2); kvsubramanyam (1); Shawn_V_Elect (1); Sparkstation (3)

 Previous in Forum: Analog and Digital Next in Forum: SENSOR TO COMPUTER INTERFACE
You might be interested in: DC Power Supplies, Power Supplies