How to Test Switch Mode Power Supplies with an Oscilloscope

How to Test Switch Mode Power Supplies with an Oscilloscope

If you’re working with switch mode power
supplies, you’re likely trying to improve efficiency, increased power density, and
reliability, comply with EMC regulations, improve power rail integrity, and
decrease thermals. This is a lot to think about and all this design optimization
can increase test time. Learn how to easily test your switch mode power
supplies and save time in the lab. Hi, I’m Melissa at Keysight technologies and in
this video you’ll learn how to test your switch mode power supplies with an
oscilloscope. The primary purpose of a power supply is to efficiently produce
well regulated and low noise DC power from an input power rail. Depending on
the part of the design you’re working on you might be looking at the switch mode
power supply holistically or only be concerned about one portion of the power
supply: the input, the switching, or the output. The input side converts and
filters your input voltage that’s typically an AC line voltage of a
110 volts in the U.S. or 220 in other regions. If you’re primarily
concerned with the input side of your supply you’ll likely be focused on power
quality, current harmonics, and inrush current. After the input your signal goes
into the main part of the supply which is the switching transistor. This
regulates the voltage with the duty cycle or the amount of time it’s on vs.
off. If this is the part of the power supply you’re focused on you’ll be
concerned about power losses, modulation analysis, slew rate, and save operating
area. After the switching transistor at the signal gets filtered again and
rectified so you get a stepped DC output which is then used to run power through
the rest of your device. If you’re focused on the output side of the power
supply you’ll want to focus on output ripple, turn on and turn off time,
transient response, power supply rejection ratio, and efficiency. An oscilloscope is the most common tool for making power supply measurements since
you can hook up both a voltage probe and a current probe to calculate power which
as you know is voltage times current. Another reason
why oscilloscopes are a great tool for characterizing power supplies are the
analysis applications that can run on them, making testing much more efficient.
Let’s set up some tests so you can see how some of these measurements are made.
If you have an Infiniivision oscilloscope and want to follow along
with the steps you can download a free trial of the power application with the
link on screen or in the description. First let’s start by looking at power
quality. I’ll clear any existing settings from previous tests by selecting default
setup, then I’ll press analyze and select the power application. You can now see the full list of the
power applications supported by the Infiniivision power application bundle.
I have a differential probe on the input side of my power supply and I connected
that to channel 1. I also have a current probe on the input side of my power
supply hooked up in the direction of current flow and that’s on channel 2.
Since we’ll calculate power from our current and voltage measurements I’ll
double-check that my channels are assigned on the scope to match this. I
choose the signals menu and the default on the scope is that channel 1 is
voltage and channel 2 is current. There’s also this handy diagram that pops up to
show you where to probe your power supply for this test. I’m also going to
look at 5 cycles. Now I can simply select Auto setup. Since I’m using a power
application, the scope automatically scales the signal properly to take
advantage of the full bits on the oscilloscope for accurate measurements
and sets up the waveform math. In this case of course it’s V times I to get
instantaneous power. Real power is the mean or average power of n cycles. That’s
the value you’re measuring directly. The application will also calculate the
other power parameters such as apparent power which is voltage RMS times current
RMS over a number of cycles. The reactive power is computed by the square root of
the apparent power squared minus the real power squared. Power factor is real
power over apparent power and the phase angle is the inverse cosine of the power
factor or real power over apparent power. We can dive into the details of power
quality results here by pressing apply. So the primary advantage of having an
automated power measurement application is that these are all calculated for you
with the press of a button, you don’t waste time working out the math with
pencil and paper or having to extract the voltage and current measurement
results to calculate power quality on a separate computer. Another important
characteristic to look at on your input line is the current harmonics. In a
switchable power supply your current is pulsed which can inject harmonics
back into the mains which can disturb other devices on your power network. If
your customer plugs your device into their power outlet at home you don’t
want the harmonics disrupting the performance of something else they have
running on the same source like their lights or TV. To prevent this IEC has
specifications that you’ll have to pass up to the 40th harmonic to make sure you
won’t be causing disruption to test. I turn on the current harmonics
measurement. We haven’t changed our probe setup but if this was the first test you
were setting up you can adjust those with the signal menus as before. In the
settings I can choose the frequency of the signal and the IEC standard that I
need to test to. My design is qualified as a Class A. Now
I’ll hit apply to begin the current harmonics measurement. The scope performs
an FFT on the current waveform and the results are listed in the top of the
screen. It also measures total harmonic distortion and tells me if I’ve passed
or failed the test. To show you an example of what happens when we fail
I’ll change the standard that I have to comply with. Now you can see where the
design would not pass these regulations and by what margin.
Let’s switch modes and look at the switching portion of the supply. You’ll
lose energy primarily during the switching phases of the transistor when
it turns on and off and during the conduction phase when the voltage is at
the transistor saturated minimum and current flows. To check if the losses
are acceptable I’ll choose the switching loss measurement. I also have to change
my probe setup to probe on the switching portion of my power supply. When I run
the application we can see two switching cycles on the top view and a zoomed in
view of one cycle on the bottom view. We also get all the power loss measurements
displayed automatically. You can zoom in and look at each phase: the conduction,
the switching, and the non conduction phase. The purple
waveform is the instantaneous power. Let’s look at one more measurement: slew
rate. This determines how fast your device turns on and off. The faster you
can turn on and off the lower losses you’ll have. We’ll follow similar setup
steps as before. The purple trace is the derivative of your voltage over time so
the peak of this is the maximum slew rate. In addition to the measurements I
cover today you can also use the Infiniivision power application to test
inrush current, effective resistance of your switching transistor, modulation, and
output ripple. Output ripple or parallel integrity measurements are particularly
important if you’re driving a high-speed digital device. The noise on the power
rail could cause jitter and timing uncertainty, and affect the validity of
your digital signal transmission. You can also measure transient response,
efficiency, power supply rejection ratio, and control loop response. You can use
these measurements to quantify and optimize your designs on the input side,
your switching, and the output side of your supply. If you want to learn more
about measuring switch mode power supplies check out the links in the
description and download a free trial of the application to try it out. Don’t
forget to follow us on Facebook Instagram and YouTube for more how-to
videos. Thanks for watching!

14 thoughts on “How to Test Switch Mode Power Supplies with an Oscilloscope

  1. To follow along step-by-step on your InfiniiVision oscilloscope, download a free trial of the power application

  2. Tv or screen? Most of the time the issue is the fact that they tape the wire directly to metal ! Just air gap and problem solved

  3. If you need to explain your customer, a ( often specialised) engineer, what to measure, how to measure and why he needs to measure that ( all rather basic things he must know already) you target the wrong market or the state of electronic engineering knowledge outside China has reached a tragic low level.

    I do not design SMPS, I only repair them, ( a lot, in test, calibration gear and specialized industrial gear, for instance from customers of Agilent gear to old for Keysight to support, and also without schematics because you do not want customers to repair their own test gear anymore (something HP and Tek always promoted)

    But regarding all your promotion/tutorials/publications etc repair techs are no target market, they cost you money. But you forget the customer that is forced to buy new often does not wants the same brand anymore. For me my scope is as important as my meters (worn agilent, a Brymen 869 that refuses to die, a Keithley 7510 and 2000) I use my scope so often that I wore down a Rigol in 1 year (10 years ago I made that mistake), My 350 MHz R&S is now 8 years old and needs replacement (repaired it twice already) . Most probes last a few moths. But I am not a hobbyist, not a student and not a product developer so who cares what repair-techs (many are also EE) do with a scope, or develop functions for them, or even tell them why you should buy a Keysight for repair. The same for multimeters (not only you) that is most times only focust on industry. My Agilent came with a rechargeable battery but who comes with the bizar idea to only charge it via the testlead sockets so you can not use it during charging (with a 24V powersupply ????).

    Sorry, I am a bit in rant mode, it is not personal. I think you do the videos very well.

  4. Down-voted because of the SIGNAL NOISE that is that annoying background "music".
    It only subtracts, not adds.
    Please cease doing this.
    It make the video seem to be a scam.

Leave a Reply

Your email address will not be published. Required fields are marked *