WEBVTT

00:00:00.000 --> 00:00:10.150 align:middle line:90%


00:00:10.150 --> 00:00:10.790 align:middle line:90%
Hello.

00:00:10.790 --> 00:00:14.210 align:middle line:84%
My name is Frank De Stasi, and
I'm an applications engineer

00:00:14.210 --> 00:00:16.320 align:middle line:90%
here at Texas Instruments.

00:00:16.320 --> 00:00:17.890 align:middle line:84%
And today, we're
going to be looking

00:00:17.890 --> 00:00:22.980 align:middle line:84%
at the EMI of a
nonsynchronous buck converter.

00:00:22.980 --> 00:00:26.490 align:middle line:84%
Now, many engineers believe
that it's almost impossible

00:00:26.490 --> 00:00:28.990 align:middle line:84%
to get the same
good EMI performance

00:00:28.990 --> 00:00:32.159 align:middle line:84%
from a nonsynchronous
regulator as it is

00:00:32.159 --> 00:00:34.240 align:middle line:90%
from a synchronous converter.

00:00:34.240 --> 00:00:35.860 align:middle line:84%
And today, we're
going to show you

00:00:35.860 --> 00:00:38.090 align:middle line:84%
how to get just as
good performance

00:00:38.090 --> 00:00:40.820 align:middle line:84%
from a nonsynchronous
regulator as you can

00:00:40.820 --> 00:00:42.890 align:middle line:90%
from a synchronous converter.

00:00:42.890 --> 00:00:45.600 align:middle line:84%
So first, let's take a
look at the differences

00:00:45.600 --> 00:00:50.630 align:middle line:84%
between a synchronous and
a nonsynchronous converter.

00:00:50.630 --> 00:00:53.270 align:middle line:84%
We see on this slide
on the left-hand side

00:00:53.270 --> 00:00:56.700 align:middle line:84%
a typical synchronous
buck converter.

00:00:56.700 --> 00:00:58.970 align:middle line:84%
Now, we can see that
the power MOSFETs

00:00:58.970 --> 00:01:02.450 align:middle line:90%
M1 and M2 are inside the IC.

00:01:02.450 --> 00:01:05.410 align:middle line:84%
So therefore, they're
in very close proximity.

00:01:05.410 --> 00:01:08.680 align:middle line:84%
The input capacitor is
also very, very close

00:01:08.680 --> 00:01:10.790 align:middle line:84%
to the IC, because
it's physically

00:01:10.790 --> 00:01:13.310 align:middle line:90%
smaller than the IC itself.

00:01:13.310 --> 00:01:16.010 align:middle line:84%
Now, the critical
loop that we have

00:01:16.010 --> 00:01:18.710 align:middle line:84%
to deal with when
dealing with EMI

00:01:18.710 --> 00:01:22.570 align:middle line:84%
is between the input capacitors
and the two MOSFET transistors

00:01:22.570 --> 00:01:24.450 align:middle line:90%
that you see here.

00:01:24.450 --> 00:01:28.180 align:middle line:84%
The reason that this
current loop is critical

00:01:28.180 --> 00:01:32.270 align:middle line:84%
is that the currents in the
loop are changing very fast.

00:01:32.270 --> 00:01:35.490 align:middle line:84%
They're turning on and
off at a very rapid rate.

00:01:35.490 --> 00:01:38.370 align:middle line:84%
And that's basically
what causes both

00:01:38.370 --> 00:01:41.380 align:middle line:90%
conducted and radiated EMI.

00:01:41.380 --> 00:01:45.290 align:middle line:84%
So the whole key to
minimizing your EMI

00:01:45.290 --> 00:01:48.395 align:middle line:84%
is to keep that critical
loop area very, very small,

00:01:48.395 --> 00:01:51.100 align:middle line:84%
or at least as
small as possible.

00:01:51.100 --> 00:01:54.550 align:middle line:84%
Basically, what happens is that
small loop between the input

00:01:54.550 --> 00:01:58.840 align:middle line:84%
capacitors and the power
MOSFETs creates a small antenna

00:01:58.840 --> 00:02:01.330 align:middle line:90%
that radiates EMI.

00:02:01.330 --> 00:02:03.530 align:middle line:84%
Now, if we look over here
on the right-hand side,

00:02:03.530 --> 00:02:06.610 align:middle line:84%
we see a typical
nonsynchronous converter.

00:02:06.610 --> 00:02:08.990 align:middle line:84%
In this case, we
have a Schottky diode

00:02:08.990 --> 00:02:11.730 align:middle line:84%
that takes the place
of transistor M2.

00:02:11.730 --> 00:02:14.080 align:middle line:84%
And that's the key difference
between these two power

00:02:14.080 --> 00:02:15.160 align:middle line:90%
supplies.

00:02:15.160 --> 00:02:19.920 align:middle line:84%
Because the Schottky diode is
external and physically larger

00:02:19.920 --> 00:02:24.400 align:middle line:84%
than typically the input
capacitor or the IC sometimes,

00:02:24.400 --> 00:02:26.150 align:middle line:84%
that necessitates
the current loop

00:02:26.150 --> 00:02:29.040 align:middle line:84%
to be actually a little
bit larger than it

00:02:29.040 --> 00:02:32.690 align:middle line:84%
is in a synchronous converter
for the same current rating

00:02:32.690 --> 00:02:35.070 align:middle line:84%
and the same voltage
and power ratings.

00:02:35.070 --> 00:02:38.210 align:middle line:84%
So you can see here, basically
with the nonsynchronous

00:02:38.210 --> 00:02:40.800 align:middle line:84%
converter, you're
almost forced to have

00:02:40.800 --> 00:02:43.760 align:middle line:84%
a slightly larger critical
loop area than you

00:02:43.760 --> 00:02:47.770 align:middle line:90%
do for a synchronous converter.

00:02:47.770 --> 00:02:49.990 align:middle line:84%
Now, here's a typical
example of a layout

00:02:49.990 --> 00:02:51.740 align:middle line:90%
for a synchronous converter.

00:02:51.740 --> 00:02:55.130 align:middle line:84%
On the right-hand side,
you see the IC itself

00:02:55.130 --> 00:02:57.540 align:middle line:84%
with the two power
MOSFETs inside,

00:02:57.540 --> 00:03:00.080 align:middle line:84%
and you see the input
capacitor right next to it,

00:03:00.080 --> 00:03:01.840 align:middle line:90%
physically very close.

00:03:01.840 --> 00:03:05.270 align:middle line:84%
So you can see by moving the
input capacitor as close as

00:03:05.270 --> 00:03:08.300 align:middle line:84%
possible to the IC
that current loop

00:03:08.300 --> 00:03:10.820 align:middle line:90%
area is going to be very small.

00:03:10.820 --> 00:03:12.720 align:middle line:84%
In this case, the
dotted red line

00:03:12.720 --> 00:03:17.340 align:middle line:84%
signifies the critical loop area
that we're trying to minimize.

00:03:17.340 --> 00:03:20.140 align:middle line:84%
In the nonsynchronous
converter, it's

00:03:20.140 --> 00:03:23.150 align:middle line:84%
easy to see that we have the
Schottky diode designated

00:03:23.150 --> 00:03:27.170 align:middle line:84%
as D1, which is
actually a little bit

00:03:27.170 --> 00:03:29.880 align:middle line:84%
larger than the IC itself
and quite a bit larger

00:03:29.880 --> 00:03:31.590 align:middle line:90%
than the input capacitor.

00:03:31.590 --> 00:03:33.520 align:middle line:84%
So it makes it a
little bit difficult

00:03:33.520 --> 00:03:37.020 align:middle line:84%
to get all the components as
close as possible-- a little

00:03:37.020 --> 00:03:40.760 align:middle line:84%
bit more difficult than it is
in a synchronous converter,

00:03:40.760 --> 00:03:43.780 align:middle line:84%
but it's not impossible
to get a good, tight loop.

00:03:43.780 --> 00:03:46.370 align:middle line:84%
But you can see here
from the dotted box that

00:03:46.370 --> 00:03:49.160 align:middle line:84%
represents the critical
loop area that it's always

00:03:49.160 --> 00:03:53.750 align:middle line:84%
going to be a little bit
more than for a synchronous

00:03:53.750 --> 00:03:57.000 align:middle line:90%
converter would be.

00:03:57.000 --> 00:04:00.810 align:middle line:84%
So the question is how do
we get good EMI performance

00:04:00.810 --> 00:04:04.160 align:middle line:84%
from a nonsynchronous
buck converter.

00:04:04.160 --> 00:04:07.530 align:middle line:84%
Well, there are three
basic principles

00:04:07.530 --> 00:04:09.170 align:middle line:90%
you have to keep in mind.

00:04:09.170 --> 00:04:12.390 align:middle line:84%
One, as we just mentioned,
is to keep the critical loop

00:04:12.390 --> 00:04:16.070 align:middle line:84%
area as physically
small as possible.

00:04:16.070 --> 00:04:21.670 align:middle line:84%
Second, use both a top side
and a bottom side ground plane

00:04:21.670 --> 00:04:23.590 align:middle line:90%
on your PC board.

00:04:23.590 --> 00:04:27.780 align:middle line:84%
Those ground planes will help
to manage the loop currents

00:04:27.780 --> 00:04:31.930 align:middle line:84%
and keep them where
you want them to flow.

00:04:31.930 --> 00:04:37.400 align:middle line:84%
And finally, make sure that
you use both common mode

00:04:37.400 --> 00:04:41.850 align:middle line:84%
and differential mode filters on
the input to your power supply.

00:04:41.850 --> 00:04:45.490 align:middle line:84%
Those filters will help to
filter out the fast rising

00:04:45.490 --> 00:04:49.300 align:middle line:84%
edges of the currents in
those critical current loops

00:04:49.300 --> 00:04:50.870 align:middle line:90%
we just talked about.

00:04:50.870 --> 00:04:54.850 align:middle line:84%
And in fact, the filters
will help with both radiated

00:04:54.850 --> 00:04:58.260 align:middle line:90%
and conducted EMI.

00:04:58.260 --> 00:05:01.900 align:middle line:84%
So we have an example
to show you here today.

00:05:01.900 --> 00:05:06.710 align:middle line:84%
We have an evaluation board that
was designed by TI engineers

00:05:06.710 --> 00:05:12.640 align:middle line:84%
specifically to minimize both
conducted and radiated EMI.

00:05:12.640 --> 00:05:17.520 align:middle line:84%
And essentially, the board
has a very tight loop area,

00:05:17.520 --> 00:05:18.710 align:middle line:90%
as we mentioned.

00:05:18.710 --> 00:05:22.460 align:middle line:84%
It has both top side and
bottom side ground planes,

00:05:22.460 --> 00:05:29.150 align:middle line:84%
and it has both differential
and common mode input filters.

00:05:29.150 --> 00:05:34.810 align:middle line:84%
This particular board is
based on the LMR 16030

00:05:34.810 --> 00:05:39.050 align:middle line:84%
simple switcher
nonsynchronous book converter.

00:05:39.050 --> 00:05:41.100 align:middle line:84%
So this is what we'll
use as our example

00:05:41.100 --> 00:05:44.920 align:middle line:84%
today in our lab to see
what kind of EMR performance

00:05:44.920 --> 00:05:46.240 align:middle line:90%
we can get.

00:05:46.240 --> 00:05:51.160 align:middle line:84%
So the next step is we'll take
this board into the EMI chamber

00:05:51.160 --> 00:05:55.000 align:middle line:84%
and go ahead and
run an EMI scan.

00:05:55.000 --> 00:05:58.450 align:middle line:84%
So here we are inside
of our EMI chamber.

00:05:58.450 --> 00:06:00.930 align:middle line:84%
And we have the PC
board we saw earlier

00:06:00.930 --> 00:06:05.100 align:middle line:84%
set up and ready to
go to run an EMI scan.

00:06:05.100 --> 00:06:08.820 align:middle line:84%
The input power
supply to our PC board

00:06:08.820 --> 00:06:11.410 align:middle line:84%
is a pair of 12-volt
batteries and series

00:06:11.410 --> 00:06:14.620 align:middle line:84%
that gives us 24 volts
for our input supply.

00:06:14.620 --> 00:06:17.050 align:middle line:84%
And it's very important
to use batteries

00:06:17.050 --> 00:06:19.790 align:middle line:84%
whenever you're doing
any kind of EMI testing,

00:06:19.790 --> 00:06:25.160 align:middle line:84%
because they provide a very low
noise source for our regulator.

00:06:25.160 --> 00:06:27.400 align:middle line:84%
The output of this
particular regulator

00:06:27.400 --> 00:06:32.600 align:middle line:84%
is designed to provide 5 volts
with a load current of 3 amps.

00:06:32.600 --> 00:06:37.020 align:middle line:84%
And again, we also need
some very low noise load.

00:06:37.020 --> 00:06:39.560 align:middle line:84%
So what we choose
to use for our load

00:06:39.560 --> 00:06:42.480 align:middle line:84%
is essentially a set
of power resistors

00:06:42.480 --> 00:06:44.220 align:middle line:90%
that are noninductive.

00:06:44.220 --> 00:06:46.620 align:middle line:84%
And if you look underneath
the turntable here,

00:06:46.620 --> 00:06:49.420 align:middle line:84%
you will see that our
load resistors are sitting

00:06:49.420 --> 00:06:51.550 align:middle line:84%
on that heat sink,
out of the way

00:06:51.550 --> 00:06:55.280 align:middle line:84%
so they don't interfere
with anything else.

00:06:55.280 --> 00:07:00.010 align:middle line:84%
On the opposite side of
the corner to the PC board,

00:07:00.010 --> 00:07:02.910 align:middle line:84%
we have our EMI
receiving antenna.

00:07:02.910 --> 00:07:04.380 align:middle line:84%
And essentially,
what will happen

00:07:04.380 --> 00:07:08.900 align:middle line:84%
is that the EMI coming
off of our PC board

00:07:08.900 --> 00:07:11.370 align:middle line:84%
will be picked up
by the antenna, sent

00:07:11.370 --> 00:07:15.170 align:middle line:84%
through the coaxial cable,
and into the receiver that's

00:07:15.170 --> 00:07:16.710 align:middle line:90%
in the lab.

00:07:16.710 --> 00:07:20.850 align:middle line:84%
So now we're all set to go
back in the lab and run a scan.

00:07:20.850 --> 00:07:24.220 align:middle line:84%
Well, we have our PC board
inside the EMI chamber.

00:07:24.220 --> 00:07:27.800 align:middle line:84%
And it's all powered up, and
we're ready to run an EMI scan.

00:07:27.800 --> 00:07:29.518 align:middle line:84%
So let's go ahead
and run the receiver.

00:07:29.518 --> 00:07:42.540 align:middle line:90%


00:07:42.540 --> 00:07:44.650 align:middle line:90%
So now the scan is completed.

00:07:44.650 --> 00:07:48.930 align:middle line:84%
We just ran a radiated EMI
scan over the frequency range

00:07:48.930 --> 00:07:51.530 align:middle line:90%
of 30 megahertz to 1 gigahertz.

00:07:51.530 --> 00:07:57.420 align:middle line:84%
And it's easy to see that our
PC board has passed the CISPR 22

00:07:57.420 --> 00:08:00.440 align:middle line:90%
class B limits.

00:08:00.440 --> 00:08:02.820 align:middle line:84%
One thing to keep in
mind is that the chamber

00:08:02.820 --> 00:08:07.470 align:middle line:84%
we have here at TI is what's
known as a three meter chamber.

00:08:07.470 --> 00:08:11.150 align:middle line:84%
That means that the antenna is
approximately three meters away

00:08:11.150 --> 00:08:15.700 align:middle line:84%
from the PC board regulator
that we're trying to test.

00:08:15.700 --> 00:08:20.110 align:middle line:84%
Typically, however, in
industry, the antenna

00:08:20.110 --> 00:08:24.790 align:middle line:84%
will be approximately 10 meters
away from the board under test.

00:08:24.790 --> 00:08:29.170 align:middle line:84%
That means that
typically the EMI results

00:08:29.170 --> 00:08:32.559 align:middle line:84%
that you get at the antenna
will be a little bit less

00:08:32.559 --> 00:08:36.940 align:middle line:84%
than what we see here in
our three meter chamber.

00:08:36.940 --> 00:08:40.530 align:middle line:84%
So even though
this EMI scan seems

00:08:40.530 --> 00:08:43.850 align:middle line:84%
like it's coming very close to
the CISPR limit, in reality,

00:08:43.850 --> 00:08:45.770 align:middle line:84%
it will actually be a
little bit further away.

00:08:45.770 --> 00:08:51.300 align:middle line:90%


00:08:51.300 --> 00:08:54.710 align:middle line:84%
This slide shows us a comparison
between a typical synchronous

00:08:54.710 --> 00:08:58.040 align:middle line:84%
converter and the
nonsynchronous converter

00:08:58.040 --> 00:09:01.030 align:middle line:84%
that we just
scanned, essentially.

00:09:01.030 --> 00:09:06.320 align:middle line:84%
These scans were taken with an
industrial 10-meter chamber.

00:09:06.320 --> 00:09:09.370 align:middle line:84%
The red graph is
the nonsynchronous,

00:09:09.370 --> 00:09:11.750 align:middle line:84%
and the black graph
is the synchronous.

00:09:11.750 --> 00:09:14.860 align:middle line:84%
It's easy to see that
both converters easily

00:09:14.860 --> 00:09:18.600 align:middle line:84%
passed the CISPR 22
class B requirement.

00:09:18.600 --> 00:09:22.100 align:middle line:84%
It's also easy to see that
the nonsynchronous converter

00:09:22.100 --> 00:09:24.770 align:middle line:84%
at some frequency
ranges will have

00:09:24.770 --> 00:09:28.110 align:middle line:84%
a little bit more EMI than
a comparable synchronous

00:09:28.110 --> 00:09:29.040 align:middle line:90%
converter.

00:09:29.040 --> 00:09:30.690 align:middle line:84%
And that's what
you would expect,

00:09:30.690 --> 00:09:33.220 align:middle line:84%
because it's a little
bit more difficult to get

00:09:33.220 --> 00:09:36.510 align:middle line:84%
the critical loop area
as small as it would

00:09:36.510 --> 00:09:39.290 align:middle line:90%
be for a synchronous converter.

00:09:39.290 --> 00:09:42.430 align:middle line:84%
So in summary, to
achieve good EMI results

00:09:42.430 --> 00:09:45.290 align:middle line:84%
from a nonsynchronous
converter, just keep in mind

00:09:45.290 --> 00:09:47.040 align:middle line:90%
these three points.

00:09:47.040 --> 00:09:51.120 align:middle line:84%
Keep the critical PC board
loop area as small as possible.

00:09:51.120 --> 00:09:53.850 align:middle line:84%
Use generous ground planes
to keep the currents flowing

00:09:53.850 --> 00:09:58.720 align:middle line:84%
in small loops, and use both
common mode and differential

00:09:58.720 --> 00:10:01.610 align:middle line:84%
mode input filters
to help filter

00:10:01.610 --> 00:10:05.370 align:middle line:84%
out some of the EMI noise
generated by the switch.

00:10:05.370 --> 00:10:07.060 align:middle line:90%
So there you have it.

00:10:07.060 --> 00:10:10.040 align:middle line:84%
We've showed you today
how you can minimize

00:10:10.040 --> 00:10:13.490 align:middle line:84%
the amount of both
radiated and conducted

00:10:13.490 --> 00:10:17.340 align:middle line:84%
EMI in a nonsynchronous
design by following

00:10:17.340 --> 00:10:19.500 align:middle line:90%
a few simple rules.

00:10:19.500 --> 00:10:22.290 align:middle line:84%
So for more information,
please see the links

00:10:22.290 --> 00:10:24.060 align:middle line:90%
at the end of this video.

00:10:24.060 --> 00:10:26.190 align:middle line:90%
And thank you for watching.

00:10:26.190 --> 00:10:43.136 align:middle line:90%