When to use AC Coupling on Your Oscilloscope

I’ve heard this comment before, and it’s dead wrong: “You should use AC coupling when looking at an AC signal.” Following this false AC coupling adage is an easy way to make some very wrong measurements.

When should you use AC coupling?

The Reason for AC Coupling on an Oscilloscope

AC coupling is intended for AC signals with a large DC offset.

For example, what if you want to look at ripple on a 12V power rail? You don’t really care about seeing the 12V DC component. What you really care about is the AC component around 12V.

Here’s the problem.

keep learning, sign up for free oscilloscope courses!

To see this ripple, you need a sensitive volt/div setting and a huge offset. An oscilloscope’s offset range is dictated by the volt/div setting, so getting a usable view of the ripple is often impossible in DC coupling mode.

That’s where AC coupling comes in.

Turning on AC Coupling

In the channel menu of your oscilloscope, you can turn on/off a couple of different settings. There’s often a 20 MHz bandwidth filter, probe settings, input impedance settings, and coupling settings. In the “coupling menu” you can switch between DC coupling and AC coupling. DC coupling allows you to see all signals from 0 Hz up to the max bandwidth of your scope. AC coupling filters out DC components.

When you enable AC coupling on an oscilloscope channel, you’re switching in a high-pass filter on the channel’s input signal path. This filters out all the DC components.

A generic low pass filter Figure 1: A generic low pass filter

The -3dB point of the AC coupling filter is typically specified in your oscilloscope’s data sheet. For example, the Keysight 1000 X-Series Oscilloscopes have a 10 Hz high pass filter. The Tektronix TDS1000B and TDS2000B also have a 10 Hz filter, and the Rigol 1000Z has a 5 Hz filter.

Figure 2: the AC coupling filter frequency of the 1000 X-Series Oscilloscopes

This seems straightforward, but there’s a catch. The AC coupling filter is an actual hardware filter. This means it’s not going to have a brickwall response - there’s going to be some roll-off.

So, a 10 Hz frequency component will still show up on screen, but it will be attenuated.

Note: As with most hardware filters, it’s often not a perfectly smooth frequency response. Sometimes there’s overshoot or inconsistencies, especially right around the -3dB point. This is normal, but means you should use AC coupling carefully. More on that in a moment.

Let’s take a look at this “roll-off” of the AC coupling filter.

Figures 3 and 4 show a 10 Hz signal with AC coupling on, and AC coupling off. Notice that the signal still shows up on screen, it just has an attenuated amplitude. It dropped from 1.03 Vpp to 0.68 Vpp!

AC Coupled Signal Figure 3: A 1 V, 10 Hz sine wave shows only 683 mVpp in AC Coupling Mode

A DC Coupled sine wave Figure 4: The same 1 V sine wave, but in DC coupling shows the proper 1 Vpp

Why is this happening? Because bandwidth isn’t as simple as we sometimes like to think. A 100 Hz oscilloscope can still see a 150 Hz sine wave, it’ll just be attenuated according to the roll-off characteristic of the oscilloscope. To learn more about the ins and outs of bandwidth and why this occurs, check out the article “What is Bandwidth? How Much Do You Need?

Keeping this in mind, there are times when you should and shouldn’t use AC coupling.

When to Use AC Coupling

Use AC coupling if you have a high frequency signal with a DC offset. This could be something like a power rail or a signal with a very low frequency drift. The goal of AC coupling is to remove DC. It’s not intended to remove non-zero, low frequency components.

I saw one situation where an engineer needed to watch a power rail turn on and then immediately see the ripple on the rail. The problem was that they couldn’t zoom in on their ripple and watch the rail turn on.

So, they double probed. They used one channel zoomed out with DC coupling mode to watch the rail turn on, and a second channel at a more sensitive time/div setting with AC coupling mode to observe the ripple.

For example, here’s a 1 kHz, 50 mV peak-to-peak sine wave riding on a 5 V DC offset. This could be a 5 V power rail, and we want to look at the ripple (Figure 5):

Can't zoom in enough in DC coupling Figure 5: We can't zoom in on our ripple while in DC coupling mode! (50 mVpp sine with a 5 V offset)

To see the ripple, we have to move the oscilloscope offset way below the screen (Figure 6):

Zoomed in as far as possible Figure 6: The best case scenario in DC coupling mode. We need AC coupling!

This view still isn’t very useful because the waveform is so small, so we want to zoom in farther. When we go from 500 mV/div to 200 mV/div, we run out of offset capability (Figure 7). The ripple is above the screen, but we can’t move the waveform any lower. Figure 6 is the best we can get in DC coupling mode.

No more room for offset Figure 7: If we zoom in farther, we run out of available channel offset!

Since DC coupling mode didn’t work very well, we can try using AC coupling. When we turn on AC coupling, as shown in Figure 8, we can zoom in as much as we want! The 5V offset is completely filtered out by the AC coupling filter.

AC coupling mode gives flexibility Figure 8: AC coupling gives us the view we need

So, this is a scenario where you’d want to use AC coupling. You shouldn’t use AC coupling just because you are looking at AC signals. You especially shouldn’t use it for something like the 50 Hz or 60 Hz wall power.

When You Shouldn’t Use AC Coupling

You shouldn’t default to AC coupling just because you are measuring an AC signal.

Turning on AC coupling mode in the wrong situation can hide DC components that you need to know about. Even worse, it can attenuate low frequency components into a warped, incorrect version of your signal.

For example, here’s a 100 Hz, 1 V square wave in DC coupling mode (Fig. 9) and AC coupling mode (Fig. 10):

A normal square wave in AC coupling mode Figure 9: In DC coupling, everything looks fine

AC Coupling causes a warped signal Figure 10: In AC coupling, the waveform is warped.

There’s a dramatic difference! The square wave we see in DC coupling mode is the actual signal, switching to AC warps the signal into something very different.

Oscilloscopes are designed to have a flat, consistent frequency response over the full bandwidth range. A little bit of offset or scaling might be required, but they can easily measure AC signals while in DC coupling mode. That’s what they’re made for.

To get more specific, you shouldn’t use AC coupling mode to view signals near the filter’s cutoff frequency. We recommend that you use AC coupling only for sine waves above 100 Hz and square waves above 200 Hz. If you try to measure signals below this range, you’ll have signal attenuation or overshoot.

It’s worth noting that it would be possible to make the filter response steeper, but a steeper cutoff leads to a lumpier frequency response. Anyone who’s implemented a hardware filter has experienced this design trade off. So, having a slower roll off means an overall smoother, more consistent frequency response.

Know your Test Gear

AC and DC coupling modes are a great example of why it’s important to have a solid understanding of your test gear. As we saw above, using AC coupling in the wrong situations will give you bad signal information and could lead you down the wrong debug path!

If you really want to keep learning and growing your test gear skills, check out a Keysight University course! Whether you are just starting out or have decades of electronics design experience, Keysight University has a course for you. I recommend Oscilloscope Probes 101 and Oscilloscope Probes 201, or maybe the Bench Power Supply Masterclass.

Keysight University

limit
3