Understanding Bandwidth - The #1 Test Gear Spec You Need to Know

What is bandwidth, really? Does it matter?
Click to subscribe! ► bit.ly/Scopes_Sub ◄
Link to the blog for a bonus tip:
bit.ly/ScopeHac...
Questions/comments? Ask me on Twitter:
@DanielBogdanoff / danielbogdanoff
Learn more about using oscilloscopes:
oscilloscopelea...
Check out the EEs Talk Tech podcast:
• EEs Talk Tech - An Ele...
Transcript:
Today, we’re going to take a look at a different way to think about bandwidth. What really is bandwidth, and what does it mean for your systems?
Hi, I’m Daniel Bogdanoff, and today we’re going to dig into arguably the most fundamental of all test gear specs - bandwidth.
In test gear 101 you hear that bandwidth is simply the highest frequency component that you can measure or produce. For example, a 100 MHz oscilloscope can measure a 100 MHz sine wave.
But, this signal is going to be attenuated by 3 dB or so, because in test gear 201 you hear that the signal will actually roll off a bit near the maximum bandwidth of your system. In fact, you can actually see frequencies that exceed the bandwidth of your scope.
For example, here’s a 40 MHz, 5 Vpp signal piped into my 1.5 GHz oscilloscope. No problem.
But then, I turn on my scope’s 20 MHz bandwidth limit, I can still see my 40 MHz sine wave, but the peak to peak voltage has attenuated down to 2.1 volts. So essentially I’m able to see a signal that’s double the bandwidth of my scope.
The same is true for the output of a function generator, but it’s harder to see that in practice because the software won’t let you set a higher frequency than the generator’s stated max output. On a 100 MHz generator, you simply can’t turn the frequency setting any higher than 100 MHz.
So, this begs the question, why don’t we, the designers of test and measurement gear, simply say the generator can produce a higher frequency and then just limit the amplitude at those extended frequencies? Or, why can’t I say my oscilloscope has a higher bandwidth? Here I am able to measure double the bandwidth on this scope. It would certainly make our specs look better!
To answer that, we need to look more closely at what bandwidth really means. As it turns out, it’s not about the Hertz, the cycles per second, it’s about rise time. It’s about how fast the signal goes up and down. Edge speed.
Let’s look an example. A delivery driver has a car that can travel at 100 km/h max. He has to make a delivery 50 km away, so it’ll take him 1 hour to deliver the package and get back. His frequency is 1 delivery per hour. Next, he has to deliver two packages that are only 25 km away. It’ll take him only 30m for each package, and his frequency will be 2 deliveries per hour. His frequency doubled, but his speed didn’t change.
This is the same with test gear and bandwidth. An oscilloscopes bandwidth or generator’s output is like the delivery rate (deliveries per second), while the rise time is like the driver’s speed.
Sure, he can deliver more packages if the deliveries are closer together, but the travel speed doesn’t change. This is the same with test gear.
Another way to think about this is in terms of frequency components. Bandwidth refers to the frequency components when you look at the frequency domain, but the rise time or edge speed in the time domain. Watch my edge speed change as a toggle the bandwidth limit on and off.
Here’s what’s happening.
When the signal goes positive, the scope trace will start to move positive. But, it has to do so linearly, meaning it can’t just magically jump up to the new voltage value. The speed that the scope adjusts to the change is dictated by its bandwidth. So, you can technically measure a frequency higher than the max bandwidth of the scope, but you can’t measure an edge speed faster than the max bandwidth of the scope.
The same is true for a function generator. The edge speed of the output maxes out at its max bandwidth. That’s why a 100 MHz function generator can create a 100 MHz sine wave, but not a 100 MHz square wave. The Hertz is the same, but the edge speeds are very different.
So, if you’re picking out a function generator, think about what types of signals you want to generate, not just the max bandwidth. A general rule of thumb for square waves is that you need a generator with 3x the bandwidth of your square wave frequency.
You can use this knowledge to push some of the specs of your scope - which is a topic for an upcoming video.
#oscilloscope #oscilloscopes #electronics #electricalengineering #bandwidth #functiongenerator #tutorial #EE #electronicsengineering #rfengineering #engineering #Keysight

[@KeysightLabs]

Thanks for watching! A couple clarifications: 1. When the input voltage is higher than the currently plotted voltage, the waveform will continue to move positive. When it's lower than the plotted voltage, the waveform will move negative. So, there will be a phase shift between the input voltage and the actually plotted waveform. The plotted waveform will also likely be distorted. 2. In the RF engineering world, bandwidth refers to a span of frequencies instead of 0 Hz -n Hz. Filtering and transmission techniques and are used to filter out unwanted frequencies. For example, cellular 4G runs on a carrier frequency of 2-8 GHz with channel bandwidths of 5-20 MHz.

[@Jameswen-electronics]

*Please scope i never have any and im not able to work with my electronics😢😢*

[@Whopperjaw.]

Excellent video! I learned something new! ...I now have the information I need to complete my space laser!

[@NLS87]

"my 1.5GHz oscilloscope" -> dies

[@tomb8430]

Very good explanation. Never properly understood this, now it's crystal clear.

[@KeysightLabs]

Glad it helped!

[@Pawelr98]

I had a friend, who happily measured 100+MHz on 35MHz analog oscilloscope. With digital scopes there's a problem of ADC speed. With multi channel operation the possible frequency goes down. For analog it all depends on components and quality. My 250MHz analog has RF transistors on input, allowing for 500MHz signals to be seen. However my slow portable 5MHz analog will have trouble keeping up even at 8-9MHz, having sync problems(unless the signal is also connected to external sync input) and introducing plenty of distortion into the signal.

[@jafarhussain482]

I have a question .. When an oscilloscope has low bandwidth then what kind of signal is not displayed correctly A- sine wave B- SQUARE WAVE C- TRIANGULAR WAVE D- MODULATED Reply me fast

[@KeysightLabs]

In general, fast edges slow down and amplitudes are reduced as the signal exceeds the system bandwidth.

[@praveenkumarsingh1436]

Suppose u have sent 100 Mhz sine on oscilloscope and scope's bw is set to 100 Mhz.... Then sine will be displayed correctly with 30% attenuation. But if u have sent 100 Mhz square, then high frequency components of square (300Mhz, 500Mhz and so on) will be lost so square signal can't be represented correctly with only 100 mhz..

[@Jameswen-electronics]

*Please scope i never have any and im not able to work with my electronics😢😢*

[@blitznoobz1457]

So this is like monkey see monkey do but the person copying ( Scope) is a slow person and can't keep up as fast ( due to the limit ) and it has to follow what the input does and I can think of it as lagging behind?

[@KeysightLabs]

Right, that's a good way to think about it

[@JimmyMakingitwork]

So we either need a super expensive high bandwidth scope to make sure we are covered, or we need to know in advance what it’s primary use will be to save money on tools. I feel a little smarter now.

[@yaghiyahbrenner8902]

Imagine the slew rate margins internally, how tight must be errh!, factor in temperature, drift over time and the expected performance of the instrument 20 years from now. Slew rate reminds me of the video: The paradox of the derivative | Essence of calculus, chapter 2

[@ameliabuns4058]

I wonder if it's possible to compensate for the effect with software?

[@Tutoelectro1]

Nice video David. Clear explanation!

[@KeysightLabs]

Thanks!

[@kruskotv1311]

Hes talking about test gear 101, is that some video ? And what about this "test gear 201" , is that a video or article ? Please link it

[@volvo09]

Thanks for that, really helps when trying to understand scope specs, and show how useless some of these low end "chinese" scopes can be with super low bandwidth.

[@Surya9201]

Hi, in digital scopes we have ADC to measure voltage at every clock/ sampling rate. And for high speed Digitizer input voltage range is merely 2Vpp at 50ohm termination. Can we do same here by selecting small input range and feeding small voltage but higher frequency's. I understand rise time issues will be improved since now it has to swing few voltage/mV at high speed.

[@RobeenaShepherd]

I knew the rules, now I understand why the rules, thank you! 💝

[@SunilYadav-th1wb]

I don't like, the wastage of paper. Rest all good thanks man

[@ronaldschild157]

I'm glad to learn I'm not the only one feeling this way.

[@T-She-Go]

I love this. Thank you

[@Alex-rv8gw]

what is the time domain and frequency domain ???

[@Motorman2112]

Is 2:56 the same thing that happens when you put a capacitor across the signal line?

[@nickstanley5064]

Yes, because the capacitor and the output impedance of your source will form an RC low pass filter, filtering out higher frequency components and reduce the signal's rise time and limit the bandwidth of your system. How much will your signal be affected depends on the RC components.

[@beachboardfan9544]

WTS why have I been missing vids! CMON YT!

[@KeysightLabs]

Hit the alarm bell icon, and you'll get notifications for every video!

[@beachboardfan9544]

I did! ...months ago!

[@grhinson]

The easel is is not facing the camera and the editing could be tightened up, but everything else is good. At least your putting out help videos

[@KeysightLabs]

Thanks for the feedback, we're definitely working on tightening everything up!

[@AvenEngineer]

but does it have a FULL BRIDGE RECTIFIER?!?!?!?

[@Fabi33677]

you mean the MOTHER of ALL FULLEST BRIDGE RECTIFIER ?

[@Eagleoneradiogod]

He was only able to obtain a one way mono use single polarity partial bridge rectifier!

[@ronvaliant9337]

Electroboom

[@3dsmaxrocks699]

Check out my ELECTRIC guitar....

[@HamedAdefuwa]

Funny to hear him say why don't manufacturers just state a higher spec and limit the amplitude, that's literally what all the Chinese cheap manufacturers do lol.

[@ovalwingnut]

👍😁 Thank you

[@B3.1425]

Amazing

[@fightocondria]

*Raises the question

[@HansDelbruck53]

My delivery driver's truck will do 100 Miles Per Hour because I live in America.

[@KeysightLabs]

#NextDayDelivery

[@kaustavsengupta8757]

*When you decided to pull out your big guns(1Ghz oscilloscope) to educate common masses onto bandwidth 101....😂😂😂

[@KeysightLabs]

Haha, yeah. I like to use the ones with bigger screens because they're easier to see on camera.

[@padmanabhaprasannasimha5385]

@@KeysightLabs that actually makes a lot of sense. I always used to think this is y'all just showing off neat gear..

[@johnavonvincentius1884]

hello?

[@johnavonvincentius1884]

bbbut i dont have a scope

[@pmcouto]

Keysight will be happy to sell you one! ;-)

[@RemcoStoutjesdijk]

So... you expect me to buy a 1.5 GHz scope, but you don't think I understand slew rate. WTF do I need a scope that fancy for?

[@KeysightLabs]

Nah, of course not! I like to use the big scope for video because it's a lot easier to see the screen.

[@johnvine5731]

There's a lot of background music consuming my bandwidth. Unlistenable.

[@davidmotley4207]

Wasting paper, wasted your video. Not too much content.

[@volvo09]

OMG, the paper! It could have been used for the the upcoming TPS report.