A Differential Probe Guide - How & Why To Use a Differential Probe With Your Oscilloscope
Vložit
- čas přidán 20. 08. 2018
- Your primer on the how to measure differential signals with an oscilloscope
Click to subscribe! ► bit.ly/Scopes_Sub ◄
Questions/comments? Ask me on Twitter:
@DanielBogdanoff / danielbogdanoff
Also check out the EEs Talk Tech podcast:
• EEs Talk Tech - An Ele...
Transcript:
Today we’re going to look at the two main reasons differential probes can save your bacon when you need to make scope measurements that aren’t ground referenced.
Hi, thanks for tuning in, I’m Daniel Bogdanoff and today we’re going to look at the two main reasons you’ll want to use a differential probe.
Before we get to that, what is a differential probe and how is it different from a standard single ended probe?
A single ended probe measures the voltage difference between earth ground and the probe tip. Like a single ended probe, a differential probe also measures the difference between two points. Unlike a single ended probe, a differential probe doesn’t have a ground clip - you can measure the voltage difference between any two points.
With a single ended probe, the oscilloscope will find the voltage difference. But for a differential probe, there’s a differential amplifier inside of the probe itself. Essentially, it pre-conditions the signal so that the oscilloscope gets a usable input. The oscilloscope hardware doesn’t care if you’re using a single ended probe or a differential probe.
We need a differential probe.
So, let’s hook it up to channel 4 and we’re now able to see what’s actually happening in our system.
These are just two simple examples, but this situation pops up pretty regularly. For example, differential probes are often used for measuring switch mode power supplies, inverters, and motor drivers - really any time you don’t want ground-referenced measurements. This is also what makes DMMs and handheld scopes so useful - they don’t make ground-referenced measurements.
Differential probes are also used to measure differential signals, which are common in for audio equipment and high speed serial communication lines. Audio and PCIe have vastly different signal speeds, they’re pretty much as far apart as you can get. So, why would they use differential signaling? Noise.
Differential signals, and hence differential probes have a tolerance for common mode noise. Common mode noise is noise that occurs on both lines of a differential signal. Differential receivers (and differential probes) only care about the difference between the two signals, not their individual magnitude. So, if both signals jump up a volt or 100 volts simultaneously, the voltage difference doesn’t change. For audio applications, this gets rid of a lot of audible noise[ramp audio noise then cut it]. For high speed digital signals, this dramatically improves the tolerance to noise sources like cross talk and power-plane induced noise.
If I’m a receiver, this is going to cause some problems. We’ll get ones where we should have zeros, and zeros where we should have ones. However, if we use a differential probe to probe the same signal, we get this!
As you can see, the sine wave is completely filtered out! Well, that’s not completely true. It’s more or less filtered out depending my probe’s common mode rejection ratio, known as CMRR. CMRR usually is dependent on the frequency of the noise.
My differential probe sees these signals the same way a differential receiver would in a communications system. You would want to use single ended probes to debug cross talk and interference, but a differential probe to analyze your communications data.
If you’re shopping for a differential probe or have the luxury of trying to decide between multiple differential probes you already have, here are a few things to consider:
First is Bandwidth - this one’s obvious, make sure to get adequate headroom in bandwidth. If you don’t have enough, it’ll affect both your measurements accuracy, especially for things like power measurements, as well as your ability to properly see high speed signals.
Second - Attenuation - attenuations vary widely, choose one that makes sense from a loading and signal amplitude perspective
Third, Input Range - How large of an input can it take? This applies to both common mode inputs and differential inputs
Finally, Common mode rejection ratio, or CMRR - If you’re expecting to deal with a lot of common mode phenomena, make sure the CMRR is adequate.
Fun fact, you can also use a differential probes as single ended probes - more on that in this video, which also shows how to use two single ended probes to measure differential signals!
#DifferentialProbe #DifferentialProbes #Probe #OscilloscopeProbe #Oscilloscope #DifferentialSignal #DifferentialSignals #ScopeProbe #Oscilloscopes #Keysight #CMRR #electronics - Věda a technologie
Oh a new video.
Really nice, this topic fits perfectly to my latest project with a differential amplifier in it.
Now i just need a scope. :)
Very well presented. Nice job!
And Boom goes the Dynamite!
Thank you so much again for awesome Video.
Thanks for adding the part at the end about hitting the bell button. Duh! i was wondering why I wasn't getting notified when a new video came out!
Thank-you Thank-you Thank-you ... for making sense out of this for me ...
Thumbs up thanks always good information
⭐⭐⭐⭐⭐
Pure GOLD 👍😁
THANK YOU Mr. D.
You are a scholar and a gentleman 🛐
Measuring SMPS high side transistor gate with a passive probe and grounded equipment sometimes reveals amazing IGBT acceleration ... :)
Awesome fundamentals explanation. This is by far the best production quality, and useful content coming from any official manufacturer's YT channel...and I am subscribed to most of them.
I have a couple of suggestions for stepping it up even further. First, it would probably benefit Keysight if you did your demonstration on the most accessible entry level scope when your explaining entry level concepts. Perhaps you could demonstrate part of the example on the smaller screen product. Viewers that want to replicate what you are using can be intimidated by a $4k to $10k+ scope. Also, I understand the difficulties of advertising a specific price on a retail item sold in multiple international markets and across multiple currencies, but entry level tools are mostly a price driven market. If you made me aware of Keysight's entry level scope, and it's functionality, passively, through demos like this, while casually mentioning it's price point, you will get the product on the radar of a lot more potential customers like myself. (My first scope is the DS1052e bc of Dave's hacking videos from back in the day, which is essentially passive marketing like this)
In my opinion, the most useful information you could follow this up with would be to demonstrate a DIY basic differential probe circuit. It may seem counterproductive to sales and marketing at first but that is the kind of upload that would clearly communicate the "we want to build a community culture" I think your going for here. Obviously something like a protoboard circuit is not going to perform as well as a production product that has made it's way to market. So demonstrate that. By showing me the fundamentals and how to build a basic version of a diff probe, your taking the first step to help me by just handing me something mildly useful. Such an effort will impact a viewers perception of the Keysight brand more than any other from of marketing, excluding word of mouth type peer to peer marketing.
That's just my $0.002 adsense on the subject as a simple hobbyist.
-Jake
Thanks, Jake! This is great feedback.
Good video, excellent information glad I found this. In the near future will us this info to look for a differential probe. By the way got a kick of your bacon sayings in the beginning. Funny part is its my last name, Lol.
Haha, thanks!
Love this info, its something that at least at my university isn't shown
Thank you!
exactly!
(also... FULL BRIDGE RECTIFIER!)
Electroboom would be excited about that
FULLLLLL BRRRRRIIDGE RECTIFIIIEEEAAAA!
thanks
can it be used for measuring impedance between two points as in RFID tag antennas
great short video
Very nicely explained.
One day I'm guessing differential, current and High Voltage modes will be built into some scopes, obviously with advance protection.
I even have a dreams of bluetooth LE probes with a built in white led.
I think you are quite on point, seeing how most general purpose scopes have bandwidths above 100mhz, having all inputs be differential (and switchable to 'regular' single ended for better noise performance) would be very interesting.
Great video as always!!
Can you make a video about measuring mains voltage
Good idea!
@@KeysightLabs thank you!!
I feel the need... the need for... differential probes!
Is it better to have the differential probe or a scopes with floating channals?
hello sir, I am trying to measure signals of an Twisted pair ethernet cable to check the difference between signals when the system is connected to internet and when its not connected to the internet?? how can that be done?
There should be a high voltage separation switch that can either connect an input to ground or disconnect it for differential readings
Love you guys. Really nice video.
Btw, could you link the video on the 7:15? Peace.
Here you go! czcams.com/video/a28lVMsWFEY/video.html
What scope are you using in this video, would like to know. Thanks in advance.
This one is a 4000 X-Series scope: www.keysight.com/us/en/products/oscilloscopes/infiniivision-2-4-channel-digital-oscilloscopes/infiniivision-4000-x-series-oscilloscopes.html
Hi, can you Keysight guys tell me how to accurately measure a very high voltage waveform, 30-50 kilovolts peak?
Use high voltage probe
Keysight obviously does not offer Differential Probes? Ask Dave..
Anyhow very nice, entertaining video, very well explained.
We have tons of differential probes! This is the one I had easy access to, but we manufacture many of our differential probes here in Colorado Springs.
Nice video. however not everyone has a bench full of test gear to generate signals. I have a DP10013 probe but haven't used it due to lack of info. I did see one young man plug it into a live power strip. Only trouble was he didn't say if it was plugged into the same power source of the scope or probe!
Yikes!
@@KeysightLabs No wonder not many people learn much from keysight with answers like, "Yikes."
Thanks, that was very well explained. A poor mans solution is to use two channels of the oscilloscope and let the scope calculate the difference.
Just guessing but i think we would use two single ended probes centre line for our connection to the required circuit jnder test and then use channel subtraction to get differential function from two single ended probes?
You definitely can, but there are some caveats.
So why are scopes grounded?
I've played around with an old CRT scope that was floating and everything seemed fine. (Famous last words?) The "ground" inputs were still shorted together obviously so I could make just one differential measurement but it was all I needed
If you are measuring mains or another high voltage circuit with a floating scope and accidentally connected your ground lead to a non-zero voltage reference, then your scope will be at that voltage potential since ground is usually connected to the metal chassis. So then if you touch the chassis, you could act as a part of the circuit, have that high voltage pass through your body towards a lower resistance pathway, and severely damage yourself. Either you conduct the signal through both of your arms, passing through you heart and killing you, or it goes to the ground through your lungs and prevents you from breathing.
There are battery-powered scopes that are left floating like the ones that Tektronix has that might work fine for that purpose, but otherwise this is not a good idea and can easily kill you. You only need a few millivolts to stop your heart.
A good application would be switching power supply troubleshooting. The line side is not ground referenced.
Sir while measuring phase voltage using differential probe (N2791A) it is not showing sine wave and instead it is showing lot of disturbances. Tried to compensate but still sine wave wasn't obtained even though the reading was 230 V in voltmeter. Can you say why is it like this?
I'd recommend contacting our support center at keysight.com/find/contactus and they should be able to help with your specific application
Always use differential probes ;)
So am i right in thinking that with a pair of differential probes, but one of them connected to ground (like an actual rod stuck into the ground), you would measure the absolute voltage, like a "normal" common-mode probe, but without the possibility that your scope will blow up (even though it's grounded). ?
Yep! That's a good point.
@@KeysightLabs Why aren't all probes like this xD
what if i supply the scope with an isolated transformer?
You will get phase offset, based on RL reactance
Scope will be always connected to gnd
My wife checked our bank account going back about 4 years since I took up electronics as a hobby because I have an outdated degree where I majored in search algorithms but went I to policing instead of moving. I've developed mental illnesses such as ptsd amd other related illnesses (no one seems to talk about like amnesia, seizures, insomnia, third party empathy as if it were my own problem, night mares every night and screaming in my sleep, I can't lay in bed and sleep or I scream all night so I sleep sitting up, I could go on). I spent my first 3 years off playing guitar non stop but it got boring after being able to literally play any song (no matter the difficulty, by reading the sheet music). There's a point here I promise.
I started buying kits and gear and supporting content creators hoping to find a way to fit in and even saved over 5k for the local makerspace and also donating guitars I'd buy either new or fix. I've basically sunk 4 years of my life to supporting many different creators at the highest level only to have a falling out with them because of my mental health issues. Most times it wasn't even disputes. The first person I met who runs a learn electronics channel took thousands from me and when I realized he accused me of harassment. I could say more about him but that's not my point. He received a scope from you worth over 5kusd and bragged about getting 1500 for it.
I just want to support people that do good work and learn and have stimulating discourse along with figure out how to entice the ne generation to get into hardware engineering. I post locally in nova Scotia every year before school starts offering free gear amd ul certified meters and such to lighten the load of crippling student loan and I don't have much but I save money and buy spare meters and whatnot to give away. Sometimes I luck out and a friend knows a friend's son who'd love to have a meter so I give them a brymen and cheap scope (amongst other things to get started).
My wife told me I've given over 10k usd to different creators and when I was learning I'd ask for a quick favor or something regarding info to point me in the right direction to learn but everyone always had an excuse ad to why they can't help, one gentleman who is well known got mad and called me stupid even though I supported him after hurricane Fiona destroyed our house and I gave away over half my stuff to have room for clothes. I've only ever wanted the best for everyone I've supported but I was never taken seriously, ignored and even asked to leave channels when asking why won't you help when I support you at a tier which offers it. I rarely asked as I'm very focused. I've gotten to the point where I've had to unsubscribe to every electronics channel where I thought I was a part of something. This profession / hobby gave me hope that ill get to go back to work someday and maybe even take engineering as I have alot of the credits. I paid people to support them but was never offered any back. I have character flaws but I'm a person too. You seem like a nice guy Daniel and I worshipp your brand. I'm not asking for anything. I'm not looking for anything. I just needed a platform to vent as I've never felt so ignored and unappreciated and despite the fact I'm leaving all the creators and this platform for the most part I still have hope. I have to believe in myself. I'm 44 and don't want to be retired the rest of my life and I love engineering which is why I've done all these things. Had I known things would end this way though I'd have an older second hand agilent mso and spectrum analyzer. I've dedicated my life to helping others but dollar signs do nothing to make me visible and feel part of a community I thought was alot more accepting and helpful to each other. I'm gonna miss big clive, dgw and your impression of noise on a dc line. Hahaha peace out man. Maybe someday I'll save up enough to buy a broken SA HAHAHAH take care.
Joey Strong
I wonder how much does it take for me to get the equipment used in the background
A lot of time hunting old bargain shops, ebay, and garage sales!
I got the necessity of differentia probes from ElectroBOOM
3:20 In this case u can use the math function summing the ch 2 & 3, right?
7:14 did you predict this? lol
Yes, it's a pretty common workaround :)
Actually, it has to use the difference function, because it needs to see the differential voltage across ch2 and 3. Also, Daniel says the disadvantages of using two single-ended probes for this at 3:55, like accuracy, mismatch and usage.
What if isolate my oscilloscope's earth from the main's input plug 🤔🤔🤔
I think there are some oscilloscopes that you can use with the standard passiv probe because they are not ground referenced 🤔 thought we had one in our university that students don’t blow up the circuit and oscilloscope 😂
It's not about ground referencing, it's that all channels share the same ground. You can isolate it from the earth but still blow your probe.
Decent overview, but it kind of glosses over the biggest reason I use one... simply for safety. Probing any higher voltage AC power circuit where there is no ground in the circuit (just like as shown in your low voltage example at 2:45), you don't want the scope to become a path to earth ground via its AC line connection should you try and hook up a standard probe's ground (black) clip by accident. It happens all the time. Best case should you do results in a blown breaker, worse case you fry the circuit, probe, scope and/or yourself!
Are you related to Wild Bill Bogdanoff. ?
If you need a diff probe but dont have one, you can use two single ended probes and subtract them (Math Function). Not ideal, but it might be sufficient.
Correct! You can do that, but there are some caveats. You lose resolution, and have to have extremely well-matched probes.
3:57 Using math you will lose a lot of accuracy? Can you explain that with more details.
To me, sounds that at this point, it is digital stuff anyhow, so unless you are subtracting two large numbers to get a small one... and even if you do that, the analog to digital process would have the same … limitations : it won't automatically change the scale, or the number of bits in the digital result because the analog signal is close to "zero", right? So, the granularity of the smaller steps won't change, … unless it is not linear. Oh, it is not linear? Well, I can be wrong, but if you can give more details, I am interested about how we could lose a LOT of accuracy using the math.
I think that I got the answer by myself. It seems to be about the scale and the number of representable values. Say that I have only 2 bits to digitalise a continuous value between 0 and 1, so, I can only get 1/8, 3/8, 5/8 and 7/8 as representable digitalized value, or, the precision is +/- 1/8. So, if I have to us a scale not from 0 to 1 but from 0 to 10, the precision is +/- 10/8 while if I can use a scale from 0 to 2, the precision will be +/- 2/8. In our case, the original signal may force me to use the scale from 0 to 10, so the math chanel will still have the same precision than the original digitalized data, that is, +/- 1.25. But if I can measure directly the analog signal of the difference, it may very well happen that I could use the scale from 0 to 2, and thus, have a precision of 0.25, or 5 times better than what the math "chanel" can deliver, in this scenario. And 5 times better, that is a lot better. That was more or less what you were having in mind at around 3:57 ?
Actually, Why is the output of lab power supplies not isolated from ground/earth?
Safety (electrical shock hazard)?
@@emmanueltrinidad7740 i mean, they only output like 5-12V maybe 20V. Thats not much of a safety issue if you ask me
Is the lab power supply plugged in to the wall outlet? Yes, 5-12V/20V is nothing, but getting exposed to 120/220 VAC will create some serious magic smoke.
Yes but a good PSU is mains transformer isolated and regulated linearly, isS't it? so there shouldn't be any connections to the high voltage side..
Just power the oscilloscope with a battery and inverter
thats what I have with my portable dmm/scope
So basicly you NEED a diffrential probe.
If only you could get a diff probe without selling a kidney...
Sorry, we don't buy kidneys.
@@KeysightLabs 😂 good one. Still, diff probe prices are prohibitive to most of us.
They are definitely more expensive than the standard passive probes, that's for sure.
They can be expensive but some are within reach, depending on bandwidth, under £200 (your currency may vary!) which sounds a lot I know, but if you need to probe switch mode PSUs they are cheaper than a new scope!
Interesting. But that music sucks.
Pathetic, i don't have to use those because -my- *OUR* old analogue oscilloscope made in USSR doesn't even have plug with ground.
any relation to WILD BILL BOGDANOFF?