Our professor made us watch this exact video in my intro to EE class before we were allowed to use oscilloscopes. It feels weird stumbling on this video a few years later.
Wish I knew these things last year. I was attempting to measure the signal at an amplifier output transistor. I quickly learnt about the long ground path through the scope leads to mains earth. Sent 5KW through a 0.33 ohm resistor. The thing glowed bright blue for a fraction of a second and vaporised like it had beamed aboard the Starship bloody Enterprise! Lesson learned.
He didn't mention one point that frighteningly few experts seem to. By attaching your oscilloscope probe-ground to a device powered through an isolation transformer, you once again reference the device to mains earth. It defeats the isolation. It's even worse than using no transformer at all, since the isolation transformer also inherently defeats whatever GFCI/RCD that you might have on your home's wiring. While this method might be okay for lower voltages, using it for 100+ volts is a serious electrocution risk.
I'm certain Dave knows this, just like any expert, and I doubt he intentionally left it out. Instead, it just seems like something that doesn't come to their minds to warn us novices about.
@@Timothymukansi Yes, I would think so. That's probably the best way. But the cheapest way I could think to do it would be to A) always plug your device into an isolation transformer and, at the same time, B) plug your oscilloscope into a GFCI (or another isolation transformer) with a "cheater plug" to break the scope's connection to mains earth. The scope's BNC connectors can only become live in one of two ways--either through the device you are testing or through a short inside the oscilloscope's own wiring. Plugging your device into an isolation transformer would protect you against the former and plugging the scope into a GFCI (or another iso transformer) would protect against the latter. The cheater plug on the scope would prevent the issue of the scope "re-referencing" the device to mains earth through the alligator clip. So long as you do both A and B at the same time, with every device, I think you'd be safe from serious shocks (so long as you didn't grab two parts of the circuit at the same time with each hand). But the differential probe seems more foolproof--if really expensive 😕.
My first oscilloscope should be arriving any minute now. I saw this video years ago and decided it was worth a second look before my scope arrives. I am very glad I watched it. This should be a prerequisite video for all new scope users/owners! Thanks Dave!
But seriously, do they teach anything useful in universities nowadays? I hated when I did my degree as a mature student. I've learned more during previous working years that many teachers knew about electrical principles. This is just a simple example of things you should learn and be aware
Universities, as they should, teach deep knowledge. It's up to you to learn the practical side. That's why you do holiday work, and gain further experience once leaving. They couldn't possibly cover it all in a few years at university.
Electronics are easy to teach and cheap for the university to afford the tools for, but they don't cause they don't give a fuck all they care about is their statics ( at least in here ) , even though I love this field I hated the study and after I graduated and got to work I learned a lot and understood what I have before.
@@flyinghuskey8466 Maybe in the US where universities are basically a scam to get 30-40k out of students per semester. All of this is taught in labs during your first year of electrical engineering in Quebec.
i am an old engineer … So I should know all the tricks, right ? NOPE ! didn't realize the USB port issue … so I agree this is a VERY VERY useful post. I pat a lot of attention to the scenarios and refresh the ideas … very good ..!! Love it.
Thank you so much, have just purchased my first oscilloscope and I am a hobbyist teaching myself. I learn so much from people like you and am grateful you share your valuable knowledge.
This should be lesson #1 in every electronics class !Safety First - thank you for a great explanation. You are Saving many fingers & expensive equipment !
If you need to make a voltage measurement between two points in a circuit, you can use two scope probes that are ground reference, then use subtract measurement on oscilloscope to get the voltage difference. Saves money on differential probes and is a simple and effective solution.
@@gasun1274 Depends on the scope but in general the higher the lowest values of the two probes, the more inaccurate the setup is. For example if your lowest signal is 10V over your mains reference and your other probe is at 11V, then most of the precision of your scope is lost between 0 and 10V. Even worst with 100V for one probe and 101V for the other probe. The difference is still 1V but here you loose 99%+ of the precision of your scope Hope that's clear enough!
A isolated tranformer to power up the load is safe, cheap and more simple. And older electrical circuits there's not grounding terminal, but in modern circuits almost all have the ground reference.
Even years later it is still valid stuff you can learn here. It looks to be really a good idea to watch this first before taking my brand new oscilloscope at risk. Thx Dave for this great lesson.
FANTASTIC video! If I could hit like 100 times I would, I've searched on the internet to no avail for a simple yet complete explanation like this. This covered exactly what I was hoping to learn. Thank You!
Tis is perhaps the funniest clip I seen, any genre. The examples on how to sabotage and get rid of your life and/or equipment are hilariously presented. "BAAAM!, you blew your scope, completely shorted it out" . I can watch this for any years and I did learn a lot. But the commentary is just, insane! Thank You. Made my day.
Thank you from 6 years in the future Dave. I've heard several times how hooking a scope up "incorrectly" can kill it, but that's usually where people leave it, so it's been in the back of my mind "what magical connection will kill my scope and blow my house up?" I've never taken the time to figure out what this scenario is. Assuming there is no other odd scenario, and it just comes down to "don't create a jumper wire with your house ground network," it's common sense, but now that I know what people where talking about, I feel better. I can definitely see someone, including me, not paying attention and doing exactly this (especially with the USB scenario), but there's no hard to identify demon in my circuits waiting to kill me.
Just blew up my Creative Tactics Sigma headset usb adapter... I was measuring with my Rigol DS1054Z the output of a class D amplifier (~50Vpp)... output, input was from mobile phone (floating) and signal was not enough to reach full amplifier potential, so I connected the input to the USB adapter of my headset.... big mistake... Oscilloscope is fine, PC USB port is fine, usb adapter has its case slighly melted on one side, 1 compoent has vaporized along with 1 layer of the PCB, one inductor is burnt and another capacitor seems burnt as well... the 3.5mm jack cable from usb adapter to amplifier was very hot... I knew it, I watched this video before, but I did the damn mistake again... Stupid me... Remembered this video and shared it on Facebook... Mistakes are so easy to make...
My dad taught me this with his Heathkit O-11 oscilloscope decades ago. I haven't used a scope in several years, but that lesson sticks with me to this day. I'm getting back into electronics and a Keytronics scope is being shipped to me, and was curious if the probe grounds are still earth ground - looks like they are. Important topic - you're doing great work!
maybe you already figured this out but there are several options if you want to avoid this, one is a usb scope, and a laptop on battery, another is an isolated usb scope, there are also battery operated scopes now or you could use a differential probe.
It's worth mentioning, even with your floating device which is safe to prob around with the scope as you demonstrate. It can become a problem if you try using two probs and attaching the earth clips at different locations. BANG!!!!
Never ever any professor has ever mentioned about this useful information in my educational period! Its great to know know how to properly use an equipment without fault. Also your way of explaining thing is amazing! Very entertaining and informative! Thank you.
I managed to short through one of the probes of my scope about 15-20 years ago. Still remember the bang! Have been very leery of scopes since then, but have come around to understanding what’s going on. This video clarifies what I have concluded. Might even pull out my scope sometime and see if it is (un)damaged - I never really used it again after that.
My favourite part of this vid is the bangs. Had a brief watch previously but didn’t fully follow. After I bought an old analog oscilloscope and played around for a while I noticed this risk myself and realised Dave may have been talking about the same thing so came back and have another view (with a thumbs up).
Another connection to avoid perhaps would be, when using an isolation transformer on a radio under test and then attaching a coaxial antenna to the rig. In my shack the outer shield of the coax is both grounded for rf and is tied back to the service entrance ground. In this case we have defeated the isolation transformer.
Wow this is a fantastic video! I just bought my first oscilloscope for working on amplifiers, and I was already looking into how not to blow up the new scope. The explanations in this are fantastic and I will be buying a high voltage differential probe ASAP. Thank you!
@@rogerd4559 Huh? If you put a cap in series with the signal, that's just AC coupling. And it doesn't do anything about the ground connection of the alligator clip :)
Dave your #3 mains reference example is the best explanation I've seen for the "non-obvious" situations you can get into while probing what appears to be an isolated circuit. AND, your example when connecting a USB is also a good example of why to isolate your device under test and not your oscilloscope. This video is an oldie but a goodie ;)
I've just seen this and other videos after almost graduating for electrical engineering. No teacher of mine EVER made this important point in any lab i've been in, and i've used an oscilliscope for almost every lab i've every had in engineering. The point being that the ground of the probes are all shorted together so you have to make sure they're all connected to the ground of the circuit. It might remind me of some of the hard lab problems i've had if I hadn't erased them from my memory for my mental health.
While it is absolutely a good idea to explain this to all new electronics students, I would dearly hope that someone graduating with an Electrical Engineering degree would by that point understand the implications of mains referenced test equipment without someone needing to tell them.
@@stargazer7644 By the time I graduated I did, thanks to youtube videos unfortuneately, but no I don't think it's an obvious fact. The most common tool anyone in electronics uses is a multimeter and they come with differential probes. To someone in school there's no reason, if you're not explicitly told, to believe that an oscilliscope probe is any different. I think when you say that by the time an engineer graduates they should know that, you are missing the fact that by the time they graduate someone in one of their labs should have told them. If anything the oscilliscopes should be the reason people know about mains referenced test equipment, as you put it.
@@kevincrowe7832 "Thanks to youtube videos"? Someone with an EE degree should be able to DESIGN this type of test equipment. If someone doesn't even understand how mains powered equipment is safety grounded, how in the world could they design it?
Freshening up on my electronics course back in 1999 for a job interview after being layed off after 25 years and need these types of videos to give me a head up on questions that might come up, thank you!!!
Be careful when considering two outlet plugs automatically as safe, in some countries the ground is connected directly to the neutral/return pin. I highly recommend using a 1:1 transformer to connect any equipment, that's is going to get probed by the scope, to the outlet. This way you don't ever have to worry about this.
Was probing a shunt resistor on the positive rail of a small electric car one time, and I wasn't worried about it because the car is on rubber tires and is battery powered by two car batteries so it should be isolated from mains. Joke was on me though, because here I am sitting in the seat of the car with the scope outsides sitting on the bench, some of the wires across my lap, when the car rolls back a few feet. Wouldn't have been a problem, but it just so happens that part of the shop I was working in has metal paneling on the wall for the first few feet above the floor. SURPRISE, that paneling is grounded, and so is the scope. All the wires between me and the scope light up BANG! The car hadn't had the fibreglass body put on it so the bare metal frame touching the paneling made a dead short across the 24v car batteries. It took me like an hour to figure out why the heck it lit up!
I think 10 times, look at the wires very carefully and then imagine the current flow before finally hooking up the clips of anything to any electrical circuit.
I did exactly the same thing when connected a pc power supply instead the batteries that has also the dc minus connected to the mains earth just to make a simple test and it couldn't got any worst.The cpu was short circuited.
Thank you!!! writing here from Argentina. Now I'm safe to make my electronic measurements! I had the same doubts about the osciloscope and you clarified them! Your diagrams were so explicative.
Yeah, it seems obvious to me; but then stupidity happens (to me, at least). Also, it seems to me that if there's an internal short somewhere in the DUT that leaks current to its chassis ground and you connect it to the scope's ground thru the probe's clip, then you might be in trouble. After all, what are GFI mains devices for? I have yet to draw the circuit in DaveCAD, so I'm not 100% certain.
Better: don't connect it at all unless you need it. If you don't get a signal from the tip alone, you need to reference the device under test to the oscilloscope. Think about that first, don't just clip on. (The thinking is what ultimately saves you.) The clip is also useful at high frequency and to remove noise. That's what it's actually designed for.
Paul F Are you saying that the positive rail of a laptop battery could be connected to mains Earth? I was considering powering my Arduino through my laptop connected to the wall and connecting the alligator clip of the probe to Arduino ground therefore Earth. Should I use a floating power supply instead to guarantee safety?
I could never understand why would anyone give a thumb down to such a great video ? what the hell if you didn't want to watch just go on and live your depressed life Bang!!!!!!!!!Bang!!!!!!!!!!Bang!!!!!!!!!!!!....coz this guy is Awesome! Bang!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
You are the boss and deserve a lot of success for your teaching, your passion is contagious and you make it really easy to understand complex stuff. I am learning my chops these days, so i am pretty sure that i will see a lot more of your videos. kudos :)
THANK YOU so much for your video blog! Makes this here newbie a little more confident. You're doing such a great service to the hobbyist community. Your infectious enthusiasm doesn't hurt a bit either.
Is that one of those deals where you have to have the washing machine on and paused with a metal zip stuck in the mechanism before the toaster will work
Thank you very much for pointing this out! I instantly stood up, got my multimeter and measured my laptop's power supply. 0.8 ohms! Also I measured between the earth pin and the USB shield, and it turned out: A dead short! I've been probing and soldering stuff connected to my computers for years an had no idea!
but earth pin you mesured but there is no connection between main and earth right? So how can you actually make a short circuit? Considering also that there is a transofrmer in the power supply which isolate the rest of the circuits, ok there is ground but still I don't see any point that the circuit could be shorted. Maybe I am missing something?
@@grzesiek1x You're missing that if there's a fault, connecting a mains referenced circuit to an earthed GND can make a dead short. It seems that the comment I originally replied to eight years ago is gone.
@@nrdesign1991 hmmm 🤔 but if there is a fault in the transformer I could be shocked even without oscilloscope (I mean probing with the scope it wouldn't change the situation because I could be shocked even connecting a diode) ? Another thing is that my USB port and also headphones port have 1 M Ohm of resistance between main's ground and them... I ask because I don't follow emotions in electronics but facts and science is important to me.
Interesting idea! -I believe he was using an analog scope in the video.- Edit: upon further inspection, it seems possible. His scope has the capability to invert CH2, and display the two signals added together! Also, fancy seeing you here! I enjoy your channel. :)
Yes you can use 2 channels. Usually the scope will have either an invert and add or subtract option for combining signals that will measure the voltage difference between the two probes like a balanced input. This is actually a very good way of making accurate measurements because you reject most if not all the common mode noise, ground loop problems, and no more shorting your ground reference, just leave your ground clip unhooked. However, you can end up using a lot of inputs this way. Your 4 channel scope has now turned into a 2 channel, which if you are reading a couple of inputs and need a reference clock to trigger off of, can be a bit of an issue.
A wire that i tried to use to measure a signal vaporized as a result of my failure to isolate my scope. Good thing nothing was harmed! My scope continues to work...
Hello from Russia, Dave! I like your videos :) Well, I'm not sure about "completely safe" power supplies with 2 pins. Sometimes capacitance between primaary and secondary windings of transformers plays role (I've read some years ago about multicontroller-based device which was unstable with transformer PSU but stable with PWM PSU which has less parasitic capacitance); sometimes there's RC filter between primary and secondary parts of PSU. From other side, there could be Y-capacitor in PWM PSU of your scope, so if you connect your scope's alligator clip to 2-pins supply's "hot" part of circuit, there could be parasitic AC-connection which at least can damage your scope.
Dave, you must know that the current carrying neutral of your mains supply is bonded to the protective earth. If you don't chech that neutral for continuity with the CUT ground rail you still blowup. CUT to protective earth is not the only risk in correctly bonded mains circuits. This is a life threatening problem in tube circuits with B+ running in the hundreds of volts. You need another video about chassis grounding. I know two people injured by this, one by heart attack.
I apologize for asking this simple question (to you guys) but I really need the answer. Crocellian really brought into focus my concern. Most of our houses in New England have three wires brought into their homes. one is 120 vac, one is 120 vac but phased shifted 180 degrees to permit 220 VAC to form and finally a "ground". The problem that I am having is understanding what happens with a 120 VAC circuit. One wire IS hooked to the ground. Caling that wire "neutral" seems erroneous. Having a separate ground plug (the round one) seems superfluous. AND it seems like I exposing ANY circuity powered by AC to earth ground exposing myself to the problem David talks about. So what do i do to attempt to create isolation? I can't really see a way unless I put an isolation transformer between the device I want to troubleshoot and the AC power source. Please help me.
william burling the ground wire is used as a safety ground. Referencing the chassis to one conductor, which is named Neutral, provides safety so if the line shorts to the chassis it will open a breaker. Otherwise it could be possible for the wash machine, for instance, have one conductor short to its chassis, and maybe the clothes dryer have its chassis shorted to the other conductor. Motors in each could became damaged and get a wire shorted to the chassis. This could electrocute a person touching each machine. So purposely referencing the chassis to the same conductor creates a safe environment.
@@KissAnalog I think what he's getting at is 'Neutral is bonded to mains earth back at the main breaker panel of a house. So, at least for American homes, seeing only two wires (hot and neutral) spade going into an AC adapter doesn't necessarily mean you are isolated. I learned that really quick about 30 years ago...don't ask how!
@@williamburling3229 Our system is called "TN-C-S grounding, single phase, with grounded center tap." In the house wiring and appliances, neutral is an electrical system wire. It is allowed to carry current in normal operation. It isn't touch-safe and it must be insulated. Code calls it the "grounded conductor" (groundED). *The neutral wire can develop several volts of AC* difference along its length during normal operation, simply because it has non-zero resistance. This means one neutral can have several volts of potential relative to nearby other neutrals or the grounding wires. It's not touch-safe or equivalent to the grounding wire. Another reason why they're not equivalent is that a GFCI outlet uses a small common-mode test voltage L&N vs G to probe for N-G continuity. This detects (and trips) if a neutral wire has rubbed or pinched against a grounded part. Despite the non-equivalence, some old work (WWII in particular) was allowed to run a common to some 240/120 appliances. This is simply less safe than a separate grounding conductor. The grounding wire is touch-safe. It can be connected in strange ways - if two appliances share plumbing or a mechanical connection, or through a metal building frame, their grounding wires will not be insulated from each other. It doesn't normally carry current. (It will carry a small amount of current due to parasitic capacitance, static electricity, etc.) Code calls this the "equipment grounding conductor" (groundING) The distribution system is different. It uses a multiply grounded power conductor. Utility electricians usually call it the "ground" wire, but I think "common" is more clear. It's touch-safe IF you only touch it at one point. It's bolted directly to utility poles without an insulator. Safety is ensured by balancing the distribution circuit and making this common wire surprisingly large on the poles. (It's at least as thick as the hots even though it is multiply grounded and carries less current.) These conflicting systems are reconciled by connecting all three of those conductor types together at one and only one point, which code calls the "service equipment." This one point is touch-safe: it touches the utility's distribution circuit at one point. The grounding conductors are at the same potential as long as they don't carry significant current. The neutral does carry significant current, so it's not touch-safe. Grounding conductors from a neighbor are generally within a few volts. Still touch-safe (except when swimming pools are involved) but if you run a wire long enough you can find a dangerous disagreement in "ground" potential. Regular earth electrodes helps a little with this problem, but they are more of a lightning-protection measure. Mostly it's not a problem because people don't run long conductors unless they are a telco or electric utility. Many old electronic devices had a "hot chassis" - connected to hot or to neutral. Neutral through a polarized plug is a little safer than fully hot. All those devices need an insulated case for safety, and they can't be probed unless they're powered through an isolation transformer. New electronics are either grounded-chassis or floating-chassis. Any low-voltage power system inside the device should be galvanically isolated from both the hot and the neutral. The low voltage system will be referenced to the chassis and might be floating or grounded relative to the environment. Modern AV equipment is often isolated from the AC input and grounded through its connections - the plug is polarized only because correct polarity reduces hum. (Quite a lot of cheap crap doesn't obey those isolation rules.) Grounded-chassis can be probed, but the reference clip can only go to a signal ground. The clip is not required for low frequency, low precision. Floating-chassis can be probed. It's a good idea to only reference clip to signal ground, but you can often get away with differential measurements. I think it's still a bad idea. If an isolation transformer is set up with a floating secondary, you can do the same thing. An isolation transformer with a floating secondary is a code violation; you're not allowed to use it as permanent wiring. In all cases the secondary-side groundING conductor must be connected to the secondary groundED (neutral) tap. That makes a floating secondary. A grounded secondary also connects them to the primary groundING conductor. Figure 2 is a good schematic of what's required for a permanent-service transformer in the US. electrical-engineering-portal.com/purpose-of-shielded-isolation-transformer
Jordan Rodrigues. Hey thanks, I’ve actually read all of that! Bunch of good info. It does get confusing for guys who work on old tube equipment. Whether or not to float scopes is usually the subject of debate. Many have done it for decades and are still alive to tell you why they do it. This is what I’ve been told.... They give up safety for lower noise. Old tube amps that are hot chassis are noisy because the caps and resistors don’t have a earth ground to drain into. So floating their scopes keeps the noise down because the amp circuit isn’t trying to drain through the scope ground. I don’t know if that’s the absolute truth of why they float scopes, but that’s what I’ve been told. I’d rather use a isolated transformer on the old tube amps and keep my scope grounded. All I’ve ever done is work on modern tube amps with chassis earth ground, so I have never worried about blowing up anything in my hand while my other hand is in my back pocket.
that's why i always use isolated differential probes whenever i am working with high voltage. Can't be so smart and aware all the time. Great video and explanation.
@ 4:06 great Back to the future reference! I love your videos.They have helped explain quite a few concepts when I get confused/stuck. Keep being awesome
isso aconteceu com minha equipe de manutenção eletrônica, o terra do osciloscopio foi usado junto com a ponta de prova para medir uma tensão em um tiristor numa retificação trifasica de 380VAC, adivinhe? a manlha da sonda ( terra) esquentou devido ao curto circuito e começou a derreter o cabo ! quase causando um incendio! isso foi uma impericia de um colega desatento! grato por seu video !!!
At 13:36, I believe what is missing in the video is to mention that if you are uncertain if connecting the black probe of the osciloscope a certain point of the circuit is going to be a problem or not, a multimeter can be used to check the voltage difference between the ground of the osciloscope and the connect point on the circuit. The voltage drop between the two spots should be 0 essentially. E.g in a valve amplifier, this might be the bolt connected to the chassis of the amplifier.
This is a very useful video. I did a Summer job as an intern with RCA in the early '80s. I was fixing broken switching mode power supplies used in studio cameras. I erroneously connected my 'scope ground to a 400V line and it DID ruin my day. Gratefully RCA had given me a big transparent box to keep the PSUs, in while they were being tested, and that took most of the flak. I also learnt that fuses are good !
I always thought that the _ground leads_ on a scope should be protected with 0.1 amp quick-blow fuses- but that's a little too obvious a solution- isn't it?
+Bob Bowie The proper solution would be for scopes to follow the principles of almost all other tools and test equipment, and be built to 'Class 2' double insulated standards, where an earth is not required. The evidence is overwhelming that Class 2 equipment has a far better safety record than Class 1 (single insulated and earthed) equipment. How often do you see drills with an earth these days? For some reason the scope manufacturers have always dug their heels in at this suggestion, and they are about the last of the Mohicans to still be building to Class 1 standards.
+ReactanceIsFutile Tektronix did produce an add-on to do that, only problem was that if the voltage to ground exceeded 50v it automatically reconnected the ground. It actually created a worst-case scenario because an ohmmeter check would suggest an earth-free probe, but the probe shield would actually short-out if it came in contact with high voltage. It could also kill the operator by providing an (unexpected) shock path to ground. Basic issue is that old-school electricians place an excessive reliance on earthing. Even in the consumer environment it is now being recognised that earthed metal is as much of a hazard as it is a safety device, because it can act as the return path for a shock from another voltage source. This is especially so on the testbench. The best arrangement is to have no exposed metal which could become live.
Good information thank you. My solution for using scope with mains circuits is to isolate the scope by plugging it into a "portable power station" (Jackery 1000) which is basically a battery and inverter.
Really simple good solution. These days there are smaller capacities that don't cost much. Because the problem is that in many places in the house in the electrical box the ground is connected to the minus.
Another big pitfall is connecting ground to the DC(-) side of a full wave rectifier. It's not obvious at first, but if you aren't using a transformer (ie. your rectifier is connected directly to mains) then connecting your scope's ground to the (-) output puts the ground lead and one one of rectifier diodes in series with mains voltage which will quickly destroy said ground lead and diode. The reason for this is that mains neutral and ground are connected inside the breaker box (that's how it is in the US anyways, might be different elsewhere). EDIT: Why is a portion of my comment crossed out? No matter, I'll put up a video later today.
This is called a "Hot Chassis". Many power supplies are built this way. In TVs to help protect service technicians there will be a silkscreen border line printed on the circuit board to warn you not to touch or connect ground to that part of the circuit. I work mostly with power equipment like inverters so I always use my portable isolated scope, the Tektronix THS720A. The scope is fully isolated and the grounds on the probes are not connected together so you can measure just about anything safely. You can measure a hot chassis with no problem with this scope.
This is one example where some people have a massive experience step! They either discover this the hard way if they aren’t too savvy or hopefully they will see your demo and think about this. When I was a young trainee engineer I was on an insulated mate and working on a mains circuit I had somehow rested my hand on a live conductor and because I was sitting on a padded seat I didn’t feel anything....that was until I picked up the scope lead with my right hand and went to attach it to the scope.... OOOH FCUK I shouted ...this scope chassis is live! Potentially the worst kind of shock you can get, across the chest through both arms. Good demo mate as always!
Got a Big Bang yesterday, luckily blew the breaker, I was clueless. And afraid to try again. Can’t wait to get back in the shop with my new found knowledge. Thank you so much!!!
For many years I have seen some people cutting the ground pin off on the scope. I never did that on mine but have just being careful about grounds. When not sure I have a plug adapter without ground. Grounds can be tricky. Thanks for the video. I have seen actually people getting shocked too. Good lesson to keep in mind.
Thanks! Great information that I did not learn in engineer school! Maybe was mentioned once in a lab to keep people from blowing up oscilloscope leads. Seeing this now, I actually think that happened to someone.
19:24 the computer did not reboot the usb hub disconnected the Arduino for drawing more current than 0.5 amps. The sound we hear is the sound of plugging/unplugging something into a usb port.
You will still charge the grounds on the scope and charge them with potential to zap you or short if the potential is not same. The ground is there to protect you not the equipment.
Very educational video about ground, earth and chassis applications. I couldn't find a video that explains the difference between those and I ended up finding it when watching this oscilloscope video.
Thank you very much Dave! Again, i watched this video and learned more than in school for 6months. Your way(approach) of teaching is completely opposite from the way how they teach us in schools. It is totally interactive and makes you ask and wonder.
Thank you very much sir ! Finally, your explanations are crystal clear ! I've seen other videos and NONE could explain it properly to a beginner like me... I understand now how I can damage the oscilloscope I've just bought recently, probing without care... I choose a wall outlet that's earth grounded for the oscilloscope and, by pure chance, the power supplies I use to power up my breadboard (one ATX SMPS and one wall-wart) are plugged in wall outlets that aren't earth grounded (no earth hole). Now I know it would be risky if my power supplies were also earth grounded...
So does this mean I can connect one probe's ground lead to a local ground on the circuit and leave the others unconnected, and they'll all reference the same ground voltage? If so, this makes connecting probes a lot easier!
+Paul Wayper Normally a probe's leads have their ground tied together inside the scope. So if you have one ground lead connected, all probes are now referenced to whatever the ground lead is connected to.
Electrically yes, electronically NO. Try doing that with mixed audio or high frequency signals and you can EASILY get into a world of bother with incorrectly shielded signals, bad return paths and/or earth loops. The main thing as in ALL electronic and electrical measurement is to properly understand WHAT you are doing, WHY you are doing it and HOW to do it correctly! SOMETIMES it may be necessary to use MULTIPLE different scopes. As an example, if you have two scope leads in an RF environment and only ONE signal ground connection back to the scope how can you be certain that currents flowing down the common shield are not interfering with currents flowing down the SAME shield from the second signal source? Or producing a skewed voltage reading due to tiny increases in voltage drop down the shielding? They are BOTH shielded but there is only ONE common signal return path. There has to be a return path but SHOULD it be down ONE shield only? If BOTH shields are connected at the scope (they almost always will be if only via chassis ground) are you creating an RF loop antenna with the two shielded cables providing a complete loop which can massively increase common mode noise? It's NOT as straightforward to get this stuff right in every circumstance as people may think! One reason why GOOD differential scope probes cost so much - it's not just the design but the precision of manufacturing for what will be a VERY SENSITIVE piece of measuring equipment.
Something I learned by looking at tv schematics is that although ground is almost always zero volts, don't ever just assume it is. Besides faults and other reasons that a ground is supposed to be zero isn't, sometimes equipment is designed with a specific voltage as ground. For example, many General Electric CRT TVs have 70V rail marked as ground throughout and chassis ground is earthed.
I think the biggest innocent mistake folks make in the beginning comes from using a multimeter as our first real piece of test gear where we have two leads/connections to hook to a circuit to test, usually a red lead and a black lead. The color of the lead connections doesn't really matter, it's the muscle-memory of making two connections to a circuit/component to make the measurement ("I must have to hook this up somewhere so here goes"). Once we become adept at using a multimeter, that muscle-memory tells us that with a scope probe, we still have two connections to make to the circuit/component that we want to test, so we connect the alligator clip to the other side of the component or IC or whatever by habit of ....connecting two leads to make a "proper measurement" and that's when the smoke happens and we are left wondering why the piece of gear we were trying to repair got worse or usually sparked and let the smoke out. I've always determined what the actual true ground of the device-under-test (DUT) is and then hooked my scope-probe alligator clip to that point only and using only one of the gator clips if using more than one oscilloscope probe (the others are unnecessary after the first). Thanks again Dave :)
So as a general rule, all I have to do is make sure the alligator clip is connected to circuit ground and I'm free to probe anywhere I want within the circuit, correct? Doing this the only thing you lose is the ability to reference the probed point against something other then ground.
Yes, for which you need to use differential-probe or 2-channels with channel1 - channel2 math operation on, most digital have this (old analog have + and invert, channle1 + invert( channel 2).
I was asked about this in an electronics interview today and had no idea about it.. Man i must have looked stupid to the 3 engineers that interviewed me :(((
Holy moly mate, I was about to probe a laptop with a similar power supply. You've just saved another noob! Thought I'd let you know... Thanks for sharing your knowledge!
Wow, am I glad I watched this video! I've started building tube powered guitar amps, and the high voltage rail can often be around 500VDC. Shorting 500VDC to ground would've been very unpleasant. Thanks Dave!
Rangy Tang similar to what happened to an acquaintance of mine he somehow got his guitar plug across the mains plug while rearranging his studio . His story never made sense but his amp took the hit the PC tracks got trashed. Luckily for him the mfr. took pity on his situation and supplied him the schematics. His luck held out and the majority of the repair was laying new tracks and 1 power transistor. Ensonified my row home neighborhood with 200 watts of mr. BoJangles. Out of his stack.
Isolate something - either the device under test or the test instrument. If the device under test is the whole electrical grid, you've got one obvious choice...
Afaik ground is connected to neutral somewhere in the grid. So you might get away with just leaving the earth clip dangling loose. I never tried it though, you need quite some faith in your science to put your 500$+ scope at potential risk. I would at least put a transformer between it with like 5v output or so.
Thank you so much Dave that you took time to make sure people pay attention to this obscure problem of ground isolation. I have fried stuff before years ago and could not figure out why.
+BenjaminGoose Isolation transformer DOES NOT equal a battery, but is close. The isolation transformer capacitively couples some line current to the output. Depending on what you are probing the coupled current may or may not matter.
I used to work in a big electronic workshop to repair industrial equipment, to avoid errors, all oscilloscopes had their main earth connection removed inside it. Not correct I know, but with so many different technicians using the same instruments it was the final solution that solved the issue of randomly fried scopes.... even with a lot of explanation, people did mistakes and blow up the probes from time to time until the earth connection was removed and the workshop had peace.
Our professor made us watch this exact video in my intro to EE class before we were allowed to use oscilloscopes. It feels weird stumbling on this video a few years later.
your professor was smart enough to know that dave was better than him by a ton
@@roseelectronics4582 What an odd thing to say.
@@roseelectronics4582 become his toilet paper then.
You undermined that professor without knowing anything, so yeah that's what you're- a toilet paper
@@ItsBoyRed you absolutely hit the nail on the head there :D
@@roseelectronics4582 it ia always wiser to admire some bette than criticize
Wish I knew these things last year. I was attempting to measure the signal at an amplifier output transistor. I quickly learnt about the long ground path through the scope leads to mains earth. Sent 5KW through a 0.33 ohm resistor. The thing glowed bright blue for a fraction of a second and vaporised like it had beamed aboard the Starship bloody Enterprise! Lesson learned.
Very useful. As a newbie, I see a 99% chance you saved my scope. Thanks for that.
I am also sure that you’ve saved my oscilloscope.
I'm pretty sure this video is the reason why my Rigol MSO1104Z-SPlus is still alive as well... :-) VERY USEFUL (and money saving, probably).
I am also sure that you’ve saved mine! Thanks
+1
And I'm thankful to still be alive, had the probe in my mouth and on my head looking for brainwaves. Smh.
9 years later and it's still an incredibly educational video. Thank you.
He didn't mention one point that frighteningly few experts seem to.
By attaching your oscilloscope probe-ground to a device powered through an isolation transformer, you once again reference the device to mains earth. It defeats the isolation. It's even worse than using no transformer at all, since the isolation transformer also inherently defeats whatever GFCI/RCD that you might have on your home's wiring.
While this method might be okay for lower voltages, using it for 100+ volts is a serious electrocution risk.
I'm certain Dave knows this, just like any expert, and I doubt he intentionally left it out. Instead, it just seems like something that doesn't come to their minds to warn us novices about.
@@gloomyblackfur399 would a differential probe negate that effect?
@@Timothymukansi Yes, I would think so. That's probably the best way.
But the cheapest way I could think to do it would be to A) always plug your device into an isolation transformer and, at the same time, B) plug your oscilloscope into a GFCI (or another isolation transformer) with a "cheater plug" to break the scope's connection to mains earth.
The scope's BNC connectors can only become live in one of two ways--either through the device you are testing or through a short inside the oscilloscope's own wiring. Plugging your device into an isolation transformer would protect you against the former and plugging the scope into a GFCI (or another iso transformer) would protect against the latter. The cheater plug on the scope would prevent the issue of the scope "re-referencing" the device to mains earth through the alligator clip.
So long as you do both A and B at the same time, with every device, I think you'd be safe from serious shocks (so long as you didn't grab two parts of the circuit at the same time with each hand). But the differential probe seems more foolproof--if really expensive 😕.
My first oscilloscope should be arriving any minute now. I saw this video years ago and decided it was worth a second look before my scope arrives. I am very glad I watched it. This should be a prerequisite video for all new scope users/owners! Thanks Dave!
The main trap for beginners is the USB powered devices that are so popular...
@@erne75 can you elaborate for a total newb?
@@gokblok its in the video lmao
A lot of regular guys here will know this but for a beginner like me, this is golden safety information. Thanks.
Learned more in 20 minutes here then 3 months in class, lol. Thank you.
But seriously, do they teach anything useful in universities nowadays? I hated when I did my degree as a mature student. I've learned more during previous working years that many teachers knew about electrical principles.
This is just a simple example of things you should learn and be aware
Universities, as they should, teach deep knowledge. It's up to you to learn the practical side. That's why you do holiday work, and gain further experience once leaving. They couldn't possibly cover it all in a few years at university.
No they don't. EE in school now is just math and theory. Even that, profs still can't teach shit but you have to learn everything by yourself.
Electronics are easy to teach and cheap for the university to afford the tools for, but they don't cause they don't give a fuck all they care about is their statics ( at least in here ) , even though I love this field I hated the study and after I graduated and got to work I learned a lot and understood what I have before.
@@flyinghuskey8466 Maybe in the US where universities are basically a scam to get 30-40k out of students per semester. All of this is taught in labs during your first year of electrical engineering in Quebec.
This has got to be one of the BEST videos that you have ever done, and it is over 7 years old, (It's June 2019 Now) THANKS.
i am an old engineer … So I should know all the tricks, right ?
NOPE ! didn't realize the USB port issue …
so I agree this is a VERY VERY useful post. I pat a lot of attention to the scenarios and refresh the ideas …
very good ..!! Love it.
Thank you so much, have just purchased my first oscilloscope and I am a hobbyist teaching myself. I learn so much from people like you and am grateful you share your valuable knowledge.
This should be lesson #1 in every electronics class !Safety First - thank you for a great explanation. You are Saving many fingers & expensive equipment !
This was the best 24 minutes and 23 seconds of Internet time have ever spent.. Thank You for this!
If you need to make a voltage measurement between two points in a circuit, you can use two scope probes that are ground reference, then use subtract measurement on oscilloscope to get the voltage difference. Saves money on differential probes and is a simple and effective solution.
That's a cheap hack and useful in most situmakations however you lose out on precision if you're logging data and not just looking at waveforms.
I second that. It leaves me wondering why Dave did not point that out. Not everyone has 300$ layin’ around to buy one of them isolation probes.
@@padmanabhaprasannasimha5385 usually how much precision is lost using that setup?
@@gasun1274 Depends on the scope but in general the higher the lowest values of the two probes, the more inaccurate the setup is.
For example if your lowest signal is 10V over your mains reference and your other probe is at 11V, then most of the precision of your scope is lost between 0 and 10V.
Even worst with 100V for one probe and 101V for the other probe. The difference is still 1V but here you loose 99%+ of the precision of your scope
Hope that's clear enough!
A isolated tranformer to power up the load is safe, cheap and more simple.
And older electrical circuits there's not grounding terminal, but in modern circuits almost all have the ground reference.
Thank you for this, about to buy my first scope, a DS1054Z, and busy learning the basics. Material like this is hugely valuable! Kudos.
Even years later it is still valid stuff you can learn here. It looks to be really a good idea to watch this first before taking my brand new oscilloscope at risk. Thx Dave for this great lesson.
12 years later, and still the best video to explain this. Very very very nice video.
FANTASTIC video! If I could hit like 100 times I would, I've searched on the internet to no avail for a simple yet complete explanation like this. This covered exactly what I was hoping to learn. Thank You!
Tis is perhaps the funniest clip I seen, any genre. The examples on how to sabotage and get rid of your life and/or equipment are hilariously presented. "BAAAM!, you blew your scope, completely shorted it out" . I can watch this for any years and I did learn a lot. But the commentary is just, insane! Thank You. Made my day.
Thank you from 6 years in the future Dave. I've heard several times how hooking a scope up "incorrectly" can kill it, but that's usually where people leave it, so it's been in the back of my mind "what magical connection will kill my scope and blow my house up?" I've never taken the time to figure out what this scenario is. Assuming there is no other odd scenario, and it just comes down to "don't create a jumper wire with your house ground network," it's common sense, but now that I know what people where talking about, I feel better. I can definitely see someone, including me, not paying attention and doing exactly this (especially with the USB scenario), but there's no hard to identify demon in my circuits waiting to kill me.
How educative it is. Vividly shown floating circuits versus grounded ones. Even though the film is 7 years old - the knowledge is timeless. Thank you.
Just blew up my Creative Tactics Sigma headset usb adapter... I was measuring with my Rigol DS1054Z the output of a class D amplifier (~50Vpp)... output, input was from mobile phone (floating) and signal was not enough to reach full amplifier potential, so I connected the input to the USB adapter of my headset.... big mistake... Oscilloscope is fine, PC USB port is fine, usb adapter has its case slighly melted on one side, 1 compoent has vaporized along with 1 layer of the PCB, one inductor is burnt and another capacitor seems burnt as well... the 3.5mm jack cable from usb adapter to amplifier was very hot... I knew it, I watched this video before, but I did the damn mistake again... Stupid me... Remembered this video and shared it on Facebook... Mistakes are so easy to make...
We all need the source of these elusive and mysterious isolation transformers that everyone speaks of
@@rogerd4559 and disable an important safety feature... There's a reason why the bnc is earth referenced.
@@rogerd4559 no, buy a differential probe and stay alive
@@rogerd4559 Fine, but the isolation transformer should be used to isolate the device being tested, not the oscilloscope!
I'm happy to see you explain this with kindness! Excellent video!
My dad taught me this with his Heathkit O-11 oscilloscope decades ago. I haven't used a scope in several years, but that lesson sticks with me to this day. I'm getting back into electronics and a Keytronics scope is being shipped to me, and was curious if the probe grounds are still earth ground - looks like they are. Important topic - you're doing great work!
maybe you already figured this out but there are several options if you want to avoid this, one is a usb scope, and a laptop on battery, another is an isolated usb scope, there are also battery operated scopes now or you could use a differential probe.
Very late comment, but this is still now, one of the most informative videos out there..!!
It's worth mentioning, even with your floating device which is safe to prob around with the scope as you demonstrate. It can become a problem if you try using two probs and attaching the earth clips at different locations. BANG!!!!
A good reminder
Never ever any professor has ever mentioned about this useful information in my educational period! Its great to know know how to properly use an equipment without fault. Also your way of explaining thing is amazing! Very entertaining and informative! Thank you.
I managed to short through one of the probes of my scope about 15-20 years ago. Still remember the bang! Have been very leery of scopes since then, but have come around to understanding what’s going on. This video clarifies what I have concluded. Might even pull out my scope sometime and see if it is (un)damaged - I never really used it again after that.
My favourite part of this vid is the bangs. Had a brief watch previously but didn’t fully follow. After I bought an old analog oscilloscope and played around for a while I noticed this risk myself and realised Dave may have been talking about the same thing so came back and have another view (with a thumbs up).
Another connection to avoid perhaps would be, when using an isolation transformer on a radio under test and then attaching a coaxial antenna to the rig. In my shack the outer shield of the coax is both grounded for rf and is tied back to the service entrance ground. In this case we have defeated the isolation transformer.
Wow this is a fantastic video! I just bought my first oscilloscope for working on amplifiers, and I was already looking into how not to blow up the new scope. The explanations in this are fantastic and I will be buying a high voltage differential probe ASAP. Thank you!
I just got my first oscilloscope and I did not know this. You may have saved me a bad day. Thanks.
dont most probes use an isolation capacitor in series?
@@rogerd4559 Huh? If you put a cap in series with the signal, that's just AC coupling. And it doesn't do anything about the ground connection of the alligator clip :)
Dave your #3 mains reference example is the best explanation I've seen for the "non-obvious" situations you can get into while probing what appears to be an isolated circuit. AND, your example when connecting a USB is also a good example of why to isolate your device under test and not your oscilloscope. This video is an oldie but a goodie ;)
I've just seen this and other videos after almost graduating for electrical engineering. No teacher of mine EVER made this important point in any lab i've been in, and i've used an oscilliscope for almost every lab i've every had in engineering. The point being that the ground of the probes are all shorted together so you have to make sure they're all connected to the ground of the circuit. It might remind me of some of the hard lab problems i've had if I hadn't erased them from my memory for my mental health.
While it is absolutely a good idea to explain this to all new electronics students, I would dearly hope that someone graduating with an Electrical Engineering degree would by that point understand the implications of mains referenced test equipment without someone needing to tell them.
@@stargazer7644 By the time I graduated I did, thanks to youtube videos unfortuneately, but no I don't think it's an obvious fact. The most common tool anyone in electronics uses is a multimeter and they come with differential probes. To someone in school there's no reason, if you're not explicitly told, to believe that an oscilliscope probe is any different.
I think when you say that by the time an engineer graduates they should know that, you are missing the fact that by the time they graduate someone in one of their labs should have told them. If anything the oscilliscopes should be the reason people know about mains referenced test equipment, as you put it.
@@kevincrowe7832 "Thanks to youtube videos"? Someone with an EE degree should be able to DESIGN this type of test equipment. If someone doesn't even understand how mains powered equipment is safety grounded, how in the world could they design it?
Freshening up on my electronics course back in 1999 for a job interview after being layed off after 25 years and need these types of videos to give me a head up on questions that might come up, thank you!!!
Be careful when considering two outlet plugs automatically as safe, in some countries the ground is connected directly to the neutral/return pin. I highly recommend using a 1:1 transformer to connect any equipment, that's is going to get probed by the scope, to the outlet. This way you don't ever have to worry about this.
Just proves that a little knowledge is dangerous. As a newbie to scopes this was a great foundation course. Thanks muchly.
Was probing a shunt resistor on the positive rail of a small electric car one time, and I wasn't worried about it because the car is on rubber tires and is battery powered by two car batteries so it should be isolated from mains.
Joke was on me though, because here I am sitting in the seat of the car with the scope outsides sitting on the bench, some of the wires across my lap, when the car rolls back a few feet. Wouldn't have been a problem, but it just so happens that part of the shop I was working in has metal paneling on the wall for the first few feet above the floor.
SURPRISE, that paneling is grounded, and so is the scope. All the wires between me and the scope light up BANG! The car hadn't had the fibreglass body put on it so the bare metal frame touching the paneling made a dead short across the 24v car batteries.
It took me like an hour to figure out why the heck it lit up!
Ouch.
That is kind of hilarious, as terrible as it must've been!
Step #1: Go change shorts.
Step #2: Come back to figure out WTF just happened :)
@Janardan S yep
I think 10 times, look at the wires very carefully and then imagine the current flow before finally hooking up the clips of anything to any electrical circuit.
Dave, back to basics, thanks for keeping us all grounded
"Please excuse the crudity of this model, I didn't have time to build it to scale or to paint it…" haha you are the best teacher around !
Yeah.... But even then, he got the symbology wrong!
Doc Brown ref. Lol
Great Scott!
I watch this video every couple of years. As a hobbyist I rarely use my scope, so I need the safety refresher.
Brilliant explanation ! At last I understand why I blow up an UPS with a single oscilloscope probe contact.
I did exactly the same thing when connected a pc power supply instead the batteries that has also the dc minus connected to the mains earth just to make a simple test and it couldn't got any worst.The cpu was short circuited.
Thank you!!! writing here from Argentina. Now I'm safe to make my electronic measurements! I had the same doubts about the osciloscope and you clarified them! Your diagrams were so explicative.
just don't connect the alligator clip to any voltage point except the ground will do.
Yeah, it seems obvious to me; but then stupidity happens (to me, at least). Also, it seems to me that if there's an internal short somewhere in the DUT that leaks current to its chassis ground and you connect it to the scope's ground thru the probe's clip, then you might be in trouble. After all, what are GFI mains devices for? I have yet to draw the circuit in DaveCAD, so I'm not 100% certain.
...but as Dave said, in rare occasions the device's power pack can be positive-rail referenced. So you need to check this first.
Doesn’t work when you have to work with differential protocols, but it’s certainly a good general rule (after verifying where ground actually is)
Better: don't connect it at all unless you need it.
If you don't get a signal from the tip alone, you need to reference the device under test to the oscilloscope. Think about that first, don't just clip on. (The thinking is what ultimately saves you.)
The clip is also useful at high frequency and to remove noise. That's what it's actually designed for.
Paul F Are you saying that the positive rail of a laptop battery could be connected to mains Earth? I was considering powering my Arduino through my laptop connected to the wall and connecting the alligator clip of the probe to Arduino ground therefore Earth. Should I use a floating power supply instead to guarantee safety?
I could never understand why would anyone give a thumb down to such a great video ? what the hell if you didn't want to watch just go on and live your depressed life Bang!!!!!!!!!Bang!!!!!!!!!!Bang!!!!!!!!!!!!....coz this guy is Awesome! Bang!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
You are the boss and deserve a lot of success for your teaching, your passion is contagious and you make it really easy to understand complex stuff. I am learning my chops these days, so i am pretty sure that i will see a lot more of your videos. kudos :)
THANK YOU so much for your video blog! Makes this here newbie a little more confident. You're doing such a great service to the hobbyist community. Your infectious enthusiasm doesn't hurt a bit either.
1:15
"the mains earth will be common and tied together throughout the whole system"
Ah, but you haven't seen the wiring in my apartment building.
Those who have done the electric wiring job have saved your oscilloscope! (but not your life)... Be Happy ;)
It might be time to split!
Somerville, Massachusetts?
Is that one of those deals where you have to have the washing machine on and paused with a metal zip stuck in the mechanism before the toaster will work
Good lesson. This should be taught to all technicians on day one.
Thank you very much for pointing this out! I instantly stood up, got my multimeter and measured my laptop's power supply. 0.8 ohms! Also I measured between the earth pin and the USB shield, and it turned out: A dead short! I've been probing and soldering stuff connected to my computers for years an had no idea!
A great case for using an isolated transformer!
but earth pin you mesured but there is no connection between main and earth right? So how can you actually make a short circuit? Considering also that there is a transofrmer in the power supply which isolate the rest of the circuits, ok there is ground but still I don't see any point that the circuit could be shorted. Maybe I am missing something?
@@grzesiek1x You're missing that if there's a fault, connecting a mains referenced circuit to an earthed GND can make a dead short. It seems that the comment I originally replied to eight years ago is gone.
@@nrdesign1991 hmmm 🤔 but if there is a fault in the transformer I could be shocked even without oscilloscope (I mean probing with the scope it wouldn't change the situation because I could be shocked even connecting a diode) ? Another thing is that my USB port and also headphones port have 1 M Ohm of resistance between main's ground and them... I ask because I don't follow emotions in electronics but facts and science is important to me.
@@grzesiek1x This isn't about shocking you, but killing your equipment with a short.
Thanks!
Dave, any reason why you couldn't just use two channels on the oscilliscope with the clips on earth and subtract one from the other?
Interesting idea! -I believe he was using an analog scope in the video.-
Edit: upon further inspection, it seems possible. His scope has the capability to invert CH2, and display the two signals added together!
Also, fancy seeing you here! I enjoy your channel. :)
Yes you can use 2 channels. Usually the scope will have either an invert and add or subtract option for combining signals that will measure the voltage difference between the two probes like a balanced input. This is actually a very good way of making accurate measurements because you reject most if not all the common mode noise, ground loop problems, and no more shorting your ground reference, just leave your ground clip unhooked. However, you can end up using a lot of inputs this way. Your 4 channel scope has now turned into a 2 channel, which if you are reading a couple of inputs and need a reference clock to trigger off of, can be a bit of an issue.
Cause you'll lose a channel and if you are on 2 channel scope that's a lot.
That is basically a homemade differential probe
why not make a diagram showing how to do that?
Thank you so much for making this video! You prevented many beginners from blowing up their scopes.
A wire that i tried to use to measure a signal vaporized as a result of my failure to isolate my scope. Good thing nothing was harmed! My scope continues to work...
I’m no expert but I thought the ‘scope should remain grounded and the gear being tested should be isolated?
10 years later, this video is still worth gold!
Hello from Russia, Dave! I like your videos :)
Well, I'm not sure about "completely safe" power supplies with 2 pins. Sometimes capacitance between primaary and secondary windings of transformers plays role (I've read some years ago about multicontroller-based device which was unstable with transformer PSU but stable with PWM PSU which has less parasitic capacitance); sometimes there's RC filter between primary and secondary parts of PSU. From other side, there could be Y-capacitor in PWM PSU of your scope, so if you connect your scope's alligator clip to 2-pins supply's "hot" part of circuit, there could be parasitic AC-connection which at least can damage your scope.
Lucky for me, I was smart enough to doubt my knowledge of electronics and look at the guy who knows. Thank you Dave, saved my oscilloscope!
Dave, you must know that the current carrying neutral of your mains supply is bonded to the protective earth. If you don't chech that neutral for continuity with the CUT ground rail you still blowup. CUT to protective earth is not the only risk in correctly bonded mains circuits.
This is a life threatening problem in tube circuits with B+ running in the hundreds of volts.
You need another video about chassis grounding. I know two people injured by this, one by heart attack.
I apologize for asking this simple question (to you guys) but I really need the answer. Crocellian really brought into focus my concern. Most of our houses in New England have three wires brought into their homes. one is 120 vac, one is 120 vac but phased shifted 180 degrees to permit 220 VAC to form and finally a "ground". The problem that I am having is understanding what happens with a 120 VAC circuit. One wire IS hooked to the ground. Caling that wire "neutral" seems erroneous. Having a separate ground plug (the round one)
seems superfluous. AND it seems like I exposing ANY circuity powered by AC to earth ground exposing myself to the problem David talks about. So what do i do to attempt to create isolation? I can't really see a way unless I put an isolation transformer between the device I want to troubleshoot and the AC power source. Please help me.
william burling the ground wire is used as a safety ground. Referencing the chassis to one conductor, which is named Neutral, provides safety so if the line shorts to the chassis it will open a breaker. Otherwise it could be possible for the wash machine, for instance, have one conductor short to its chassis, and maybe the clothes dryer have its chassis shorted to the other conductor. Motors in each could became damaged and get a wire shorted to the chassis. This could electrocute a person touching each machine. So purposely referencing the chassis to the same conductor creates a safe environment.
@@KissAnalog I think what he's getting at is 'Neutral is bonded to mains earth back at the main breaker panel of a house. So, at least for American homes, seeing only two wires (hot and neutral) spade going into an AC adapter doesn't necessarily mean you are isolated. I learned that really quick about 30 years ago...don't ask how!
@@williamburling3229 Our system is called "TN-C-S grounding, single phase, with grounded center tap."
In the house wiring and appliances, neutral is an electrical system wire. It is allowed to carry current in normal operation. It isn't touch-safe and it must be insulated. Code calls it the "grounded conductor" (groundED). *The neutral wire can develop several volts of AC* difference along its length during normal operation, simply because it has non-zero resistance.
This means one neutral can have several volts of potential relative to nearby other neutrals or the grounding wires. It's not touch-safe or equivalent to the grounding wire. Another reason why they're not equivalent is that a GFCI outlet uses a small common-mode test voltage L&N vs G to probe for N-G continuity. This detects (and trips) if a neutral wire has rubbed or pinched against a grounded part.
Despite the non-equivalence, some old work (WWII in particular) was allowed to run a common to some 240/120 appliances. This is simply less safe than a separate grounding conductor.
The grounding wire is touch-safe. It can be connected in strange ways - if two appliances share plumbing or a mechanical connection, or through a metal building frame, their grounding wires will not be insulated from each other. It doesn't normally carry current. (It will carry a small amount of current due to parasitic capacitance, static electricity, etc.) Code calls this the "equipment grounding conductor" (groundING)
The distribution system is different. It uses a multiply grounded power conductor. Utility electricians usually call it the "ground" wire, but I think "common" is more clear. It's touch-safe IF you only touch it at one point. It's bolted directly to utility poles without an insulator. Safety is ensured by balancing the distribution circuit and making this common wire surprisingly large on the poles. (It's at least as thick as the hots even though it is multiply grounded and carries less current.)
These conflicting systems are reconciled by connecting all three of those conductor types together at one and only one point, which code calls the "service equipment." This one point is touch-safe: it touches the utility's distribution circuit at one point. The grounding conductors are at the same potential as long as they don't carry significant current. The neutral does carry significant current, so it's not touch-safe.
Grounding conductors from a neighbor are generally within a few volts. Still touch-safe (except when swimming pools are involved) but if you run a wire long enough you can find a dangerous disagreement in "ground" potential. Regular earth electrodes helps a little with this problem, but they are more of a lightning-protection measure. Mostly it's not a problem because people don't run long conductors unless they are a telco or electric utility.
Many old electronic devices had a "hot chassis" - connected to hot or to neutral. Neutral through a polarized plug is a little safer than fully hot. All those devices need an insulated case for safety, and they can't be probed unless they're powered through an isolation transformer. New electronics are either grounded-chassis or floating-chassis. Any low-voltage power system inside the device should be galvanically isolated from both the hot and the neutral. The low voltage system will be referenced to the chassis and might be floating or grounded relative to the environment. Modern AV equipment is often isolated from the AC input and grounded through its connections - the plug is polarized only because correct polarity reduces hum.
(Quite a lot of cheap crap doesn't obey those isolation rules.)
Grounded-chassis can be probed, but the reference clip can only go to a signal ground. The clip is not required for low frequency, low precision. Floating-chassis can be probed. It's a good idea to only reference clip to signal ground, but you can often get away with differential measurements. I think it's still a bad idea. If an isolation transformer is set up with a floating secondary, you can do the same thing.
An isolation transformer with a floating secondary is a code violation; you're not allowed to use it as permanent wiring. In all cases the secondary-side groundING conductor must be connected to the secondary groundED (neutral) tap. That makes a floating secondary. A grounded secondary also connects them to the primary groundING conductor. Figure 2 is a good schematic of what's required for a permanent-service transformer in the US.
electrical-engineering-portal.com/purpose-of-shielded-isolation-transformer
Jordan Rodrigues. Hey thanks, I’ve actually read all of that! Bunch of good info.
It does get confusing for guys who work on old tube equipment. Whether or not to float scopes is usually the subject of debate. Many have done it for decades and are still alive to tell you why they do it. This is what I’ve been told....
They give up safety for lower noise. Old tube amps that are hot chassis are noisy because the caps and resistors don’t have a earth ground to drain into. So floating their scopes keeps the noise down because the amp circuit isn’t trying to drain through the scope ground.
I don’t know if that’s the absolute truth of why they float scopes, but that’s what I’ve been told.
I’d rather use a isolated transformer on the old tube amps and keep my scope grounded.
All I’ve ever done is work on modern tube amps with chassis earth ground, so I have never worried about blowing up anything in my hand while my other hand is in my back pocket.
that's why i always use isolated differential probes whenever i am working with high voltage. Can't be so smart and aware all the time. Great video and explanation.
@ 4:06 great Back to the future reference! I love your videos.They have helped explain quite a few concepts when I get confused/stuck. Keep being awesome
just bought an oscilloscope and this video will definatly help in the not blowing things up part
isso aconteceu com minha equipe de manutenção eletrônica, o terra do osciloscopio foi usado junto com a ponta de prova para medir uma tensão em um tiristor numa retificação trifasica de 380VAC, adivinhe? a manlha da sonda ( terra) esquentou devido ao curto circuito e começou a derreter o cabo ! quase causando um incendio! isso foi uma impericia de um colega desatento! grato por seu video !!!
At 13:36, I believe what is missing in the video is to mention that if you are uncertain if connecting the black probe of the osciloscope a certain point of the circuit is going to be a problem or not, a multimeter can be used to check the voltage difference between the ground of the osciloscope and the connect point on the circuit. The voltage drop between the two spots should be 0 essentially. E.g in a valve amplifier, this might be the bolt connected to the chassis of the amplifier.
I've got really good at mimicking the "Hi!" at the start of every one of these :-)
ha ha me too
Watch them at 0.5 speed :)
+Paulo Constantino hahaha oh man now that's stuck in my head
lmao, I keep imagining him saying all kinds of stuff. Like on his introduction, to me he says "I'm your host, David Fucking Jones."
I think we pretty much all do that. Lol
This is a very useful video. I did a Summer job as an intern with RCA in the early '80s. I was fixing broken switching mode power supplies used in studio cameras. I erroneously connected my 'scope ground to a 400V line and it DID ruin my day. Gratefully RCA had given me a big transparent box to keep the PSUs, in while they were being tested, and that took most of the flak. I also learnt that fuses are good !
I always thought that the _ground leads_ on a scope should be protected with 0.1 amp quick-blow fuses- but that's a little too obvious a solution- isn't it?
+Bob Bowie 0.1A is still enough to damage circuitry. But it would certainly be safer, yes.
+Bob Bowie The proper solution would be for scopes to follow the principles of almost all other tools and test equipment, and be built to 'Class 2' double insulated standards, where an earth is not required. The evidence is overwhelming that Class 2 equipment has a far better safety record than Class 1 (single insulated and earthed) equipment. How often do you see drills with an earth these days? For some reason the scope manufacturers have always dug their heels in at this suggestion, and they are about the last of the Mohicans to still be building to Class 1 standards.
It just comes down to money. Remember the old CRT television sets with the _live_ chassis? The fix was.........to go to LCD.
+Phantasmotronogun
One other addition to this would be good-- a switch for floating each probe's ground.
+ReactanceIsFutile Tektronix did produce an add-on to do that, only problem was that if the voltage to ground exceeded 50v it automatically reconnected the ground. It actually created a worst-case scenario because an ohmmeter check would suggest an earth-free probe, but the probe shield would actually short-out if it came in contact with high voltage. It could also kill the operator by providing an (unexpected) shock path to ground.
Basic issue is that old-school electricians place an excessive reliance on earthing. Even in the consumer environment it is now being recognised that earthed metal is as much of a hazard as it is a safety device, because it can act as the return path for a shock from another voltage source. This is especially so on the testbench. The best arrangement is to have no exposed metal which could become live.
Good information thank you. My solution for using scope with mains circuits is to isolate the scope by plugging it into a "portable power station" (Jackery 1000) which is basically a battery and inverter.
Really simple good solution. These days there are smaller capacities that don't cost much. Because the problem is that in many places in the house in the electrical box the ground is connected to the minus.
Another big pitfall is connecting ground to the DC(-) side of a full wave rectifier. It's not obvious at first, but if you aren't using a transformer (ie. your rectifier is connected directly to mains) then connecting your scope's ground to the (-) output puts the ground lead and one one of rectifier diodes in series with mains voltage which will quickly destroy said ground lead and diode. The reason for this is that mains neutral and ground are connected inside the breaker box (that's how it is in the US anyways, might be different elsewhere). EDIT: Why is a portion of my comment crossed out? No matter, I'll put up a video later today.
This is called a "Hot Chassis". Many power supplies are built this way. In TVs to help protect service technicians there will be a silkscreen border line printed on the circuit board to warn you not to touch or connect ground to that part of the circuit. I work mostly with power equipment like inverters so I always use my portable isolated scope, the Tektronix THS720A. The scope is fully isolated and the grounds on the probes are not connected together so you can measure just about anything safely. You can measure a hot chassis with no problem with this scope.
Robert's Electronic Hobbies great scope for the job
This is one example where some people have a massive experience step! They either discover this the hard way if they aren’t too savvy or hopefully they will see your demo and think about this. When I was a young trainee engineer I was on an insulated mate and working on a mains circuit I had somehow rested my hand on a live conductor and because I was sitting on a padded seat I didn’t feel anything....that was until I picked up the scope lead with my right hand and went to attach it to the scope.... OOOH FCUK I shouted ...this scope chassis is live! Potentially the worst kind of shock you can get, across the chest through both arms.
Good demo mate as always!
I'm scared.
Got a Big Bang yesterday, luckily blew the breaker, I was clueless. And afraid to try again. Can’t wait to get back in the shop with my new found knowledge. Thank you so much!!!
4:07 Love the reference to Back to the Future 1
For many years I have seen some people cutting the ground pin off on the scope. I never did that on mine but have just being careful about grounds. When not sure I have a plug adapter without ground. Grounds can be tricky. Thanks for the video. I have seen actually people getting shocked too. Good lesson to keep in mind.
👌🏽great examples and explanation of the potential differences!!
Thanks! Great information that I did not learn in engineer school! Maybe was mentioned once in a lab to keep people from blowing up oscilloscope leads. Seeing this now, I actually think that happened to someone.
Currently watching this while my eyes heal from staring directly at a wire that vaporized after I made this mistake.
Thanks Dave for the video. Even after a decade, this is still relevant. your videos ages like wine!!
19:24 the computer did not reboot the usb hub disconnected the Arduino for drawing more current than 0.5 amps.
The sound we hear is the sound of plugging/unplugging something into a usb port.
I know this was made 11 years ago, but still very helpful today! You're the man!
Why not leave the ground clip unconnected and just use the probe? The oscilloscope gets ground reference from mains? Any reason to not do this?
You will still charge the grounds on the scope and charge them with potential to zap you or short if the potential is not same. The ground is there to protect you not the equipment.
I have a scope that is powered on USB and I can power it via battery just for such. The battery isolation keeps the boom potential lower still. 😊
Very educational video about ground, earth and chassis applications. I couldn't find a video that explains the difference between those and I ended up finding it when watching this oscilloscope video.
"...has become a real potential problem!"
Ha ha...
Thank you very much Dave! Again, i watched this video and learned more than in school for 6months. Your way(approach) of teaching is completely opposite from the way how they teach us in schools. It is totally interactive and makes you ask and wonder.
Learn how to release the blue smoke with this one weird trick1
Thank you very much sir !
Finally, your explanations are crystal clear ! I've seen other videos and NONE could explain it properly to a beginner like me...
I understand now how I can damage the oscilloscope I've just bought recently, probing without care...
I choose a wall outlet that's earth grounded for the oscilloscope and, by pure chance, the power supplies I use to power up my breadboard (one ATX SMPS and one wall-wart) are plugged in wall outlets that aren't earth grounded (no earth hole).
Now I know it would be risky if my power supplies were also earth grounded...
So does this mean I can connect one probe's ground lead to a local ground on the circuit and leave the others unconnected, and they'll all reference the same ground voltage? If so, this makes connecting probes a lot easier!
+Paul Wayper
Normally a probe's leads have their ground tied together inside the scope. So if you have one ground lead connected, all probes are now referenced to whatever the ground lead is connected to.
Electrically yes, electronically NO. Try doing that with mixed audio or high frequency signals and you can EASILY get into a world of bother with incorrectly shielded signals, bad return paths and/or earth loops. The main thing as in ALL electronic and electrical measurement is to properly understand WHAT you are doing, WHY you are doing it and HOW to do it correctly! SOMETIMES it may be necessary to use MULTIPLE different scopes.
As an example, if you have two scope leads in an RF environment and only ONE signal ground connection back to the scope how can you be certain that currents flowing down the common shield are not interfering with currents flowing down the SAME shield from the second signal source? Or producing a skewed voltage reading due to tiny increases in voltage drop down the shielding? They are BOTH shielded but there is only ONE common signal return path. There has to be a return path but SHOULD it be down ONE shield only? If BOTH shields are connected at the scope (they almost always will be if only via chassis ground) are you creating an RF loop antenna with the two shielded cables providing a complete loop which can massively increase common mode noise? It's NOT as straightforward to get this stuff right in every circumstance as people may think! One reason why GOOD differential scope probes cost so much - it's not just the design but the precision of manufacturing for what will be a VERY SENSITIVE piece of measuring equipment.
Rule of RF: Every wire is an antenna. If you want your probe leads to NOT be an RF radiator, connect the ground ... that's why probes use coax!
Only for low frequency work.
Something I learned by looking at tv schematics is that although ground is almost always zero volts, don't ever just assume it is. Besides faults and other reasons that a ground is supposed to be zero isn't, sometimes equipment is designed with a specific voltage as ground. For example, many General Electric CRT TVs have 70V rail marked as ground throughout and chassis ground is earthed.
I think the biggest innocent mistake folks make in the beginning comes from using a multimeter as our first real piece of test gear where we have two leads/connections to hook to a circuit to test, usually a red lead and a black lead. The color of the lead connections doesn't really matter, it's the muscle-memory of making two connections to a circuit/component to make the measurement ("I must have to hook this up somewhere so here goes"). Once we become adept at using a multimeter, that muscle-memory tells us that with a scope probe, we still have two connections to make to the circuit/component that we want to test, so we connect the alligator clip to the other side of the component or IC or whatever by habit of ....connecting two leads to make a "proper measurement" and that's when the smoke happens and we are left wondering why the piece of gear we were trying to repair got worse or usually sparked and let the smoke out. I've always determined what the actual true ground of the device-under-test (DUT) is and then hooked my scope-probe alligator clip to that point only and using only one of the gator clips if using more than one oscilloscope probe (the others are unnecessary after the first).
Thanks again Dave :)
I just did that innocent mistake :)
How did you determine the true ground of the DUT?
When using multiple probes, the safest thing is to unclip (and set aside) all but one of the clip-on alligator wires.
AND!!! Remember that you can also change voltages on the circuit and kill devices, even if the scope is fine! Great video!
So as a general rule, all I have to do is make sure the alligator clip is connected to circuit ground and I'm free to probe anywhere I want within the circuit, correct? Doing this the only thing you lose is the ability to reference the probed point against something other then ground.
Yes, for which you need to use differential-probe or 2-channels with channel1 - channel2 math operation on, most digital have this (old analog have + and invert, channle1 + invert( channel 2).
Hello, Can i put the alligator to the building ground and go with the probe measuring the 400v phases? (one at time of course)
Thanks, always a great reminder to mind your grounds between components, especially audio amplifiers and power supplies.👍👍
I was asked about this in an electronics interview today and had no idea about it.. Man i must have looked stupid to the 3 engineers that interviewed me :(((
...... No one who wants to learn looks stupid.
@@robbieaussievic You're right, I actually ended up getting that job.
Holy moly mate, I was about to probe a laptop with a similar power supply. You've just saved another noob! Thought I'd let you know... Thanks for sharing your knowledge!
Thank God for quick acting circuit breakers, my scope could have died about a week ago...
Wow, am I glad I watched this video! I've started building tube powered guitar amps, and the high voltage rail can often be around 500VDC. Shorting 500VDC to ground would've been very unpleasant. Thanks Dave!
Rangy Tang similar to what happened to an acquaintance of mine he somehow got his guitar plug across the mains plug while rearranging his studio . His story never made sense but his amp took the hit the PC tracks got trashed. Luckily for him the mfr. took pity on his situation and supplied him the schematics. His luck held out and the majority of the repair was laying new tracks and 1 power transistor.
Ensonified my row home neighborhood with 200 watts of mr. BoJangles. Out of his stack.
So What if I want to measure the mains with my scope, I'll short live to earth?
just change ".. i'll short live" .. to: " ...i'll short life ..."
Isolate something - either the device under test or the test instrument. If the device under test is the whole electrical grid, you've got one obvious choice...
only if you connect your probe's earth lead to live wire in your wall socket.
Afaik ground is connected to neutral somewhere in the grid. So you might get away with just leaving the earth clip dangling loose. I never tried it though, you need quite some faith in your science to put your 500$+ scope at potential risk. I would at least put a transformer between it with like 5v output or so.
Simply Electronics find a 240 outlet with a scope that is capable of that voltage rating and probes that meet or surpass 240vac
Thank you so much Dave that you took time to make sure people pay attention to this obscure problem of ground isolation. I have fried stuff before years ago and could not figure out why.
Thanks heaps Dave, this was extremely informative!
One of the best educational materials I've watched recently. You are the man!
Which,boys and girls, shows why you should run your Arduino, /Rasberry Pi etc. from a battery if you're going to probe it.
+Robert Langford Or use an isolating transformer, which you'll be doing anyway unless you want to supply your arduino with 230v!
+BenjaminGoose Isolation transformer DOES NOT equal a battery, but is close. The isolation transformer capacitively couples some line current to the output. Depending on what you are probing the coupled current may or may not matter.
SirDeanosity Ah ok, I'm a bit of a beginner so I don't know much about this sort of stuff. Thanks for the response.
YES! and STAY AWAY from live mains ... and that includes Main's earth connection!!
I used to work in a big electronic workshop to repair industrial equipment, to avoid errors, all oscilloscopes had their main earth connection removed inside it. Not correct I know, but with so many different technicians using the same instruments it was the final solution that solved the issue of randomly fried scopes.... even with a lot of explanation, people did mistakes and blow up the probes from time to time until the earth connection was removed and the workshop had peace.