I finally did it. After years of wanting one (and pushing off projects because I didn’t have one) I finally bought an oscilloscope. Over the years I read and watched a ton of content about how to use a scope, you’d think I would know what I’m doing. Turns out that, like anything, hands-on time with an oscilloscope quickly highlighted the gaping holes in my knowledge. And so we begin this recurring column called Scope Noob. Each installment will focus on a different oscilloscope-related topic. This week it’s measuring a test signal and probing Alternating Current.
Measuring a Signal
Hey, measuring signals is what oscilloscopes are all about, right? My very first measurement was, of course, the calibration signal built into the scope. As [Chris Meyer] at Sector67 hackerspace here in Madison put it, you want to make sure you can probe a known signal before venturing into the unknown.
In this case I’m using channel 2. Everything on this scope is color-coded, so the CH2 probe has blue rings on it, the probe jack has a blue channel label, and the trace drawn on the screen is seen in blue. I’m off to a fantastic start!
This scope, a Rigol 1054z, comes with an “auto” button which will detect the signal and adjust the divisions so that the waveform is centered on the display. To me this feels like a shortcut so I made sure to do all of this manually. I started with the “trigger” which is a voltage threshold at which the signal will be displayed on the screen. The menu button brings up options that will let you choose which channel to use as trigger. From there it was just a matter of adjusting the horizontal and vertical resolution and position before using the “cursor” function to measure the wave’s voltage and time.
I played around with the scope a bit more, measuring some PWM signals from a microcontroller. But you want to branch out. Because I don’t have a proper signal generator, the next logical thing to measure is alternating current in my home’s electrical system. I suppose you could call it a built-in sine wave source.
Probing Alternating Current
I sometimes take criticism for never throwing things away. Seven years ago we had a cat water fountain whose motor seized. It was powered by a 12V AC to AC converter seen here. Yep, I kept it and was somehow able to find it again for this project.
Of course at the time I thought I would build a clock that measures mains frequency to keep accurate time. This would have done the trick had I followed through. But for now I’m using it to protect me (and my fancy new scope) from accidental shock. I’ll still get the sine wave I’m looking for but with a source that is only 12V at 200 milliamps.
Don’t measure mains directly unless you have a good reason to do so.
Continuing on my adventure I plugged in the wall wart and connected the probe to one of the two wires coming out of it. But wait, what do I do with the probe’s reference clip? I know enough about home electrical to know that one prong of the plug is hot, the other is neutral. The clip itself is basically connected directly to mains ground. Bringing the two together sounds like a really bad idea.
This turns out to be a special case for oscilloscopes, and one that prompted me to think about writing this column. Had this been a 3-prong wall wart, connecting the probe’s reference clip to one of the wires would have been a very bad thing. Many 3-prong wall warts reference the mains earth ground on one of the outputs. If that were the case you could simply leave the clip unconnected as the chassis ground of your scope is already connected to mains ground via its own 3-prong power cord and the reference clip is a dead short to that. If you did need to probe AC using the reference clip you need an isolation transformer for your scope. There are bigger implications when probing a board powered from mains which [Dave Jones] does an excellent job of explaining. Make sure you check out his aptly named video: How NOT to blow up your oscilloscope.
As I understand it, and I hope you’ll weigh in with a comment below, since the wall wart I’m using has a transformer and no ground plug I’m fine using the ground clip of the probe in this case. Even though I’m clipping it to an AC line, the transformer prevents any kind of short between hot/neutral mains and earth ground (via the probe’s ground clip). What I don’t understand is why it’s okay to connect the transformed side of the 12V AC to mains ground?
At any rate, the screenshots above show my progress through this measurement. I first connected the probe without the ground clip and got the sad-looking trace seen on the left. After conferring with both [Adam Fabio] and [Bil Herd] (who had differing opinions on whether or not I should “float the scope”) I connected the ground clip and was greeted with a beautifully formed sine wave. I’m calling this a success and putting a notch in the old bench to remember it by.
What’s Next?
I don’t want to get too crazy with the first installment of Scope Noob so I’ll be ending here for now. I need your guidance for future installments. What interesting quirks of an oscilloscope should a noob like me explore? What are your own questions about scope use? Leave those below and we’ll try to add them to the lineup in the coming weeks.
Homework
For next week I’m working my way through the adventure of rectifying this 12V AC signal into a smoothed DC source. Here you see a teaser of those experiments. I’ve built a full-wave rectifier using just four diodes (1N4001) and will plunk in a hugely-over-spec’d electrolytic capacitor to do the smoothing. If you want to follow along on the adventure you should dig around your parts drawers for these components and give it a try yourself this week. We’ll compare notes in the next post!
Hey Mike, If that scope is anything like the DS2072A then it has a ‘trigger on mains’ function.. which triggers on the 60Hz (or 50Hz) mains voltage coming into the scope! I thought it was an interesting feature that could possibly be used to check out noise sources that you think might be coming from the mains.
Hey that’s a good tip, thanks! This setting is in the trigger menu –> Source –> AC
I have the full-wave rectifier hooked up. When I select AC as the source the scope registers a frequency of 120Hz, I assume because it measures the inverted half-wave as it’s own period.
What’s strange to me is that instead of showing 0v as the bottom peak it’s showing 7.9v for high and -10.9v for low. I’ll have to play around with this to understand what’s going on.
Dave Jones did a good video on EEVblog #685 – What Is Oscilloscope AC Trigger Coupling?
He did an excellent job of describing it even for an old analog scope like mine.
https://www.youtube.com/watch?v=y5aAjd9YPok&list=UU2DjFE7Xf11URZqWBigcVOQ
Look forward to more issues of Scope Noob.
Stupid YouTube.. fogot to takeout the “playlist” part of the link. I HATE YouTube playlists,
https://www.youtube.com/watch?v=y5aAjd9YPok
I came here to post the same vid, so informative!
As for the signal ranging between +7.9V and -10.9V, when you expected 0 to 17V or so, I suspect you have the channel set to AC coupling. That removes the DC component of the signal, allowing you to see small AC signals superimposed on large DC offsets (useful for looking at power supply ripple, for example.) But if you want to know the DC value of something, be sure your channel is set to DC coupling. (the issue of the channel coupling is distinct from the issue of trigger coupling, which is explained in Dave’s video that Trav linked to).
I’ll note that, in the picture you posted, Channel 1 was AC coupled. I can see the yellow box on the lower left corner of the picture has a ~ symbol, meaning AC coupled, and it also says the scale is set to 10.0V per division. When you count the number of divisions of the wave, it’s a bit more than 4 divisions peak to peak (more on this later…)
When you’re set up with DC coupling, the output of the bridge rectifier should be roughly 0 to 18V or so. Recall that, even though the transformer label says it outputs 12V, that’s normally an RMS voltage. For a sine wave, the peak voltage is sqrt(2) times the RMS voltage, so 12V RMS should be equivalent to about 17V peak. If you’re seeing something a bit higher than that, you may have run into the fact that, with an unregulated transformer, the no-load voltage is a bit higher than the voltage under load. The placard on the transformer gives the output voltage when the designed load is attached, so the no-load voltage will be a bit higher. Note that your posted picture of the raw output of the AC transformer shows a bit more than 20V p-p. When you rectify that, you’ll basically flip the bottom half of the waveform up to the positive side, and you’ll lose a couple of diode drops, so 18V or so peak is about what you’d expect.
Great explanation Richard! Mike was AC coupled – I had told him to switch over when he was looking at the original AC output of the transformer. I forgot to tell him to go to DC when he mentioned trying to rectify the signal! (whoops!) All worked out in the end though, so no harm done.
Note that, when probing the output of the transformer secondary, it shouldn’t make any difference whether the channel is set to AC coupled or DC coupled. The DC component should be zero. If you doubt this, try switching back and forth between the two types of coupling.
As a rule of thumb, DC coupling should be the default unless you have a reason to switch to AC coupling. If you need to look closely at an AC signal that has a DC offset, switch to AC coupling, but be aware of what you’ve done.
This was something I found confusing when I first started using an oscilloscope. I think it would help if instead of being labeled DC and AC it was labeled something more like high pass filter on/off.
Awesome. I have to shame myself and admit I have to look up how to use the thing each time I break it out. Thanks for this.
Having been out of school ~20 years and not working in the EE field, I also find myself looking up reference materials. I’m here to learn, and I’m not ashamed to admit it…
Hi Mike,
The reason it is okay to connect to the ground on the 12v “transformed” side is due to the transformers construction. Normally, a transformer will have two connections per coil, Live and Neutral. These connections correspond with each end of the coil’s wire. However, this transformer has a special 3rd connection called a “Centre Tap”. It is a wire connected to the exact centre of the coil, and thus at zero volts.
Hope this helps!
That is not really true… He is measuring the secondary of the transformer so he is isolated from the mains. The two output pins of the transformer are basically floating at some unknown potential ant when the ground clip is connected it just pulls one potential to earth. The voltage between both leads is unaffected by this.
If you have three wires secondary this does also not change. But your measurement result changes depending on where you set your refence with that ground clip.
Cheers ;)
Hi Mike,
I agree with the Redbaron. A transformer isolates the main completely. So there is no reference yet! The secondary of a transformer is just like a battery only AC. If you want to probe it you have to set the reference (by connecting the reference wire). You may choose you’re reference as you like as long as the system is isolated. But remember, you can only choose one reference point (for both probes). Most of the times you want to set you’re reference in a way all other voltages are positive or at a center point if it’s symmetrical. And remember, the reference pin of the scope is connected to mains earth. So by connecting the reference wire to you’re system (battery, plug pack, lab power supply) the system isn’t isolated anymore! You made a connection to mains earth.
And about plug packs. Most plug packs (with or without mains earth plug) are on the secondary not connected to mains earth (or hot or neutral). Dave Jones explains it in his ” How NOT to blow up your oscilloscope” video. But it’s better to check with a DMM before you connect the reference pin to a plug pack. So what you state in the article is wrong.
But as you have seen, if there is no reference (isolated from mains and reference pin not connected) you get some weird stuff. That’s because the only coupling between your system and the reference of the scope is trough some capacitive air coupling. You can only measure volt between something (like you can only define distance between two points). Unless ohm it has no meaning without reference set. Just like you can’t measure voltage with your DMM if you only connect one lead. (You get a weird reading as well.)
Timo
I’m also a ‘scope noob, but think I spotted a couple of things you missed. One of the output of your transformer might have been connected to one of the mains terminals – hopefully not the hot/live but possibly the neutral. It might have been wide to use a multimeter to check first. Also just to clarify, never float the ‘scope with an isolation transformer, always the device under test.
That is wrong. The output wires of a wall wart are never connected to the mains.
That’s the sort of assumption that can damage your scope (or in rare cases get you electrocuted). They’re usually not. They shouldn’t be. 99% of the time you’ll be right.
Some DC output wall warts might tie their GND output to mains earth. That cheap piece of junk you got from China via eBay or salvaged from an old piece of kit… who knows how that’s been wired? Poor quality kit and differences between national conventions might catch you or the equipment manufacturer out. Just get a multimeter and check the resistance between the output pins and mains terminals before you connect it to your scope.
I do speak from experience by the way. When I was about 12 I borrow a ‘scope from school and was checking the output of a 30VAC transformer. One of the outputs was earthed and I ended up completely destroying a BBC Micro. Luckily the borrowed ‘scope survived. (It was quite a while ago and I was trying to make an EPROM programmer at the time.)
No its not
https://en.wikipedia.org/wiki/Ground_and_neutral#Earthing_systems
I may be completely wrong, but that 12v could be grounded to the same olace as “earth” pin. The inly difference is that the connection between “earth” and 100v “nuetral” must occur in the main house panel. Same copper going to same place, (of course with the possibility of load or voltage from shorted equipment on the wire run) perfect world 2 prong devices ARE grounded, at least on the neutral. You can do a quick continuity check to see if that is true for your AC-AC device. Also you can’t trust earthing to nuetral if the socket is reverse wired. (Of course the same holds tru for ground, but it should be theoretically more difficult to mis wire that. Bare wire and usually doesnt connect to the sides of recepticle.
For myself it might be a moot point. I would likely just probe the 120v, or my car ignition. Would I need a 1/1000 probe to measure a 10-40kv system? What would be safe there?
Regarding probing 10-40kV: Be very careful here!
There’s a max voltage at the scope’s input terminals, and a typical passive probe has a voltage divider that means the voltage at the source is higher than the voltage at the scope input. However it does NOT follow that you can always safely operate a probe with a source voltage that will deliver the scope’s rated voltage to the BNC jack. The probe itself will have its own maximum rating at the input, due to the limitations of its insulation and construction.
To probe high voltage, you need a probe that simultaneously meets two criteria: It must be rated to safely handle the peak voltage you’re wanting to probe, and it also must divide that voltage down so that the voltage it delivers to the scope is within the scope’s input limits. You must check both of those, and don’t proceed if only one of them is met. Finally, if the power source is high energy or high voltage, you should be using the proper personal protective equipment, like blast shield face masks, properly rated gloves, proper shoes, etc. And you shouldn’t be working alone. Be very careful.
I know I’m digging up an old thread here, but there’s another issue that can trip you up when measuring high voltages. It’s not safety-related but might still leave you scratching your head when you don’t the get the results you expect. If your HV source is high impedance (i.e, it can only supply a small current) the impedance of the probe will become significant, and can load down the HV source. So you’ll see a signal that might be the same shape as what you expect, but the voltages will be too small.
Now, of course you should be compensating for the impedance of your measurement setup no matter what you’re measuring, but when working at 10kV, the relative magnitude of the errors are not what you’re accustomed to if you normally just probe 5V TTL circuits.
And yeah, be careful with big voltages, even if the current is small!
I was always told when probing around high voltage (300+) to work with one hand in your back pocket. This way you don’t get the voltage straight across your heart.
That’s helpful, but if the current passes through your arm to ground through some other part of your body (like your legs, feet, or torso below your heart), then you could still get a fatal shock. So be careful.
If you’re looking to learn something, try measuring a digital pulsetrain with different lengths of ground wire. Start with a modified wire from the probe to be a short as possible. You can use stiff wire and place it less than a cm from the probe tip. Try to measure as close to the chip as possible. Then measure at the terminating end of the trace. Zoom in to the signal and look at the ringing on the waveform. Now start increasing the length of the ground wire. Start connecting to a gnd further and further away from the source chip. What you will find is that your scope will lie to you, and the longer the ground, the more ringing and distortion you will see. This is especially problematic if you are debugging transition levels.
Totally agree with this comment. The difference in observed signal is huge, and it would be good for HaD to illustrate it in a future post in this series.
Another thing to try is to test the Nyquist frequency, or maximum sampling frequency of your scope (fmax = fsample/2). If your scope samples at 10 MHz, then the fastest waveform you can observe correctly is a 5 MHz. If you try to look at a faster signal, you will get aliasing and will see an incorrect signal.
A couple things that apply here:
To get a good signal without added noise, you should always connect the ground for the probe (there is more to it, but really not anything to consider at this point in your ‘scope-life), even for ground referenced signals. BE SURE THAT THE POINT IT IS CONNECTED TO IS THE GROUNDED POINT. This avoids showing mains ground noise and ground-loop pickup on your signal. IF you are using the high impedance input (not 50ohm), and there is a ground loop concern or worries about differential voltage at ground (which can happen to the tune of several volts, even on circuits fed from the same panel), then a resistor can be used in series with the probe ground connection. I generally go with about a 10K resistor in such cases, after checking to be sure that there is minimal differential using a DVM, and leave the DVM across it.
If you are using more than one probe on the same circuit, you usually don’t need to connect the ground strap for the second probe unless the frequency is very high. How high depends on the probe, scope inputs, and a bunch of other things. Guideline might be 10 to 20MHz you need each probe strap connected to the circuit under test.
Another key thing is that one leg of the 120V outlet (US, Canada, and elsewhere) is connected to ground at the service enterance to the building, and is therefore grounded. On a 240V circuit, BOTH legs are hot (120VAC to ground). In most places, similar will be true, but the details vary. Some old equipment (mainly televisions and radios, but other things as well) with two prong plugs counted on this for shielding. If the plug was backwards, hum city. Led to all kinds of fun debugging and repairing.
The output of the transformer can be connected to ground (On one side) because it IS isolated. Just like you can connect one terminal of battery to ground. Without a second connection, there is no circuit.
This is why I keep an isolation transformer around and why I still have my old tube scope with full-float inputs
I’ve seen people who should know better with scope setups that are a couple of times more expensive than my house assume that high speed, high impedance diff probes don’t need to be grounded and the scope mains isolated, only to discover that the common-mode glitching goes away when they unplug other mains-connected equipment from the target.
I am amused with this . Get a scope , use it a few times ,write an instruction blog !
A bit like build a boat ,write a book to tell everybody how to do it.
Its a bit premature and arse about you see
Normally you put in 50 years experience first , then write the book.
I think you miss the point. There are plenty of tutorials written by experts. See the eevblog video I linked.
Here I’m writing about my adventure in learning to use the scope, and showing that it’s not something to be afraid of. Everyone’s a beginner at some point.
Mike some people are confused because when they read “Scope Noob” and your introduction paragraph they are expecting the “sage of scopes” to spew enlightenment. Happens all the time.
Seriously though, this is great! I think Me’s comment below really highlights the value of your column. I think this is a fantastic idea and I’m looking forward to your future installments.
Must resist urge to buy scope right now….
Yes, after you are so far removed from the beginners that you don’t even know what questions they might have… then you write a big, heafty book to discourage the next generation.
A “guide” that is as much “follow me as I learn” as it is guide could be a good way for a new person to pick up beginner information. It’s likely to hit on the same points that any other beginner will need. They can read your 50 years of experience later when they are ready to become real experts.
This isn’t an instruction blog, it’s a learning experience blog. Beginners can read about someone else starting out with a scope and feel like they can do it as well. “Hey, isn’t that the guy I saw walking around the Makerfaire with a foam Skull and wrench on a stick stuck in his backpack? If he can do this anyone can!”
When I started to learn how to use a scope I didn’t use the probes, just a wire to the BNC connector. That kept me from making a stupid grounding error. When I needed more stable signals I’d switch to a probe.
Forgot: Check the rating on the scope and the probe before going at high voltage (higher than 10V or so). Some scopes have very low V spec for the input, as they expect a X10 probe to always be used, which is a good idea under most conditions anyway, as the response and test circuit loading will both be improved over a X1 probe.
A 50ohm input may be only a couple volts max
Good point. I didn’t mention it but I am using 10x probes here.
Glad to see the column became a reality. Good write-up, I look forward to future posts.
1. You MUST to connect the ground clip somewhere. Otherwise the best you can get is qualitative information like: there is a signal, but without any true information about its quantity (voltage). That is how voltage measurement works: you measure voltage between two points.
2. Transformers have a primary (input) and a secondary (output) winding. In case of wall warts the primary is the one connected to the high voltage (110-230V) and the secondary provides low voltage. In general there is no technical reason not to use such transformer to raise 12V AC to 230V AC. Let’s remember: the primary winding is connected to the source of energy and the energy is drawn from the secondary winding.
3. The secondary winding of a transformer connected to the mains is isolated from the mains. There is no voltage between any of the secondary winding terminals and any wire of the mains. This is because there is no direct connection between primary and secondary windings. You can consider the secondary winding of a transformer an “AC battery”. If you connect your ground clip to the secondary winding nothing bad will happen because the circuit (scope (mains) ground , wall wart (mains) live wire) is NOT closed. This is the reason 1:1 isolation transformers work.
PS. You MUST NOT connect the protection (“grounding”) hole/prong in your socket to the ground wire inside the socket. If you do so you may kill someone or get killed yourself if: the ground wire is broken AND the insulation inside a device connected to such socket is broken too. In such (unlikely, though possible case) there will be high voltage on the metal (supposedly grounded) surfaces of the device. The protection (PE) wire is supposed to be separate and go straight to the main house panel and make residual-current devices work.
You might want to use a different diode. The 1N4001 is only rated for 50v. I never use anything less than a 1N4004 (400v) when dealing with AC, even low voltage AC. Better safe than sorry, and the higher voltage diodes are the same price as the 50v ones.
Why?? A 12V plug pack has a top of around 17V. Because it’s not regulated, let’s say 25V. That’s much lower than 50V. So why use a bigger diode if he has a 1N4001? It’s well suited :D
17 volts AC RMS is 48 volts peak to peak.
True, but he has a 12V plug pack. That makes a top of 17V. But let’s say, because it’s unregulated, it puts out 17V. The peak value is only 24V. And it’s the peak voltage that matters, not the peak-to-peak. So 1N4001 (even with a large margin) is a factor 2 over dimensioned. So no pain there.
Why?
Why not? Why use the weakest possible component when you could use one that costs the same and won’t crap out on you the first time a spike comes through.
Always overspec your components.
There is usually a pseudo-differential mode setting for most scopes. It basically uses CH1 as + and CH2 as – inputs, so you want the scope to do the math for (CH1-CH2). On some scope it is marked as “CH1+CH2” at Ch1 and “Inv” at Ch2. It should be in your manual/app note for your scope. You can use that if you have to measure AC mains. This would work for measuring low frequency signals while letting you have isolation without floating the scope ground. Not the same as floating the scope ground, but in most cases it is good enough.
both inputs to the *same* proper range and make sure that scope and both probes are rated to handle the voltages. You also have to make sure that both channels are calibrated to same gain as any mismatches would affect your measurements and also common mode rejection. Just about the only safe ground reference point for ground clips for the probes would be Earth Ground and make sure that both the scope and what you are measuring is on the same Earth ground connection.
Ok, we all make mistakes when learning, and have to ask for advice… But I’m literally wincing at how much bad and confusing advice has been given.
So I’ll made it simple and clear. For this measurement, where the transformer provides full isolation from the mains, the second scope connection (reference) should have been made to the other wire coming out of the transformer on the low voltage side (12VAC).
And definitely NOT to ground, as found on the 3rd prong on the AC plug. As [Steelman] said, when you’re measuring voltage, you’re actually measuring the *difference* between two points. By connecting to AC ground, you’re measuring between two circuits which have no connection (due to full isolation by transformer in this case), and therefore no clearly defined relationship to each other. You may still find it works, and get a clean trace; but that is in some part by luck (things outside your direct control). Try the exact same improper measurement in another building, and you might get a different result. Or plug the wall wart in “backwards” (if it has non-polarized prongs, which it likely does), see if your measured voltage changes.
As for diode drops, center tap transformers, and more hat other folks have listed… None of that applies to the measurement in question, and is really only confusing the issue when the basics need to be clarified.
Three important points regarding line voltage:
1) ALWAYS use an isolated scope or use a 1:1 transformer between the wall to power the scope. This will prevent ground loops from destroying your scope and your electronic Device Under Test (DUT).
2) If you have unisolated circuits on your oscilloscope (meaning all the grounds are tied together on the probes), don’t probe around foolishly and ground your probe to different potentials. This can also harm the device. In fact, a lot of people use the ground clip to connect to their singular ground point and then
3) If you have unisolated circuits on your oscilloscope and need to probe differently grounded potentials, then use differential probes. (Yeah, they aren’t cheap, but you don’t have to diagnose your oscilloscope when you’re done.)
Don’t hold me accountable for actions. I do this type of work all day, and I know how to do it safely.
Figure out before you start which type of oscilloscope you have:
1) isolated system, isolated circuits
2) isolated system, unisolated circuits
3) unisolated system, isolated circuits
4) unisolated system, unisolated circuits
One very easy trap to fall into, is that both ground plugs on the scope probes are connected to eachother inside the scope. So you can’t use channel 1 referenced (ground clip connected to) say 10V and channel 2 referenced to ground, it will short the 10V source. With a small adaptor powered thing, not much will happen except that your charger might break, but with (lithium-something, big NiMH and NiCD) batteries, things might turn ugly for the scope.
Happy probing!
How to measure “mains” signals directly with ab o’scope.
DANGER HIGH VOLTAGE PRESENT, ALWAYS COVER EXPOSED WIRING with some type of insolation.
Assumptions – 2 channel scope (minimum) with input range higher than high voltage to be measured. Same for scope probes, must be capable of handling rated voltage.
I use a power cord that I cut the “equipment end” off to gain access to the wires to be connected. Terminating these wires with small ring lugs makes connections to scope probes easier.
Attach scope probe #1 to one of the power cord wires and tape up sufficiently to provide shock protection. Attach scope probe #2 to the other power cord wire in the same fashion. If you have a 3 wire earthed power cord, you can attach the scope grounds to the earth ground conductor. Make up the connections to this power “test” cord BEFORE plugging the cord into the voltage source to be measured. Verify all connections and safety insulation is correct and in place before connection to the test source. DO NOT connect cord to source until AFTER scope setup is complete below.
Scope settings. Set voltage range for both channels to a value that will allow viewing of voltage range expected. For 120 volt circuit, I would suggest something on the order of 50 volts/centimeter (assuming scope gradient is in centimeters). Set “Vertical Mode” to ADD (A+B). What you are doing is to setting the scope up to show the voltage difference between the two scope probes. In this setup, you are NOT connecting the scope ground (via the scope probe ground leads) to either of the power legs of the circuit. One scope lead will measure the hot leg, and the other the neutral leg (in a 120 Vac US circuit) or each will be a hot leg in the 240 Vac European circuit. The scope will display the voltage difference between the 2 scope lead inputs just as a voltmeter would measure the voltage between the meter leads.
After you have set up the scope and connected the scope lead test cord to the scope, you may proceed to connect the test cord to the voltage source to be measured. ALWAYS be aware of where the scope leads and test cord are to assure that you do not provide either a shock hazard or a shorting hazard.
Practicing this method on a AC “wall wart” transformer would be good idea, giving you the chance to go through the process using safer voltages. The process is the same, and even if the transformer does tie the secondary side of the transformer to the primary side ground, you will still be safe from destroying an expensive piece of test equipment.
Using the method of measuring differential signals on the 2 channels of the scope can also be used for other “isolated” measurements such as current flow through a system. Having a power cord with an appropriate shunt resistor provides the ability to measure the voltage ACROSS the shunt providing the ability to see the current waveform in a system. Using the differential measurement method provides the ability to connect to both sides of the energized shunt resistor. You can observe the current waveform and with the appropriate math, can calculate the current value from the measured voltage and shunt resistance value.
I still say never float the scope. Also, never give your managing editor advice that may electrocute him. :)
I bought my first scope, a 100M HP analog dual channel from a guy that had the same philosophy. He’d burnt out both channels and sold it to me for $5. I took it to a repair shop where they repaired it, calibrated it and gave me the service manual on microfiche for $100. I still have that scope and haven’t burnt a channel out yet because sometimes when I am first looking around I float the scope and do not use the reference ground.
The major advantage of using a transformer (apart from the fractional voltage output) is the isolation of input to output. The output is electrically isolated and only magnetically coupled to the input, allowing safe circuitry on the output side that isn’t directly connected to mains.
The output of a transformer is always floating at an unknown poitential until you tie one terminal to a known voltage. You can happily connect either terminal (but not both, ideally!) to ground with the scope probe.
Nice work, Mike. I acquired a DS1074Z (the 1054Z was out of stock and I didn’t want to wait), and it’s my first big boy scope, so I’m glad you’re writing up your experiences.
Maybe we’ll have to collaborate on some upcoming installments since we’re both local?
Wow, working with a scopes which cost as much as my house did, pretty much daily, all day, everyday and on high voltage using 1000:1 (or more) dividers with 65kV and up standoff that cost as much as if not more than the scopes, this entire comment section is making my hair stand on end right now!
I don’t see anything that Mike is doing wrong, he’s learning!
All this gibberish about floating his scope on an iso, or clipping the ground pin, or any of that is pure hogwash! using a wallwart is fine. The primary and secondary are isolated from each other like others have said. And even if it did have a ground pin, it make be just for an isolated shield between the windings. And if it is referenced to one of the secondary wires, a DMM should be used to ohm out which one, the ground clip would connect to it.
The ground clip is probably the most dangerous part of the scope as it is very much ground as long as he’s using a normal power cord. If that were to contact an actual AC live wire, or something that is referenced to live, sparks WILL fly! I have seen switch mode wall warts and dimmer switches that have their internals referenced to live, and that will fry stuff in a hurry if you don’t isolate your scope! That is when you want to float your scope for sure. I do it with a UPS that is unplugged (or if the scope has a battery pack) to power my scope for a few minutes to make a measurement. I can float the whole thing at high voltage, which is fine as long as you’re not touching it!!!! That is key. DO NOT TOUCH A FLOATING MEASUREMENT DEVICE! You can make a “chicken stick” for tweaking your scope safely (only if absolutely needed) with a NON-CONDUCTIVE fiberglass rod (DO NOT USE CARBON!) and some heatshrink, I’ll let you figure out the details… You’ll also want you scope either on the floating chassis itself or sitting on something that is non-conductive. again DO NOT TOUCH FLOATING MEASUREMENT EQUIPMENT!!!
My advice is that a good old DMM is your friend! if you are thinking of clipping your ground clip to something, simply measure between your clip and whatever it was you were going to clip to, everything powered as you’d expect for your measurement, if you see anything nasty, don’t clip to it! You can also do a simple resistance measurement between your clip point and ground, if it is below a few ohms off of ground, you should be fine.
I’ve also invest in some 100X probes at some point in the future. You can never have too many good scope probes.
I once vaporized several traces on a television main board because I didn’t observe precautions regarding earth ground when using an oscilloscope. The stench from blown components was indescribable and I’m d–n lucky the smoke cloud didn’t set off the fire alarm system in the dorm.
Since then (that was about 15 or so years ago) I always use an isolation transformer anytime I’m not sure of how anything is grounded. The time to set it up is practically zero and the savings in time/effort and safety is invaluable. I actually think it would be *better* if scopes came with an isolation switch to ensure that kind of thing never happens again. (oscilloscope companies, if you’re reading this and want a killer feature, you’re welcome).
Even so, when trying to measure the output of your wall wart, it’s not a bad idea to take a meter and verify that the output isn’t directly connected to the neutral wire before you hook it up. Though it’s against code, it wouldn’t surprise me much if some designer assumed that it would be safe to connect the neutral to the output on the wall transformer. No one would know that it was a problem until it somehow interacted with another earth grounded system.