Safest way to use an oscilloscope ?

[ytterligare]

I've realized that I seldom use an oscilloscope, and I regret not exploring its capabilities further. I want to delve deeper into diagnosing and probing various parts of circuit boards. However, being aware of my tendency to make mistakes, I'm concerned about potentially causing damage by shorting the wrong terminal and creating an unexpected electrical event, like "BOOM!" 

To address this concern, I'm considering three options to enhance safety, for which I'd like your opinion :

1) Isolation transformer to "unground" the scope
2) Using a differential probe
3) Using a portable ( battery powered ) oscilloscope.


I hope my question does make sense, please just in case try to connect the missing dots....
Thank you
A.

[bdunham7]

4) don't go exploring high energy things like mains powered circuits until you have more experience and skills, as well as the proper probes and safety equipment.

[JustMeHere]

Run your DUT from a battery....or actually most switched mode wall adapters.  They  are by nature isolated.

A bench power supply.is probably the best way to go.  It can be set up to protect your circuit from a fault.

[BrokenYugo]

1. No, you never "float" the scope, if anything the device under test is isolated.
2. Maybe sometimes, depends what you're doing.
3. Not really.

It's unfortunately a bit of a learning curve, but you really have to understand what the scope/probe represents electrically and how adding that circuit to the circuit under test will work (or fail, perhaps with energetic results). Like draw the probe schematic, any earth reference Included, onto whatever you're working on and ask yourself if it makes sense. It's something you have to evaluate on a case by case basis,  you pick a node of interest and then figure out if you have means to probe it safely. Like with basic probes you must at least know that the ground clip/spring isn't going to cause a short, at higher voltages and frequencies high frequency derating of the probe (see data sheet included with probe for curve) must be done, etc.

The advice to avoid mains/high energy in general for now is good advice, you must be certain of your actions when mistakes can harm yourself or costly equipment.

[coppercone2]

make a god damn work plan with a multimeter on where you can probe and then its safe

the dangers you are worried about come from defective clips and haphazard operation.

its when you feel pressed for time then you should stop working on something dangerous

[tggzzz]

I've realized that I seldom use an oscilloscope, and I regret not exploring its capabilities further. I want to delve deeper into diagnosing and probing various parts of circuit boards. However, being aware of my tendency to make mistakes, I'm concerned about potentially causing damage by shorting the wrong terminal and creating an unexpected electrical event, like "BOOM!" 

To address this concern, I'm considering three options to enhance safety, for which I'd like your opinion :

1) Isolation transformer to "unground" the scope
2) Using a differential probe
3) Using a portable ( battery powered ) oscilloscope.

Good to see someone thinking, so they can "have fun, safely".

1) Never ever float a scope by disconnection the "protective mains earth". It can cause loud noises, bad smells, and deathly silences

2) HV differential probes are often a good choice for probing high voltage and/or where neither points being probed are at earth potential

3) a ordinary scope powered by a battery is no better than a floated scope. A few scopes are specially designed with internal isolation between channels and protective mains earth, and some of those can be battery powered

4) don't probe "high energy" circuits. High energy usually involves high voltages, but a 12V car battery can deliver high current sufficient to melt metal.

FFI, have a look at the references mentioned: https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Even after all that, pause and think It is all too easy to make a "stupid" mistake when tired or rushed

[ytterligare]

Thank you guys, you have all understood the point well ( sorry for my english as I'm not native speaker )
I might have been biased ( and scared ) by a video posted on YT by Dave some times ago, but better exceed in caution than reconstructing what happened while healing wounds from second degree burns on the flesh.

Anyhow, ungrounding a scope /portable scope, seems to be a bad choice, although not everyone has the same clarity of vision as you guys,  as I clearly remember having read somewhere these were viable options : at this point, more than me having little knowledge and experience, it seems that tangible troubles comes from who's spread partially true or false knowledge for the sake of showing off.

Two sparse thought.
It seems that a differential probes might be a "condom" to avoid troubles, although I suspect that the result might not be as accurate as using normal passive probes : anyhow, this comes at a cost, as diff. probes at 150 Mhz are in the 350 Euro ballpark, but they could be an introduction then to high power devices territory, to be avoid at the beginning.

Also, I get from Dave's video (*) other than floating the scope, which should remain connected to ground, better float the DUT with the aforementioned isolation transformer ( or USB isolator in case of PC USB powered devices )

Last (debatable ) thought : with experience, mistakes count will became thinner, but maybe having some power sockets attached to an isolation transformer in the lab, dedicated to the DUT tobe probed, is not a bad idea....

(*) Dave's video :

[coppercone2]

i almost wanna pay you 350 euro to not see another floating oscilloscope thread


clearly your interested in high voltage. get the probes and you might actually accomplish something rather then endlessly worrying about all the BS

I got one for you. With a probe you get used to using a probe. With a stupid isolation capability you might be in a rush to fix something and use the wrong socket. Everyone acts like they always are correct with the power plugs. Clearly the amount of destroyed side cutters and wire strippers show that man is not good at determining the correct cord.

This is what is gonna happen

your gonna do it alot then have two IEC cables, so you can easily switch it near the scope, then your gonna plug in the wrong one or forget which one was connected when there is some kind of problem you are frustrated about. or get hurt by a damaged power cord or touch the prongs.

you know when the first hard circuit comes along and you are steaming mad that you fried hundreds of dollars of chips, if you actually do something challenging.

[tggzzz]

Thank you guys, you have all understood the point well ( sorry for my english as I'm not native speaker )
I might have been biased ( and scared ) by a video posted on YT by Dave some times ago, but better exceed in caution than reconstructing what happened while healing wounds from second degree burns on the flesh.

Anyhow, ungrounding a scope /portable scope, seems to be a bad choice, although not everyone has the same clarity of vision as you guys,  as I clearly remember having read somewhere these were viable options : at this point, more than me having little knowledge and experience, it seems that tangible troubles comes from who's spread partially true or false knowledge for the sake of showing off.

That is a sensible attitude.

There are many people, especially on yoootooob that see something, half understand it, decide to make an "exciting accessible" video about it. They end up very confidently making statements equivalent to "I've run into the road without looking many times. Therefore it isn't dangerous. Therefore you can do the same".

Knowing how to spot such people is a valuable life skill.

[pcprogrammer]

Knowing how to spot such people is a valuable life skill.

Here is an example for you. Nobody objects, so I'm right. 

About using a scope, always keep your focus on what you are trying to do. For measuring on low voltage systems, lets say up to 40V AC or DC, using a standard probe is fine. Make sure that the probe is on 10x range if it is switchable and look for the proper ground in the device under test. Here you connect the ground clip of your probe and use the tip of the probe to check the signals. Most of these devices will have some sort of transformer in them to have a galvanic separation from the mains.

The moment you are looking into measuring on systems with higher voltages look for solutions like differential probes or special high voltage probes, but still make sure about what is what in the device under test. Don't just go probing into some system you know nothing about. Nothing good will show up on the scope anyway if you don't know what to expect.

So first find all the necessary information about what you are going to measure on, and then take appropriate actions based on that information.

[tggzzz]

Knowing how to spot such people is a valuable life skill.

Here is an example for you. Nobody objects, so I'm right. 

Oh. Him.

The recent deluge of crap suboptimal posts on this forum means I've recently started populating my "ignore poster" lists   and

Unfortunately all that does is hide the contents of posts; I'd prefer it to hide the existence of posts and replies to the posts. Bring back usenet readers

[nightfire]

Option 3) is from theory a good one. But: For this you need some industrial-grade unit, such as the scopemeters from Fluke- they are really isolated resp. giving you isolated channels. Most of the cheaper units that also offer USB connectivity are not isolated , so you will get those nasty ground loops descibed in Daves Video linked above.

For a beginner, that explores the possibilities of scopes, probably the combination of a decent entry-level scope (maybe used) and, depending on the measurement scenario, a differential probe should be the best and cost-efficient way to go.

I myself have bought an entry-level Rigol scope and a micsig DP10007 probe is high on my shopping list for this year.

[audiotubes]

Don't forget Dave's point about USB being a surprise.

It seems that the issue is putting the scope ground clip on something that really isn't at ground. I don't know why anybody would do that though. Probably it would help if there was a little explanation on how to use a handheld multimeter to make sure what you think is ground is really ground, before attaching the scope ground clip.

I know Dave covered that a bit but it was regarding how earth ground may wind up on your board in an unexpected way. For those of us getting started (or in my case restarted, last time I used a scope seriously was in the 1970s) it would help to show a safety procedure explicity.

[bdunham7]

It seems that the issue is putting the scope ground clip on something that really isn't at ground. I don't know why anybody would do that though.

You might want to look at a signal between two points where neither of them is at ground.  The way you'd do it with an isolation transformer is that you'd isolate the DUT, then intentionally ground one of those two signal points and then look at the signal.  Depending on the exact circumstances this may work really well and that is why you'd want to do it.

[tggzzz]

Don't forget Dave's point about USB being a surprise.

It seems that the issue is putting the scope ground clip on something that really isn't at ground. I don't know why anybody would do that though.

Consider an H-bridge motor driver, and you want to be able to monitor fast transients on the high-speed transistor, V_be or V_gs as appropriate.

Where would you put the probe shield? Don't forget CMRR at high frequencies!

[audiotubes]

Yeah I meant without an isolation transformer or differential probes.

[bdunham7]

Yeah I meant without an isolation transformer or differential probes.

OK, in that case the reason you'd want to connect your scope ground to a live circuit point is because your name is Electroboom and you're making another bad comedy YouTube video.