Oscilloscope probing mains AC DUT questions

[mart1n]

Hello Folks,

I've been wanting to learn more about AC circuits, antique electronics and power supplies. I have a few switch mode power supplies that I've started to study/analyze that are mains earth referenced that I would like to probe with my scope.
I just finished watching the EEVBlog episode on how not to blow up your scope (Thanks Dave!) and have a quick question that I want to verify to maintain my safety.

First off, I don't yet have a differential probe or an isolation transformer.

Question #1:
If I'm extra careful to only attach my ground lead clip to a mains earth referenced rail, I should be safe to then probe the circuit, is that correct? Is there anything else I need to be careful of?

Thanks!

[james_s]

No no NO, just don't.

At the bare minimum an isolation transformer is *mandatory*, this is not optional. A differential probe is highly encouraged but don't even think about poking around inside a live mains powered switchmode PSU without isolation. Really you should avoid live testing at all unless absolutely necessary.

You can buy small isolation transformers for under 20 bucks from arcade machine parts suppliers as most of the older CRT monitors required them. Seriously, this is for your own safety and that of your equipment, don't try to take shortcuts, these things are lethal. A switchmode PSU is much more dangerous than the high voltage on a CRT, the 340V rail can pack a hell of a punch.

[nigelwright7557]

SMPS often have a bridge rectifier on the front end which means trying to probe anything after that will be just garbage on the scope screen.
You need an isolation transformer..
Then you can connect scope ground to bridge -ve end out and probe circuit and get good results.

You really do have to play safe even with an isolation transformer.
Always turn off and wait for caps to discharge before touching the circuit or attaching/un-attaching scope probes.

On high voltage circuits I always added an LED and high value resistor across main cap to show when it is still charged.
Not fool proof but better than nothing.

[tggzzz]

Firstly, an isolation transformer doesn't make the setup safe. It can also interact badly with GFCIs.

Secondly, if you have really understood the failure mechanisms, use an isolation transformer on the DUT. Never ever float the scope.

Thirdly, try and find another way of achieving your ends.

Finally, read and understand the safety references and probe types at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

[mart1n]

Thanks for all the tips, warnings and information!
I still feel like my question wasn't answered though...
Assuming I never actually connect my scope probe(s) to an AC mains earth referenced DUT without proper safety equipment (isolation transformer, differential probe), I would
still like to know the answer to my questions for the purpose of technical understanding.

Scenario #1:
With the DUT powered off and any component able to store energy discharged...
I use my DMM to test resistance between test point A and the earth pin on the DUT's power cable to verify that test point A
is indeed a direct connection to mains earth. With the DUT still powered off, I connect my scope probe's ground lead to test point A.
I connect the scope probe to some other point in the circuit and then plug the DUT's power cable directly into a wall outlet.

Assuming mistakes were not made during the setup of Scenario #1 (I understand this is the point of the safety gear as mistakes do happen).
We're also assuming that we power down and discharge the DUT if/when we move the scope probe to different test points, while maintaining
test point A as our ground lead location.

Scenario #1 Questions:

• Is this "safe"?
• Are all probe points "safe"? (Assuming we're not moving the ground lead)
• If some probe points are not "safe", what would those be and why?

I think the answer to these questions are that this is OK. I couldn't think of any reasons why it wouldn't and that's why I'm asking.
Not to take short-cuts with safety gear, but because I want to solidify my understanding.

FYI - I've purchased a bench isolation transformer from Jameco since first posting this question.

[bob91343]

While it's true that you are okay with using a common ground of the power mains, your oscilloscope displays won't be too useful, as one has mentioned.  Also, a fault on the line (even remote from your test station) could lift the ground enough to cause problems.  It's not worth risking blowing up your self or your gear.

You can cobble up an isolation transformer with a pair of back to back transformers.  Just respect the power ratings.

[bdunham7]

Scenario #1 Questions:

• Is this "safe"?
• Are all probe points "safe"? (Assuming we're not moving the ground lead)
• If some probe points are not "safe", what would those be and why?

I think the answer to these questions are that this is OK. I couldn't think of any reasons why it wouldn't and that's why I'm asking.
Not to take short-cuts with safety gear, but because I want to solidify my understanding.

FYI - I've purchased a bench isolation transformer from Jameco since first posting this question.

If  you use a proper probe, I think that is safe.  A proper probe might be a 10:1 10M or 100:1 100M rated for whatever your expected voltages are. 

An isolation transformer is a must for the test bench, IMO, but it will not necessarily improve the safety of a ground-referenced device and may give you a false sense of security.  As previously mentioned, it may also actually defeat a GFCI that would otherwise protect you. 

If you don't need a very high frequency response, it is a bit safer and more convenient to run a separate ground lead and just leave the ground lead off the probe itself.  Then you can probe and poke with the single probe (and one hand!) with relatively little chance of anything going wrong.  Keep in mind that you won't be able to measure a lot of things, especially in a SMPS, because the measurement you may need will not be referenced directly to ground.  A HV differential probe or isolation amplifier made just for this situation is really the only way to handle it properly.

[tggzzz]

If you have fully understood and fully described the situation, then you may be right. But if you haven't then you should, at the very least, read learn and inwardly digest this post:
https://www.eevblog.com/forum/projects/isolation-transformer-for-scope/msg2259465/#msg2259465

So, how lucky are you feeling?

General point applicable to far more than this subject... The many apparently arcane parts of safety regulations are the response to people having been injured due to subtle and surprising causes. They contain a lot of hard won knowledge.

[tautech]

Hello Folks,

I've been wanting to learn more about AC circuits, antique electronics and power supplies. I have a few switch mode power supplies that I've started to study/analyze that are mains earth referenced that I would like to probe with my scope.
I just finished watching the EEVBlog episode on how not to blow up your scope (Thanks Dave!) and have a quick question that I want to verify to maintain my safety.

First off, I don't yet have a differential probe or an isolation transformer.

Question #1:
If I'm extra careful to only attach my ground lead clip to a mains earth referenced rail, I should be safe to then probe the circuit, is that correct? Is there anything else I need to be careful of?

Thanks!

OK it might seem you're at the point where you want to investigate high voltages and that's a perfectly natural progression from working on LV circuits but not without dangers to both you and your scope.
Yes mains earth is at the same potential as the reference lead of your probe however it's wise to always check with a DMM in case the mains protective ground has a break somewhere or it's not properly wired.

100x probes are wise for increased safety for both yourself and your scope and it's not like they are costly like a differential probe however to get very far doing work on SMPS you should get one of them too.

Working on elevated voltages requires a clear head and NO distractions and most importantly sound methodology to prevent accidents. Old timers would keep one hand behind their back or in a pocket so to not expose the risk of an electrical path across the chest and heart and this is still wise practice today.

Stay safe and be careful !

[Brumby]

I can only echo the comments above.  I would also like to reinforce the point that if the scope and the DUT are referenced to the same earth, the level of risk is lowered only if that relationship is maintained.

All too often, things you want to inspect can have, or can develop, a fault that compromises this relationship.  Were such a fault to occur, it might well be non-obvious in the devices normal operation or using a DMM.

With a DMM, you can get away with this sort of thing where the DMM is a floating instrument.  Most bench scopes are not.


Just remember this ancient saying: "One flash and you're ash."  It applies directly to the scenario you describe.

[calzap]

Martin,

In Scenario #1, you’re making an assumption that point A remains connected to earth ground when the DUT is powered-up.   All it would take is the operation of a DPDT relay in the DUT to swap point A and the point where the probe is attached.  There are lots of other circuit scenarios where things change a wee bit when power is applied.

If the DUT is mains-powered, at least, power it via an isolation transformer.  Oh, and check with a DMM that the secondary of the isolation transformer really is isolated from the primary before you use it.  Not every transformer sold as an isolation transformer is designed for true electrical isolation.  Some are sold to isolate the output from noise on the input lines, but the secondary is not electrically isolated from the primary.  One of the secondary lines is electrically connected to the metal case which is connected to mains ground, which in turn, is connected to the neutral in the service entrance panel.

Mike in California