Probing mains with an oscilloscope

[Elandril]

Since I'm experiencing some strange fluctuations on a few dimmer channels in my home installation (240V EU mains), I'd like to see if I can probe those channels to get an idea what kind of noise/problem I'm dealing with. So my idea was to connect my oscilloscope (Rigol 1052E) with a 10:1 probe to have a look. Since I'm completely inexperienced with this, I'd like to know what is "best practise" to do this safely and if I have to be cautious about something?

I would appreciate any input from you pros here. :-)

PS: The dimmer is a 4-channel KNX/EIB bus driven one from GIRA. The flickering happens on 2 channels with dimmable energy saving lamps and can be slow (1 flicker every 10 secs) or fast (up to ~10 per sec).

[David_AVD]

Can you add an incandescent load in parallel with the offending channel outputs for testing purposes?  Some TRIAC dimmers don't like non resistive loads.

Also, is the dimmer a leading or trailing edge type and are the lamps specified for leading and/or trailing edge dimming?

[Elandril]

Can you add an incandescent load in parallel with the offending channel outputs for testing purposes?  Some TRIAC dimmers don't like non resistive loads.

I'll try this. Thanks for the suggestion.

Also, is the dimmer a leading or trailing edge type and are the lamps specified for leading and/or trailing edge dimming?

The dimmer is actually capable of both modes and even has an automatic detection feature. I've actually tested both modes, but the situation did not really change/improve. The lamps are marked as "capacitive dimming load" type, so I would expect the trailing edge dimming to be the best mode.

[amyk]

http://www.eevblog.com/2012/05/18/eevblog-279-how-not-to-blow-up-your-oscilloscope/

[AndyC_772]

http://accessories.picotech.com/active-oscilloscope-probes.html

[vk6zgo]

Since I'm experiencing some strange fluctuations on a few dimmer channels in my home installation (240V EU mains), I'd like to see if I can probe those channels to get an idea what kind of noise/problem I'm dealing with. So my idea was to connect my oscilloscope (Rigol 1052E) with a 10:1 probe to have a look. Since I'm completely inexperienced with this, I'd like to know what is "best practise" to do this safely and if I have to be cautious about something?

I would appreciate any input from you pros here. :-)

PS: The dimmer is a 4-channel KNX/EIB bus driven one from GIRA. The flickering happens on 2 channels with dimmable energy saving lamps and can be slow (1 flicker every 10 secs) or fast (up to ~10 per sec).

Oscilloscope cases (& their probe earth clips) are normally connected to the mains EARTH,which is,in turn,connected to the NEUTRAL line at the incoming power connection to your house.("the "fusebox")

If,when testing a pair of Mains connections,you inadvertently connect the probe clip & earth lead backwards,you will directly connect the ACTIVE line to the NEUTRAL line,via the Oscilloscope& the EARTH connection,putting a short across the Mains,hopefully blowing the Mains fuse,but also possibly  blowing the earth connection inside the 'scope.
Even if you get it right,it is illegal in most countries to have any other connection between EARTH & NEUTRAL
in addition to the primary connection at the house "fusebox".

In your case,you are not worried about high frequency information,so you should be able to use the following simple method:

Don't connect the probe earth connection to anything,just use the probe tip(remove the earth clip lead,or tie it back to the probe lead)------the other side of the connection will be provided via the 'scope EARTH connection via its power cord.

[Elandril]

Thanks for the suggestions, but I'm coming up with even more questions... ;-)

In your case,you are not worried about high frequency information,so you should be able to use the following simple method:

Don't connect the probe earth connection to anything,just use the probe tip(remove the earth clip lead,or tie it back to the probe lead)------the other side of the connection will be provided via the 'scope EARTH connection via its power cord.

So I could use two channels in diff math mode by connecting CH1 to the mains phase L1/L2/L3 and CH2 to the mains N (while shorting or omitting the two GND leads)? This should effectively measure L-GND and N-GND which are then combined to SIG=(L-GND)-(N-GND)=L-N, right? I mean with the 1M resistance of the oscilloscope plus the 9M from the prove the max current between the probe tip and GND should only be ~ 24 microAmps which should not trigger the RCD protection.

The whole setup is a bit more complicated, since my home has multiple RCDs: one for each floor, one each for garage and outside lights, one for the hobby workshop, and one for devices in the electric enclosure (where my dimmer sits) plus one last-resort-if-everything-fails general RCD that comes directly after the point where the mains (L1-3 + PEN) enter my home. So the affected dimmer channels are on different neutral lines in respect to each other and the oscilloscope would probably also be connected to a different neutral line. Would this lead to any problems?

Would an isolation transformer be of any help?

[Bored@Work]

Watch the video amyk referred to.

[Elandril]

I did watch the video, but unfortunately, Dave doesn't mention the diff mode approach, so I'm still kinda unsure, if this is "safe enough" to do. I principle, if I understood oscilloscope+probe internals correctly, I should work a treat, right?

[alm]

Yes, that would work fine. So would using an isolation transformer and put the device under test (eg. dimmer) on the secondary side. Just don't put the scope on an isolation transformer to defeat its ground connection.

[vk6zgo]

Thanks for the suggestions, but I'm coming up with even more questions... ;-)

In your case,you are not worried about high frequency information,so you should be able to use the following simple method:

Don't connect the probe earth connection to anything,just use the probe tip(remove the earth clip lead,or tie it back to the probe lead)------the other side of the connection will be provided via the 'scope EARTH connection via its power cord.

So I could use two channels in diff math mode by connecting CH1 to the mains phase L1/L2/L3 and CH2 to the mains N (while shorting or omitting the two GND leads)? This should effectively measure L-GND and N-GND which are then combined to SIG=(L-GND)-(N-GND)=L-N, right? I mean with the 1M resistance of the oscilloscope plus the 9M from the prove the max current between the probe tip and GND should only be ~ 24 microAmps which should not trigger the RCD protection.

The whole setup is a bit more complicated, since my home has multiple RCDs: one for each floor, one each for garage and outside lights, one for the hobby workshop, and one for devices in the electric enclosure (where my dimmer sits) plus one last-resort-if-everything-fails general RCD that comes directly after the point where the mains (L1-3 + PEN) enter my home. So the affected dimmer channels are on different neutral lines in respect to each other and the oscilloscope would probably also be connected to a different neutral line. Would this lead to any problems?

Would an isolation transformer be of any help?

Well,actually,I thought you just wanted to look at one dimmer channel with one 'scope channel,which is why I suggested the method I did.
It is really just measuring between ACTIVE & EARTH instead of ACTIVE & NEUTRAL,but using the 'scope internal earthing as the EARTH connection.

The "high frequency" comment was due to the fact that the long connection back to NEUTRAL via the house wiring can severely attenuate any higher frequency components in the observed voltage.

Your suggested method is a better one,but the simple one may be sufficient.
Many times during troubleshooting inside equipment,you find that access to both ACTIVE & NEUTRAL at the same time is difficult,but it is possible to reach each side one at a time  & test them w.r.t EARTH.

If you do this with an old analog multimeter you get to test the RCD too!

[The Electrician]

You can just connect a small power transformer across the mains where you want to observe.  I used a 12 VAC out, 30 VA transformer.  I used a Tektronix TPS2024 scope which has isolated inputs, so you can connect the inputs directly across the mains safely.  I wanted to compare the directly observed signal with the output of the transformer to verify that the frequency response of the transformer would be adequate.

I have a triac dimmer in series with an incandescent lamp.  The scope is connected directly across the lamp (orange trace) and the other channel (blue trace) is looking at the 12 volt secondary of the transformer which is also connected across the lamp.



Here is the same rising edge at a much faster sweep speed.  The rise time of the dimmer output can be seen to be only a little slower at the transformer output.



I then made a simple high pass filter consisting of a 200 pF capacitor and a 1k resistor to eliminate the 60 Hz component of the line voltage.  I then connected the scope across the 1k resistor to see what kind of RF was present on the line.  This image shows the FFT of the high passed line.  Mainly what I see are the local AM radio stations, greatly attenuated.



So, you can just use a small power transformer to isolate and protect your scope.

[Elandril]

Wow, thanks a lot going to all this efford, The Electrician!

In this case, I think that I'll prefer this method. I hope to find the reason for the flickering this weekend.