Proper grounding for measuring wall AC signal with oscilloscope

[mosfet9]

Hi all,

I need to simultaneously record the AC signal from two separate wall 120v AC signals with a scope. I've got a metal enclosure with two step-down transformers (neutral and hot of each input go the transformer, and ground goes to the enclosure). The secondary of each transformer goes to a scope input.

I'm not sure the proper way to connect the grounds. My questions:

* The two inputs need to be able to be connected to different outlets, different circuit breakers, or possibly even different neighborhood transformers. Is it a good idea to connect only one of the ground connections (on my input wires) to my enclosure, or should they be connected together?

* On the output side of my transformers, (the input to the oscilloscope), the "+" signals on the oscilloscope probes will connect to one terminal on the secondary of each transformer. The oscilloscope input ground will be connected to the other terminal on the secondary of both transformers. Should this ground be connected to my enclosure ground and therefore to the ground in the wall plug(s), or should it float with respect to my wall ground?

* Alternately, I can leave the ground from the wall unconnected from my system, and ground my enclosure to only the oscilloscope input ground. Would there be any reason not to do this?

Thanks!

 

[MacGyverS2000]

You don't sound very familiar with electronics, so I highly suggest you find someone local who can help you with this... connecting a mains-powered circuit to a scope can be a quick trip to the morgue (for the equipment at a minimum, and for you at the worst).

Dan - Owner

http://www.Hi-TecDesigns.com

 

[waross]

Yes I agree with Gunnar. Grounding to two different transformer fed circuits simultaneously with a small conductor is an invitation to the insurance adjuster.
It may be more helpful to all of us if you tell us why you want to record waveforms (that may well be distorted waveforms through a small transformer). Someone may be able to give you an answer directly or at least direct you to some method of measuring safely and accurately.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[mosfet9]

Thanks for the quick answers. Some more details about what I'm doing: I need to simultaneously record PLC signals from two different powerline systems, which may or may not be on the same transformer.

Please let me know what other info you'd like me to provide in order for you to make a recommendation.

To simplify the question:
* are there any issues with connecting the ground of a scope input (which should be internally connected to the ground of the scope power cable) to the ground pin of a plug in my neighbor's house, for example ?
* it is not unsafe to scope the secondary of a step-down transformer when the primary is connected to neutral and hot, correct?

thanks for any additional advice you can give.

 

[Skogsgurra]

I assume (sorry again Jeff) that you are measuring Power Line Communication PLC and not Programmable Logic Controller PLC?

And, if that is so, not the Turtle type, which is rather tough to get a grip on, but those Echelon and similar signals in the 50 - 100 kHz band or thereabouts.

It is a bad idea to measure the complete waveform. You will have hundreds of volts at 50 or 60 Hz and very low kHz levels.

I use an isolation transformer on the signal (not power) side and put a blocking capacitor on the primary of that transformer. That way, I can get a solid waveform to analyze. See attached pdf.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[aginajm]

I thought you are trying to measure between hot and neutral of the transformer secondary. In your last post you mention connecting the ground pin of the o-scope to your neighbors house. Why?

You can safely obtain a measurement between the hot and neutral of the 120Vac mains at your house as long as:
1. The signal at the instrument input terminals is stepped down to within the specified input range.
2. The instrument is connected to the mains through an isolation transformer (this will prevent an accidental short circuit).
3. The isolation transformer is connected to a power bar with over-current protection.

[!]Attention![/!]
With respect to grounding you need to consider the following:
1. There is likely a potential difference (around a few hundred millivolts) between the ground at your house and your neighbors house. Connecting the two grounds (your and your neighbors) may introduce a ground current loop.
2. In some cases the potential difference between the two grounds can exceed several thousand volts (for example, a lighting bolt hits your neighbors house), a potentially fatal situation.

I recommend that you measure each location simultaneously with two separate instruments (one instrument located at each location). The data acquired can be synchronized through time stamping of the samples (or blocks of data) and then combined lated on. I doubt if your scope has this functionality. Check out ni.com, they might have the product for your application.

 

[peebee]

I'm a bit scared to answer your question and won't really attempt to do so. But I'd advise against connecting the grounds between two separate systems without giving that some serious consideration first. My advice would be to ground to only one of the two systems in question. If you are determined to bond the grounds from both systems, I'd recommend doing with the systems de-energized, and I'd use some big thick heavy bonding cable to do so -- cable roughly the same size as the utility service entrance cable feeding the panelboard. You might be surprised at the voltage difference between the two systems prior to bonding them (shock hazard when you try to connect them) and you might be surprised at the current flow through any bonding conductor between the two that you install (fire hazard).

Any possibility of putting an opto-isolator on each source?

 

[sibeen]

I'd suggest that hiring an oscilloscope with isolated inputs may be the best solution. Something like a Tektronix THS720P or similar.

 

[zappedagain]

How about using two oscilloscopes and then save the data and compare it later?

Z

 

[biff44]

I 2nd the fiberoptic thought. I would definately use the step down transformers, 115 to 12 VAC for example. I would get two cheap USB type oscilloscopes, hook one to circuit 1, and the other to circuit 2, and connect the DATA ports to a computer. If there is any distance between outlets, I would use a fiberoptic usb extender.

Hooking the chassis ground of an oscilloscope to one outlet, and having the scope plugged in with a 3wire plug into another outlet could easily blow the scope, and if the house is miswired (more common than you would think) could be deadly. Even using a step down transformer might not save you, as there can be leakage currents thru the windings.

http://www.MaguffinMicrowave.com

Maguffin Microwave wireless design consulting