Monitoring power line conditions

[Skaperen]

I would like to do some monitoring of power line conditions. To start with, I just want to monitor and record the voltage wave form on the two split phases. I would dedicate a small computer to the purpose. The idea is to reduce the voltage from the power line in a safe way and feed that through the audio input of the computer. A program would then capture that audio input, compress it in a lossless manner (a 60 Hz sine wave should be very compressible), and store it on disk space to be fetched by another computer over the network on a regular basis.

I'm not sure what I will get out of it. But I just want to tackle the issue of getting it in a reliable and safe way for now. In particular, I want to know what method might be best for reducing the voltage and still not alter the waveform any more than would be acceptable. Would a transformer be adequate for this, especially if it is kept well away from the saturation level? Or would it be best to have a resistor divider network?

I don't even know what impedance and levels the audio inputs have, yet. But I know I don't want to go buy the off-the-shelf device I once saw, since it could not be automated as I would want (had to manually move a memory device from the unit to the storage computer, which would apparently also mean a gap in measuring), and it would cost $3500.00 for this inconvenience. I figure I could do better with $50 of parts and an old used netbook worth maybe $100 today (and feel safe about fetching the data via wifi).

One motivation to do this is that at different times, I detect behavior of different devices changing, presumably due to changes in the electricity. Sometimes my UPS clicks over to battery and back off again. Was there a transient? A couple lost cycles? Sometimes my lights just dim for a while. Sometimes my fan in the floor pumping air into a small room I use for a couple computers will slow down and speed up. These different things do NOT happen at the same time so I am curious about this. Voltage change? Transient? Frequency change?

Any suggestions to set up power line monitoring like this?

 

[berkeman]

I would like to do some monitoring of power line conditions. To start with, I just want to monitor and record the voltage wave form on the two split phases. I would dedicate a small computer to the purpose. The idea is to reduce the voltage from the power line in a safe way and feed that through the audio input of the computer. A program would then capture that audio input, compress it in a lossless manner (a 60 Hz sine wave should be very compressible), and store it on disk space to be fetched by another computer over the network on a regular basis.

I'm not sure what I will get out of it. But I just want to tackle the issue of getting it in a reliable and safe way for now. In particular, I want to know what method might be best for reducing the voltage and still not alter the waveform any more than would be acceptable. Would a transformer be adequate for this, especially if it is kept well away from the saturation level? Or would it be best to have a resistor divider network?

I don't even know what impedance and levels the audio inputs have, yet. But I know I don't want to go buy the off-the-shelf device I once saw, since it could not be automated as I would want (had to manually move a memory device from the unit to the storage computer, which would apparently also mean a gap in measuring), and it would cost $3500.00 for this inconvenience. I figure I could do better with $50 of parts and an old used netbook worth maybe $100 today (and feel safe about fetching the data via wifi).

One motivation to do this is that at different times, I detect behavior of different devices changing, presumably due to changes in the electricity. Sometimes my UPS clicks over to battery and back off again. Was there a transient? A couple lost cycles? Sometimes my lights just dim for a while. Sometimes my fan in the floor pumping air into a small room I use for a couple computers will slow down and speed up. These different things do NOT happen at the same time so I am curious about this. Voltage change? Transient? Frequency change?

Any suggestions to set up power line monitoring like this?

The easiest starting point would be to use two AC "wall wart" transformers like in the picture below. You don't need a high current one like in the picture, though. You will likely want to attenuate the output down to whatever the sound card's input voltage range is.

Sounds like an interesting project. You may want to look at other commercial AC Mains power monitoring equipment datasheets, to see what kinds of things you should look for (surge, sag, cycle dropouts, etc.).

 

[jim hardy]

""Would a transformer be adequate for this, especially if it is kept well away from the saturation level? Or would it be best to have a resistor divider network?""

A transformer such as Berkeman suggested gives you a VERY significantly safer hookup.

You mentioned avoiding saturation, which is a good idea to mantain fidelity.
To that end if you could find 230 volt step down transformers, on 115 they'd be operating at half flux.
Furnaces use a 230::24 control transformer. They're not prohibitively expensive at a good hardware store.



In my power plant in early 1970's we used magnetic tape recorders for the monitoring you decribe.
As a customer you'd probably only see off-frequency immediately before a huge blackout. Should you capture one the utility will probably be interested in your data especially if you can synchronize your time to WWV. But you'll need a UPS..
Voltage sags and missing cycles from line switching or faults is probably what you'll detect most often.

 

[Skaperen]

The easiest starting point would be to use two AC "wall wart" transformers like in the picture below. You don't need a high current one like in the picture, though. You will likely want to attenuate the output down to whatever the sound card's input voltage range is.

Looks like a good idea. How can I verify that these devices do not include things that might alter the signal accuracy, like filters? Ideally a straight transformer with nothing added would seem right. I do want a wide bandwidth, at least to the Nyquist limit on my current sampling rate (48 kHz sample rate, so 24 kHz bandwidth). That should not be saturation issue (even below 60Hz as long as it is not a strong signal).

Another idea I thought of since posting was to build some kind of self contained sensor that could isolate the hazards better, and have it do PWM or PCM of an LED to pass on the voltage readings. I understand some LEDs can be modulated on/off well past 1 MHz. I'd then need some kind of device to pick up this light and pass it into the computer. Maybe a transformer box like shown could be the host for that LED (removing the LV cord).

 

[Skaperen]

Furnaces use a 230::24 control transformer. They're not prohibitively expensive at a good hardware store.

Good idea. I was looking at some buck-boost transformers, as they usually have a pair of 120V windings for primary that can be wired in series. But those are typically well north of $100.

Something else I thought of are those voltage sensors people (especially electricians) get to detect presence of AC voltage real close. I don't yet know how those pick up the voltage. I'm curious if that could be used with any accuracy of waveform.

As a customer you'd probably only see off-frequency immediately before a huge blackout. Should you capture one the utility will probably be interested in your data especially if you can synchronize your time to WWV. But you'll need a UPS..

Maybe the netbook battery will hold it up long enough to get a good recording.

Hey, I did see some voltage/current phase diagrams of certain grid points during that big 2003 east coast blackout. Crazy.

Doing WWV was something I was also thinking about for another project (have the audio system emit a carrier fed to an external circuit to step it up to a harmonic close to WWVB, mix that from the antenna, do low pass, and bring it back to the audio input, and average the carrier over days in software).

Voltage sags and missing cycles from line switching or faults is probably what you'll detect most often.

Like the sag that happens many nights right around 12:30 plus or minus 20 minutes, and lasts for 10 to 20 minutes then resumes normal. I haven't gotten a voltage reading on it, yet. My little floor fan slows down enough to notice the change in pitch. Yet the lights do NOT dim (OTOH, they are all have dimmers, so that may be voiding that).

One other thing I do want to monitor is the surges during storms. These cause a very noticeable increase in brightness (the dimmers don't seem to stop that), and when we had old CRT based TVs, they affected them as well. I've generally written these off as one of the other two phases going to ground in the MV distribution. But I'd be curious if I can see any fluctuations in the waveform indicating arcing.

 

[jim hardy]

""Hey, I did see some voltage/current phase diagrams of certain grid points during that big 2003 east coast blackout. Crazy.""

indeed - people don't think of the grid as the flexible, dynamic thing that it is.

"" Nyquist limit on my current sampling rate (48 kHz sample rate, so 24 kHz bandwidth). That should not be saturation issue ""
oops - i thought you were speaking of magnetic saturation in the transformer cores. That's why i suggested the furnace transformers operated at half voltage.
Amazing, how fast one becomes obsolete . You mean a Nyquist isn't a delicious tropical fruit?

""Another idea I thought of since posting was to build some kind of self contained sensor..."
There exist operational amplifiers with isolation between input and output for applications like this.
http://www.analog.com/static/imported-files/tutorials/MT-071.pdf
Several other brands out there so look around.

""... and have it do PWM or PCM of an LED to pass on the voltage readings. I understand some LEDs can be modulated on/off well past 1 MHz. ""
FWIW:
The analog tape recorders we used in early 1970's were FM, where the signal to be recorded modulated an FM carrier in the audio range and that carrier was recorded. Playback was by FM demodulator.
The carrier frequencies it used were in accordance with IRIG B standards i think around 7khz. They gave acceptable fidelity on playback. I built one channel for a special measurement, used LM566 Phase Lock Loop VCO for V/F modulator and LM386 to drive the tape head directly. It could have as easily driven a LED.
Point being if you're a creative tinkerer you can do this with simple equipment .
Just keep it safe - isolate from those power lines, and fuse your sense lines..


Good luck !

old jim

 

[Skaperen]

Oh, I did mean magnetic saturation of the core, too. I don't want a 50% swell to be distorted.

FM might be a workable method, too. That should not be too hard to "decode". But something that can give me a lot in a limited bandwidth would be good, too.

The smaller I can make the power sensor, the less is exposed to possible power damage. The more isolated it is, the less the risk of a shock hazard somewhere. A small circuit that can fit inside a small wall wart, with an LED on the front to emit signal, would be good isolation. I was thinking along such lines once when I wanted to do a current transformer, given the even greater hazard these things can have if they go open circuit (fuses don't play well with these).

 

[jim hardy]

you're well on your way.

Are you in a part of the world where you could get 230 volt wall-warts with low AC out?

Ahhh ebay...

 

[Skaperen]

you're well on your way.

Are you in a part of the world where you could get 230 volt wall-warts with low AC out?

Ahhh ebay...

Not readily. Would have to order from one of those "220 appliance" places, or from EU (whoever can ship to USA w/o VAT).

I don't do the Ebay thing anymore. Too risky.