POWER QUALITY METER

[aswin p ajayan]

Hai friends
I am trying to build a power quality meter to measure supply fluctuations and harmonics. My problem is that i cant find a proper circuit process the supply so that it can be supplied to pic's ADC pin. Could i use a tranformer to step down the supply . If i do so will it block the higher harmonics .

please help me even with the wildest of your ideas

yours
aswin

 

[alec_t]

For safety's sake you must use a transformer, assuming the supply you are talking about is the mains supply. But all mains supplies have a lot of interference superimposed on them; how will you distinguish that from harmonics?

 

[aswin p ajayan]

thank you alec
i was thinking about fourier analysis so that i can display the power spectrum in a glcd screen.
but i was afraid wether the inductance of transformer will block the noise spikes.

if it does what alternative i can have

 

[alec_t]

Can't think of any alternative which would be as safe.
I'm curious as to why you'd want to check for harmonics? Most of us simply have to assume their presence in today's world.

 

[aswin p ajayan]

thanks again for showing up
i am an electronics student and doing this project as a part of my curriculum. I want to build a power quality analyzer ,which should detect
volltage sag ,swell and presence of harmonics and glitches in supply, and record their time of occurrence and magnitude . For the project part it's just enough for me to show the power spectra on a glcd. As far as know these machines are commercially available and are used to ensure quality of power supply.

 

[alec_t]

I've just checked the output from a small standard mains 12V transformer, by feeding the output via an attenuator to my laptop's mic input and recording a wav file using Audacity. Here's the FFT as generated by LTspice, clearly showing many harmonics plus noise and so proving it's ok to use a transformer for your project.

 

[aswin p ajayan]

you are wonderful alec . You don't know how much this mean to me!
I was just going through a circuit using op-amp in summing amplifier config . just thought if i placed
an opto-coupler at its output i could isolate it . circuit diagram is attached with this .
wish to hear your opinion about it.
Ri = 150k
Rf = 1k

 

[audioguru]

What is wrong with your electricity supply that causes its voltage to sag and swell? Do you live on a farm that is hundreds of miles from the generating station?
My electricity in the city has rock solid voltage that never has a brown-out and is never too high.

Also I have never had a problem with harmonics on my electricity supply. I would use a transformer to "break in" a new speaker for a few hours. I never heard loud harmonics but I heard a click when something powerful turned on or off.

 

[aswin p ajayan]

its not about supply at my home. its just about the project . You mentioned about the cliks heard when switching on some devices , now think of an industry with a lot of heavy machinery
.this type of meters are used to measure power fluctuations caused by such machines

 

[canadaelk]

An analog optocoupler (like th HP HCNR2000 or HCPL4562) would give you isolation with good linearity and bandwidth. Read app. note AN107 at www.ixys.com E

 

[crutschow]

i was thinking about fourier analysis so that i can display the power spectrum in a glcd screen.
but i was afraid wether the inductance of transformer will block the noise spikes.

Any power transformer you use should have sufficient frequency response to pass all the spikes and fluctuations of interest. The primary inductance of the transformer determines the magnetizing current, it does not affect the frequency response of the transformer (although leakage inductance can). The primary inductance is not in series with the signal going from primary to secondary.

 

[alec_t]

circuit diagram is attached with this . wish to hear your opinion about it.

Don't use it; it could become 'live' if the input resistor failed. Why not use a transformer, since you may need one anyway to provide power to your circuit?

 

[aswin p ajayan]

Any power transformer you use should have sufficient frequency response to pass all the spikes and fluctuations of interest. The primary inductance of the transformer determines the magnetizing current, it does not affect the frequency response of the transformer (although leakage inductance can). The primary inductance is not in series with the signal going from primary to secondary.

well i think i would use a transformer to step down the voltage and then use that circuit to further step down to 5v and level shift it by 2.5 volts
could you suggest me a low cost rail to rail op-amp

 

[dr pepper]

A standard mains transformer might well pass higher order harmonics, however the core probably wont cope well with lower order harmonics and attenuate them, doesnt mean to say you cant use one however.

 

[ben7]

You don't NEED a transformer to do it. There are such things as opto-isolators, and using more than one MCU.

 

[dr pepper]

On that route you can get linear opto isolators which could measure the voltage indirectly then you could do the job with only one mcu.

 

[NorthGuy]

It all depends. If you fry the optoisolator and surrounding traces, it could be about the same amount of work/money to repair it as if you fried MCU.

I don't think the resistor has a high probability of failing short if you get a resistor of good wattage and high value. I have my power measured that way.

From the safety standpoint, one of the main wires has to be "grounded" (e.g. neutral wire in North America or ground wire in Europe) while the other is "hot". If you connect your MCU ground to the "grounded" wire, it'll be safe. You can make it more safe if you make your PCB in such a way that in the case of failure the connection to the ground stays. This way, even if the resistor fails, the thin PCB traces to the "hot" main will burn down quickly and your device will stay connected to the "grounded" wire - will be safe to touch.

 

[Reloadron]

This link may be of some help.

This is not a solution but should give you a few ideas. Years ago HP (Hewlette Packard) marketed a distortion analyzer known as the HP 332 Distortion Analyzer. It looked at distortion on a fundamental frequency. If we look at what Alec posted you can see the major distortion was harmonic distortion followed by what is likely hum and noise on the fundamental. The unit I linked to uses a modified wien bridge as a tuned rejection filter to reject the fundamental. With the fundamental attenuated (eliminated) all that remains is the harmonics, hum and noise from the fundamental.

Anyway, the circuits and more important the theory of operation may be of some help to you. Keep in mind this is late 1960s technology so doing it today is much easier than 50 years ago.

Ron