diyAudio (
-   Power Supplies (
-   -   Distortion in ac mains supply ? (

percy 11th March 2012 05:47 AM

Distortion in ac mains supply ?
1 Attachment(s)
This the spectrum I get for my household 120V 60hz AC line. I measured this by just dropping the voltage down to line level (~2V) using a simple resistor voltage divider and recording it with my soundcard at 96Khz/24bit.
Pretty strong and extended odd order harmonics and a fair amount of even order as well.
Is this normal ? Is this what typical household ac looks like ? Frankly I wasn't expecting this, I was pretty taken aback but has anyone measured the spectrum on their ac line ? I am very curious how does it look for others.

lanchile 11th March 2012 05:53 AM

1 Attachment(s)
yeap, most the time you will get a lot of RF/EMI! I mesured mine and I could hear radio station and noise. after using an AC filter I got nothing!!! I measured mine with my toy.

zigzagflux 11th March 2012 05:58 AM

Very typical. IEEE Red Book recommends 5% THD or less at the service entrance, which you meet. Likely your utility will consider this acceptable. Harmonic spectrum is normal given today's electronic loads.

percy 11th March 2012 06:04 PM

an ac filter might help suppress >100Khz I guess ? but what about the lower order harmonics ? Which brings up the question - Do those lower order harmonics really matter though ?

simon7000 11th March 2012 06:15 PM

No, it is quite a bit noisier than what I would expect. Try running it through an audio transformer for isolation. I think you are seeing RF past the Nyquist limit fold back down.

The first 5 or 6 spikes look about right, it is the 1 K and sidebands that don't look right.

zigzagflux 11th March 2012 06:18 PM

Correct, an AC filter is designed to block/suppress RF frequencies, but will do nothing to the harmonics between 60 and 20k. As far as whether or not those harmonics matter depends on the application.

When I AC heat DHT's, I prefer to run my heater source through a ferroresonant transformer, which produces harmonics only out to about the 11th. Higher order trash is eliminated. So you get slightly cleaner and regulated power. But for converting AC to DC for amplifier power supplies, this source at your house panel is just fine. Having a cleaner sine wave won't really improve the situation. In fact, it is the use of rectifying power supplies in your home and on the grid that causes the voltage to look like this in the first place.

Any attempt to clean up the source harmonics will normally result in higher source impedance as seen by your amps, which is usually not the goal.

percy 11th March 2012 06:55 PM

simon, actually those are indeed 60hz harmonics. What you see 1Khz is actually the 17th harmonic (1020hz). I will try to just attach the wav file so that anyone can look at it directly.

if this doesn't matter when converting ac to dc, then why does the higher frequency interference (emi/rfi) matter and is usually desired to be filtered out ?

AndrewT 11th March 2012 07:41 PM

Curl suggests wiring a 10uF film cap across the Live & Neutral poles.

This sinks the higher harmonics more than the lower harmonics.

I note that the 3rd, 5th and 9th are each ~1% of the 60Hz fundamental.

I suspect the total distortion must be around that 5% upper limit.

zigzagflux 11th March 2012 08:16 PM

Soundcard indicated 4.5%

zigzagflux 11th March 2012 08:31 PM


Originally Posted by percy (
if this doesn't matter when converting ac to dc, then why does the higher frequency interference (emi/rfi) matter and is usually desired to be filtered out ?

Not sure why we are discussing emi/rfi at all in this specific case; you have not shown RF to be an issue on your supply. All you have is a typical distribution of harmonics due to rectifying loads.

These harmonics can be easily filtered out by a power supply filter (series inductors and parallel capacitors). So they are not detrimental to the performance of a power supply. In fact, a square wave AC input rectifies very nicely, with low peak currents (improved crest factor).

While the inductance equation XL=2pifL implies RF will be greatly attenuated at the series choke, the reverse is true. RF can fly right through an inductor's stray winding capacitance. Similarly, RF will not see a parallel capacitor as a short circuit (1/(2pifC)), rather it will see the lead inductance as significant. So RF is not attenuated by a power supply filter like the lower order 60Hz harmonics are. Once you have RF inside your power supply, tubes or transistors are happy to amplify or oscillate in response to their presence. Bad news.

So maybe a question for you to keep the thread on course- why are you concerned about the typical FFT spectrum of your AC line?

All times are GMT. The time now is 02:04 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2