I thought AC was meant to be a sine wave

[Mark R]

So I plugged my my sound card into the mains (*)

This is what happened:

I was wondering why my halide lamps were a bit flickerier than I was used to in my old home. Those nice shelves at zero volts obviously do wonders for the arcs.

* via an appropriate transformer and voltage divider

(Unfortunately, my old power analysis program met with an unfortunate accident, in that I accidentally excluded it from my backup schedules, and then my old comp blew up). However, I've never seen a power waveform remotely as bad as this one, ever.

 

[Jeff7]

Eww, nasty. I don't suppose you happen to have an oscilloscope handy for comparison?
I know the AC at work, and also here at home, is a pretty reasonable sine wave, without much noise at all.

 

[edro]

Are you using any type of solid state power converters in your house?
Variable speed drives, lighting dimmers, solid state relays, etc?

That is a pretty typical waveform for these types of devices... but most don't corrupt the main feed... just the output.

 

[PottedMeat]

(Unfortunately, my old power analysis program met with an unfortunate accident, in that I accidentally excluded it from my backup schedules, and then my old comp blew up). However, I've never seen a power waveform remotely as bad as this one, ever.

wow that's fairly noisy

can you dump the raw data into excel and do an FFT?

 

[MovingTarget]

Negative. It is a phase shifted cosine wave.

 

[MovingTarget]

Also, I wonder why you have such a noisy signal. Are you near anything that could cause such signal interference?

 

[bobdole369]

the quick answer is here in the OP:

via an appropriate transformer and voltage divider

Whats your voltage divider look like?

Any transformer is hugely inductive and quite likely to distort the wave a bit unless its power draw is within a reasonable range. (Usually only a big issue when the core is saturated, but at that point its getting ridiculously hot) but undersaturation can do something like that too.

However, I believe that even with a simple transformer/voltage divider you ought to get a fairly good representation of a decent sine wave. I've never seen anything BUT a normal looking sine wave when I stuck a 10x probe in the wall.

Is there by chance a transfer switch somewhere? Or is this the output of a power strip or a UPS? Is there a rusty old fridge from the 60's chugging along somewhere in your house, or a weird inductive stove? How bout ham radio operators or big rig CB'ers nearby?

The period looks to be .02 seconds, so you are on 50 cycle power yes?

 

[PowerEngineer]

Yikes!

One thing I note is that it appears to be 50 Hz, so I gather you must be somewhere other than North America.

I'm having a hard time believing that your AC power is that badly warped. The distortions in the wave are so nearly identical in each cycle that you're either sitting right next to a huge solid-state motor controller (where the on-off switching can definitely screw up the supply side voltage) or there's something (maybe filtering) in your measurement set-up or sound card that is actually responsible.

 

[LordMorpheus]

Those zero-crossing hits look a lot like a PWM driver operating without any kind of dead-time compensation. There is probably a circuit similar to that in one of the devices between your instrument and the power outlet.

 

[SonicIce]

how do u hook that up?

 

[halik]

I am so lost in this thread

<- not EE

 

[Mark R]

I've now recreated my analysis program - it wasn't as difficult to put together as it was the first time.

Pics of instrument:

Analysis pictures
Voltage waveform

Frequency trace
[img=http://pics.bbzzdd.com/users/MarkR/instrument5.png[/img]

Voltage spectrum (log amplitude scale)

Quantitative harmonic distortion analysis (showing the dramatic distortion changes with time)

Voltage change over time (strange spike seems to be some sort of weird inductive artefact from light switches - makes the whole instrument go nuts)

 

[Analog]

Use a nice Analog scope ftw.

 

[SonicIce]

how can i do the audacity thing do i need a bunch of equipment

 

[Leros]

I just hooked up my scope to a 24v AC adapter and I get a nice sine wave from that. I can only deduce that my mains has a nice sine wave as well.

 

[Jaepheth]

how can i do the audacity thing do i need a bunch of equipment

I think you just take a microphone, cut the cable, and then stick the wires you find into a wall socket.

(Not intended to be a factual statement)

 

[Leros]

OP,

Can you post a schematic of your circuit? I'm wondering if your circuit isn't screwing up your readings.

 

[PottedMeat]

I think you just take a microphone, cut the cable, and then stick the wires you find into a wall socket.

(Not intended to be a factual statement)

heh yeah don't do this

 

[PottedMeat]

OP,

Can you post a schematic of your circuit? I'm wondering if your circuit isn't screwing up your readings.

think there's something else between the potted toroid ( least that's that it looks like to me ) and the voltage divider?

 

[Leros]

think there's something else between the potted toroid ( least that's that it looks like to me ) and the voltage divider?

What toroid? All I see are 4 resistors. Is that all the circuit is, a voltage divider?

 

[PottedMeat]

What toroid? All I see are 4 resistors. Is that all the circuit is, a voltage divider?

next one is the voltage div

 

[Leros]

next one is the voltage div

Oh, I complete ignored that photo because I can't really tell what is going on.

 

[Howard]

What toroid? All I see are 4 resistors. Is that all the circuit is, a voltage divider?

encapsulated toroidal transformer (round thingy with wires out of it)

 

[Mark R]

Whats your voltage divider look like?

It's a 100:1 divider. 5 kOhm/50 Ohm.

Any transformer is hugely inductive and quite likely to distort the wave a bit unless its power draw is within a reasonable range. (Usually only a big issue when the core is saturated, but at that point its getting ridiculously hot) but undersaturation can do something like that too.

Transformers do distort a bit. Cheap transformers can be bad, because the cores saturate, and they have high internal resistance. I've tested these before, on my old program, and they would add 3-4% distortion by flattening off the top of the sine waves.

The current toroidal transformer was a find on ebay, which I picked up for basically free. It's a 225 VA 18V transformer with oversized core (it's an 'audio grade' transformer which oversizes the core to keep the magnetic field down, for reasons of noise, but it also means that the transformer has a huge margin for saturation, so is extremely linear).

At my old house, I used the same transformer and same sound card to give virtually perfect sine waves (< 1% distortion at times). See:

However, this is the first time I've put this setup together at my new house.

Is there by chance a transfer switch somewhere? Or is this the output of a power strip or a UPS? Is there a rusty old fridge from the 60's chugging along somewhere in your house, or a weird inductive stove? How bout ham radio operators or big rig CB'ers nearby?

The period looks to be .02 seconds, so you are on 50 cycle power yes?

No idea what is around here. No large factories for sure. Not much in my house, and even turning everything else off doesn't really do anything. Maybe it's the rectifiers for the subway trains, but there are a lot of 9th/15th harmonics which shouldn't occur on 3 phase rectifiers. Perhaps, it's just because it's a very densely populated residential area, and it's just the effect of a huge number of non-PFC PSUs and long supply cables.

It is indeed 50 Hz.

 

[KGB]

One way of testing your test rig:

If you have a UPS unit, charge the batteries... pull the AC cord and hook your test rig up to the output.