How Clean is your Mains?

[FreeThinker]

Just a general query about mains distortion.
Some time ago we had a guy come to the factory where I was working who gave a presentation on mains borne distortion in a 3 phase system causing very high neutral currents to flow.The main cause (acording to him) was the injection of 3^rd and 5^th harmonics from the likes of switch mode power supplies, Inverters , electronic ballasts (in lighting) plasma tv and the like.He droned on for what seemed like hours about rotating vectors and such and the upshot was that our sinusoidal mains was being polluted with harmonic frequencys and was no longer sinusoidal.This was some years ago and I've not given it any thought since I left the meeting room until I read an artical in EPE Oct 2010 p19 entitled forgone conclusion where he brings up the same problem but from a mainly RFI viewpoint.So the question is how bad is the problem?What pecentage of harmonics is in your mains?Are we slowly creeping towards a square wave and if so what can be done?Filtering at source is one option, but who pays?Is this a real problem or another millenium bug?I know that high neutral currents are becoming an issue in some places noteably office blocks with high densitys of computers and fluorescent lighting with electronic control gear, any one any knowledge of this?

[saturation]

As I recall, the field that studies and mitigates problems with the mains, be it for residential or industrial customers, is called power line quality.  Here's a link, I think it discusses in more detail issue you raise, but not how to solve it!  This is more for electrical engineers, and specifically power line specialists, so its a bit out of my league.

http://www.pscpower.com/Whitepapers/Power%20Quality.pdf

[Mechatrommer]

not sure, but will be interesting to do experiment to check our main. any thought? mine is to just connect the dso to main through a transformer and see whats the wave like
i got situation where my friend complaint that when he connected the flourecent bulbs in parallel, he usually got burnt bulb frequently, i suspect is this not clean main supply, but thats just my speculation, not sure whats really happening. as i will be doing this setup in my new home, so this thing quite become my concern. hope to listen from others too.

[FreeThinker]

As I recall, the field that studies and mitigates problems with the mains, be it for residential or industrial customers, is called power line quality.  Here's a link, I think it discusses in more detail issue you raise, but not how to solve it!  This is more for electrical engineers, and specifically power line specialists, so its a bit out of my league.

http://www.pscpower.com/Whitepapers/Power%20Quality.pdf

Bit out of my league too.Wish I had paid a bit more attention all those years ago.The problem as i understand it is that these harmonics are causing ever decreasing efficiencys in the power distribution network.The power companys do not like this and have legal powers (in the uk at least) to fine the polluter, but who is the polluter? the equipment manufacturer or the user? Who is the easiest to catch and fine? Go figure.More to the point as this harmonic distortion increases how will your equipment stand up to being fed a square wave? You could end up paying at both ends.

[FreeThinker]

not sure, but will be interesting to do experiment to check our main. any thought? mine is to just connect the dso to main through a transformer and see whats the wave like
i got situation where my friend complaint that when he connected the flourecent bulbs in parallel, he usually got burnt bulb frequently, i suspect is this not clean main supply, but thats just my speculation, not sure whats really happening. as i will be doing this setup in my new home, so this thing quite become my concern. hope to listen from others too.

My thoughts would be to monitor the mains as you suggest and use the FfT to show the harmonics.Don't have that sort of equipment myself but I think the rigol does.

[FreeThinker]

Just read this from this same site as Saturation linked to
http://www.pscpower.com/Whitepapers/Harmonics%20&%20Harmonic%20Distortion.pdf
These cases only deal with single phase.With three phase distribution the problems are far worse as the loads become unbalanced and hence the large neutral currents.These cause heating and wasted power and can be a great problem.Also 3ph motors connected to a badly distorted system are prone to burn out due to this imbalance.The answer that was preposed at the time was to filter the incoming mains with large chokes to stop the distortion leaving the "polluter" and being retransmitted to the mains network.This did nothing to stop the distortion "consumer" side but helps prevent it being exported to neighbouring consumers.Lots of control manufacturers now include rfi chokes to limit the effects but MOST supplys are extremely distorted and it's getting worse.

[Zero999]

Yes I did quick module about this at college. The problem is odd harmonics are reveres clocked and try to drive AC motors in the opposite direction which effectively increases the loading causing them to draw more current and overheat. Harmonics can also add in the neutral conductors which are generally undersized to account for some load balancing between phases. The solution is to use power factor corrected SMPSes but small PSUs such as those found in wall warts, fluorescent lighting and computers aren't power factor corrected. In some cases, simply going back to old magnetic fluorescent ballasts for the factory lighting can solve the problem but often expensive harmonic filters are the only solution.

[Mechatrommer]

this is come out of 2nd wind transformer connected to our main, 240Vac in, 20Vac out. looks clean to me except there is glitch (as shown), not sure what it is.

[sigxcpu]

a dimmer?

[DJPhil]

this is come out of 2nd wind transformer connected to our main, 240Vac in, 20Vac out. looks clean to me except there is glitch (as shown), not sure what it is.

a dimmer?

I think so, that's exactly what I get on my scope with a variable speed fan on full.

[PetrosA]

It's definitely becoming a problem for electricians and engineers. I'm working on saving up for power quality analyzing equipment, but it's still outrageously expensive (in the thousands of dollars for ones that can do more than one phase/leg at a time).

Even on a balanced three phase load where the neutral load should be zero, it can actually be higher than any of the phases. It's gotten to the point where some electrical engineers are speccing oversized neutrals for buildings where there is a high chance of nonlinear loading since equally sized neutrals will burn up. The problem isn't only limited to the customer's premises, either. The harmonic distortion will travel back through the distribution lines and affect other customers service as well. In extreme cases isolation transformers have to be installed. Ultimately, it's a problem that the whole world will have to deal with and spend billions of dollars in doing so. That's progress for ya

[Mechatrommer]

I think so, that's exactly what I get on my scope with a variable speed fan on full.

is this means a vari fan (with dimmer) can screw up the mains?

[PetrosA]

I think so, that's exactly what I get on my scope with a variable speed fan on full.

is this means a vari fan (with dimmer) can screw up the mains?

To some degree absolutely. Most dimmers and fan controls nowadays are using PWM and variacs instead of rheostats. That means a non linear load

[Mechatrommer]

To some degree absolutely. Most dimmers and fan controls nowadays are using PWM and variacs instead of rheostats. That means a non linear load

and this is propagated outside our home mains? to the electric pole? can we feel the other house's dimmer? sorry noob question. but qurious.

[jahonen]

Here are some measurements of local mains, done with Agilent MSO6034A and Fluke DP120 CATIII differential probe. Time- and frequency domains are in separate pictures to get better views in each domains.

It seems that 5th harmonic is dominant here, but mains is still relatively clean. Not as clean as in PowerWare 9120 online UPS I have, though.

One can reduce the mains pollution by preferring equipment which has power factor corrector, as that will make the input current more sinusoidal, thus creating less distortion. Distortion problem is worse on those countries which use lower mains voltage, since currents are higher.

I think that if the power distribution would be invented today, DC would be much better choice since most equipment runs with DC anyway, and polyphase AC motors could be fed with the VFD. With DC, there would not be distortion problems, and no AC losses on the distribution networks, although that would make voltage transform a bit more difficult, but I think that would still be solvable problem using today's technology.

Regards,
Janne

[FreeThinker]

It's definitely becoming a problem for electricians and engineers. I'm working on saving up for power quality analyzing equipment, but it's still outrageously expensive (in the thousands of dollars for ones that can do more than one phase/leg at a time).

Even on a balanced three phase load where the neutral load should be zero, it can actually be higher than any of the phases. It's gotten to the point where some electrical engineers are speccing oversized neutrals for buildings where there is a high chance of nonlinear loading since equally sized neutrals will burn up. The problem isn't only limited to the customer's premises, either. The harmonic distortion will travel back through the distribution lines and affect other customers service as well. In extreme cases isolation transformers have to be installed. Ultimately, it's a problem that the whole world will have to deal with and spend billions of dollars in doing so. That's progress for ya

When I was A Lad.... the neutral conductor was normally of REDUCED csa to save weight/cost the theory being that it would carry little in the way of current, as you rightly point out it is now the norm to OVER spec the neutral conductor due to burnout.Neutral conductor failure can have dire results like 240v equipment floating to way above this voltage.

[PetrosA]

To some degree absolutely. Most dimmers and fan controls nowadays are using PWM and variacs instead of rheostats. That means a non linear load

and this is propagated outside our home mains? to the electric pole? can we feel the other house's dimmer? sorry noob question. but qurious.

Dimmer switches and fan controls probably aren't the main culprit when issues propagate beyond a single customer. Hundreds or thousands of fluorescent/HID lights, computer power supplies, UPS equipment, variable frequency drives, etc are the things that can create the issue. I suspect that in EU, where compact fluorescent is now mandatory, many apartment buildings will show greater THD than they would have 10 years ago. Most appliances with non-linear components like motors now have PF correcting caps installed to get a PF rating near 1.0, but the electronic stuff is dirty as hell.

[alm]

Most appliances with non-linear components like motors now have PF correcting caps installed to get a PF rating near 1.0, but the electronic stuff is dirty as hell.

I believe anything designed to draw a significant amount of power, like a computer power supply, is required to have active PFC in the EU, so from cos(phi) perspective, they should be very close to 1.0.

[DJPhil]

I was getting ready to stow the fan for the winter anyhow, so I cracked it open to see what was inside. It's a simple circuit wrapped around a T435-400T triac from ST (it's an old one, datasheet's from the new >=600V version). I took a picture of the board and a few of the scope with the fan running at different settings. Sorry for the crappy pics, but still getting used to cameras and linux.

The single dial on the fan moves from off to max, and then down to low speed. To get the traces on the scope I just held the probe near the dial with my finger on the tip of the probe, so a lot of the nasty noise is due to the fact that this measurement was made with me as an antenna. It's crappy, but it gets the idea across. You can see the glitch at max grow to garble up 90 degrees of phase on each side of the waveform. A good measurement wouldn't show quite as much mess, but it's still pretty nasty and the sharp transit time begets harmonic pollution.

and this is propagated outside our home mains? to the electric pole? can we feel the other house's dimmer? sorry noob question. but qurious.

Others have answered well on this, but I thought I might add that the effect of a little fan like this is visible in a very local sense (i.e. the same power strip). I'd always figured that the harmonics from relatively small household dimmers and the like were largely radiated or filtered by line inductance/capacitance. As others have described it's a much bigger issue with a factory full of motors.

Hope that's useful to someone.

[FreeThinker]

As I understand it PF is due to inductive loads causeing the current and voltage to get out of step by 90⁰.This causes the true power to be greater than the apparent power by a certain factor(power factor) ie at 240v 10 amps the power should be 2400 watts however a watt meter would show less than this say 2200w if the the load is inductive (or capacitive) this equates to a pf of approx 0.91.This is NOT a power loss only a reduction in the efficiency of the supply.If the generator was a 1 meggerwatt at 250v then it can supply 4000 amps.As your power factor goes down the power companys profits go down because they charge you for the apparent power used (kw) and not the reactive power (Kvar) so 40 customers could get 100 amps each at unity pf but at 0.91 pf would only pay for 100 * 0.91 * 250 watts = 22750w and not the 25000w it's v*a would imply.As stated previously this is not a power loss as such just an inefficiency due to the current and voltage not being in step but is a reduction in capacity of the network overall.However with harmonic distortion there are much bigger issues at stake which have nothing to do with PF.

[FreeThinker]

Yes I did quick module about this at college. The problem is odd harmonics are reveres clocked and try to drive AC motors in the opposite direction which effectively increases the loading causing them to draw more current and overheat. Harmonics can also add in the neutral conductors which are generally undersized to account for some load balancing between phases. The solution is to use power factor corrected SMPSes but small PSUs such as those found in wall warts, fluorescent lighting and computers aren't power factor corrected. In some cases, simply going back to old magnetic fluorescent ballasts for the factory lighting can solve the problem but often expensive harmonic filters are the only solution.

Yep I remember now why the presentation was held in the first place, we had lost a lot of 3 ph motors in a short period of time due to burn out.Someone had heard about this company that sold these filters which might cure the problem.As you seid they are very expensive and needless to say the cost of rewinding the motors was a lot less than the cure! But this was 10 15 years ago the problem is much worse now.

[Zero999]

I was getting ready to stow the fan for the winter anyhow, so I cracked it open to see what was inside. It's a simple circuit wrapped around a T435-400T triac from ST (it's an old one, datasheet's from the new >=600V version). I took a picture of the board and a few of the scope with the fan running at different settings. Sorry for the crappy pics, but still getting used to cameras and linux.

It looks like a standard phase control lamp dimmer to me.

The schematic will look similar to this. The extra capacitors and the inductor are for RF suppression and there'll be a snubber network to protect the TRIAC from the back EMF generated by the motor.

http://www.wiringdiagrams21.com/search/12+volt+lamp+dimmer

However with harmonic distortion there are much bigger issues at stake which have nothing to do with PF.

That's not true, a non-linear load which draws a non-sinusoidal waveform and causes harmonic distortion will have a power factor of much less than 1. What you said about inductance, reactance and power factor is completely true though.

Yep I remember now why the presentation was held in the first place, we had lost a lot of 3 ph motors in a short period of time due to burn out.Someone had heard about this company that sold these filters which might cure the problem.As you seid they are very expensive and needless to say the cost of rewinding the motors was a lot less than the cure! But this was 10 15 years ago the problem is much worse now.

But surely the cost of rewinding motors adds up so you'd save money in the long term. The question is how long the filter will take to pay for itself and whether it's worth taking a lone to buy one if one doesn't have the capital.

[jahonen]

However with harmonic distortion there are much bigger issues at stake which have nothing to do with PF.

That's not true, a non-linear load which draws a non-sinusoidal waveform and causes harmonic distortion will have a power factor of much less than 1. What you said about inductance, reactance and power factor is completely true though.

This can be understood via the fact that distorted current has significant harmonic frequencies, and only same frequency components in voltage and current will yield to net power transfer, i.e. P=?û*sin(n*x)*î*sin(m*x) = 0 if n ? m (n and m are integers).

More general definition of the power factor is real power divided by apparent power (PF=P/S), the "phase shift" aka cos ?-version applies only with linear loads, i.e. only those with resistors, capacitors and inductors.

Generally, linear reactive power is not a great problem, since no distortion is produced in the mains. Those loads with just a bridge rectifier and filter capacitor will create distortion (even with the transformer, so your audio amplifier will distort its own power). Active PFC can be used to make use the whole sinusoidal cycle, and thus lowering the current peaks and distortion.

Regards,
Janne

[FreeThinker]

However with harmonic distortion there are much bigger issues at stake which have nothing to do with PF.

That's not true, a non-linear load which draws a non-sinusoidal waveform and causes harmonic distortion will have a power factor of much less than 1. What you said about inductance, reactance and power factor is completely true though.

I Think you have misunderstood my comment, I didn't mean to imply that non linear loads do not effect PF far from it but the main consideration with harmonic distortion is the non sinusoidal nature of the mains and the effects it can have on your equipment eg the braking effect on 3 phase motors causing extra load on the motor and even premature failure.And yes it would of made sense to fit the filters but how many accountants do you know who can quantify such items on a balance sheet?BOTTOM LINE RULES common sense is not in their vocabulary.

[Zero999]

This is a bit off topic but I've noticed something about the dimmer switch: it uses a potentiometer with a plastic shaft. If you build a mains powered dimmer switch, never use a potentiometer with a metal shaft, unless you mount it to an earthed metal panel, the shaft should always be plastic.