Show us your mains waveform!

[Simon]

The reason for BLDC motors is speed control. Small induction motors will be less affected by poor supply because the losses are not much. An industrial motor may have issues.

[Simon]

"replacing induction motors with BLDC motors will only make the harmonic problem worse." Why is that? I'd think any sizeable BLDC will have a PFC front-end, so no harmonics.

Because they are not 50Hz sine driven off the monophase mains. There are 3 phase machines and the reason to use tem is speed controllability. My new boiler went from a fixed speed combustion air blower to a speed controllable one. The mains will be rectified and then 3 phases of AC at some non 50Hz frequency will be synthesized at around 10kHz. The saving graze will be that with decent bulk capacitance and with the drive being 3 phase it will likely even out but not be power factor corrected. You need a neet 3 phase AC syncronous motor for PF of 1 or more not one with a bunch of noisy electronics in the middle.

[uer166]

The mains will be rectified and then..

Why would they not be rectified via a PFC front-end that makes the load look exactly (or close to) resistive? These are mandated for power levels above a certain level. A BLDC with a proper controller with PFC will look like a kettle from mains' perspective, and will help the harmonics situation.

[nfmax]

The mains will be rectified and then..

Why would they not be rectified via a PFC front-end that makes the load look exactly (or close to) resistive? These are mandated for power levels above a certain level. A BLDC with a proper controller with PFC will look like a kettle from mains' perspective, and will help the harmonics situation.

Well that's just the point, isn't it? If PFC is mandatory for loads of any significant power, why is the mains waveform clipped as badly as it is?

[uer166]

Well that's just the point, isn't it? If PFC is mandatory for loads of any significant power, why is the mains waveform clipped as badly as it is?

Because of loads that are below that threshold that don't happen have PFC front-ends 

[Simon]

The mains will be rectified and then..

Why would they not be rectified via a PFC front-end that makes the load look exactly (or close to) resistive? These are mandated for power levels above a certain level. A BLDC with a proper controller with PFC will look like a kettle from mains' perspective, and will help the harmonics situation.

Whats the limit, what power are those fans and pumps? - not much. What manufacturer is going to give a shit about PFC if they can legally make their product cheaper?

Same with phone chargers and many other devices. You have to be over 0.5kW or 1kW to be legally obliged to put PFC in. Some (lab) power supplies do it in the several hundred watts range but last time I checked it was 1kW by law.

[uer166]

I believe that limit is 75W, not 500W, and certainly not 1kW. Supposedly according to EN61000-3-2.

[Simon]

Your boiler combustion blower could well be 75W or less. I don't know what counts as PFC and if the higher you go the more stringent it gets but at low powers I believe a choke in series with a rectifier is deemed OK. there was certainly not much in the last boiler fan I opened, they don't even use full 3 phase drive as there were only 4 MOSFETs not 6.

[Gyro]

... My new boiler went from a fixed speed combustion air blower to a speed controllable one....

Arghh, another domestic 'appliance' switched from decades proven shaded pole motor to electrolytic infested monster! 

[Simon]

... My new boiler went from a fixed speed combustion air blower to a speed controllable one....

Arghh, another domestic 'appliance' switched from decades proven shaded pole motor to electrolytic infested monster! 

i think it makes the boiler more efficient or something. The last boiler was 86% efficient, this one is 96%, if that electrolytic infested monster is helping me save over 10% in gas I'll take it.

[coppice]

There are growing demands for PFC in electronic equipment, but hardly anything is said about the severe harmonic distortion in the current waveform that most PFC hardware causes.

This is incorrect.
The purpose of PFC circuits is to substantially reduce the current waveform distortion caused by rectifier circuits. They may not produce a perfectly clean waveform but they do considerably improve the distortion compared to non-PFC rectifiers.

Have you actually looked at some of the weird current waveforms you get with typical PFC designs? They move the current waveform to be centred around the voltage waveform, and a power factor meter shows an excellent power factor. However the waveform is usually a weird shape, rich in harmonics. If you look in the documentation for PFC chip sets you will actually see these weird waveforms.

[engrguy42]

Seems to me this comes down to whether equipment (utility or customer) is operating within spec or not. Not sure what UK utility specs are for power quality (IEC?), but under normal conditions I think it's reasonable to assume all their equipment is operating withing specs. Which I *assume* results in fairly clean voltage waveforms, though I'm not sure to what extent that's true. It seems fairly certain that a multimillion $$ synchronous generator used by the utility is putting out a very clean voltage. OTOH, if you're being supplied by a utility's portable diesel you can fit on a flatbed it might not be so clean, but still I'm sure Caterpillar makes sure they're within spec.

So then it comes down to whether the ugly waveforms are, although kinda ugly, still acceptable or is there something else going on? My tendency is to first assume the customer has some equipment operating that might distort the waveform.

So I think a useful exercise would be for someone to take their waveforms, do some analysis, and compare to their utility's specs for power quality and see if they're within spec. In the US I think most/all states have a Public Utilities Commission (PUC or PSC) that has rules that utilities need to follow under normal conditions for voltage magnitude, power quality, etc., and you can get them on their website.

Of course under abnormal conditions it can be anything.
 

[engrguy42]

Ooooo...apparently there's a function in Matlab to calculate THD 

Maybe just save a CSV from the scope and bring into Matlab to do the magic ?

Wow, that would be a great science project. Gather a bunch of ugly waveforms and calculate THD so we can see what correlates 

"r = thd(x) returns the total harmonic distortion (THD) in dBc of the real-valued sinusoidal signal x. The total harmonic distortion is determined from the fundamental frequency and the first five harmonics using a modified periodogram of the same length as the input signal. The modified periodogram uses a Kaiser window with β = 38"

[Simon]

The reason for the flattening is fairly obvious. Standards are there because if there were none it would be impossible to operate. But we all know that even with standards you don't get the ideal situation. It can be argued that smaller devices cannot have lots of money spent on their making just to correct for PF when their impact is supposed to be low. But you also have to consider that if slapping a CE mark on your product is all that is needed then who knows what Chinese built down to the cent bought off ebay/amazon shite there is out there.

If I were to find a product that failed PF rules I would not even be able to report it in the UK. Trading standards no longer talk to the public due to cut backs, I have to call consumer direct, a charity help line staffed by numpties that don't even know basic consumer law never mind what PFC is.

CE marking is all well and good but it's hard to trace back to who made it once it's out there and burnt your house down.

[engrguy42]

The reason for the flattening is fairly obvious.

And that reason is?

[Simon]

overlay the current draw waveform of any rectifier driven device and consider the change in load impedance over the waveform. Sure decent computers may have PFC but I bet monitors don't, laptops? there are a ton of devices out there sub 75W. They all add up.

[engrguy42]

overlay the current draw waveform of any rectifier driven device and consider the change in load impedance over the waveform. Sure decent computers may have PFC but I bet monitors don't, laptops? there are a ton of devices out there sub 75W. They all add up.

Like I said before, in my case I had 3 desktops and other equipment running, then turned them all off so there was nothing running in my house whatsoever, and there was no difference in waveform.

So it sounds like you're assuming that either the user here or their neighbors have some unnamed equipment running that distorted the waveform into a flat top?

Do we have any proof of that? Like a before & after waveform? 

[coppice]

The reason for the flattening is fairly obvious.

And that reason is?

If you do simple rectification into a large reservoir capacitor the rectifier doesn't turn on until the incoming voltage exceeds the voltage across the capacitor plus the drop across the rectifier. It then turns off as the voltage drops below that threshold. So, the current draw is focussed near the the peaks of the voltage waveform. Ohmic losses then mean the load only drags down that voltage near its peaks. The result is flattened peaks.

[engrguy42]

The reason for the flattening is fairly obvious.

And that reason is?

If you do simple rectification into a large reservoir capacitor the rectifier doesn't turn on until the incoming voltage exceeds the voltage across the capacitor plus the drop across the rectifier. It then turns off as the voltage drops below that threshold. So, the current draw is focussed near the the peaks of the voltage waveform. Ohmic losses then mean the load only drags down that voltage near its peaks. The result is flattened peaks.

I agree, hypothetically there are a bunch of reasons that CAN cause a flattened sine wave. But the question is which of those things are ACTUALLY causing the issue, and is it acceptable or is there something bad happening? The utility could be overvoltaging a transformer causing saturation and flattened peaks, but that's not normal or good.

Personally I think it's more useful to identify exactly what is causing it in a particular situation rather than assumptions and generalizations.

[ocw]

It can be more interesting to compare commercial power to that coming from an emergency power generator.  The attachment shows a single phase of three phase 208 VAC power from both the commercial source, plus the generator.  The other phases were about the same.

[engrguy42]

It can be more interesting to compare commercial power to that coming from an emergency power generator.  The attachment shows a single phase of three phase 208 VAC power from both the commercial source, plus the generator.  The other phases were about the same.

Wow, ocw, that's interesting. By emergency generator do you mean one of those tiny 5kW things, or a big industrial Caterpillar-type thing?

[nfmax]

It can be more interesting to compare commercial power to that coming from an emergency power generator.  The attachment shows a single phase of three phase 208 VAC power from both the commercial source, plus the generator.  The other phases were about the same.

Congratulations on having the cleanest sinewave so far sighted in this thread! OTOH, as far as I can tell, all the other waveforms were taken from domestic supplies, not industrial. Is this significant, I wonder?

[Simon]

overlay the current draw waveform of any rectifier driven device and consider the change in load impedance over the waveform. Sure decent computers may have PFC but I bet monitors don't, laptops? there are a ton of devices out there sub 75W. They all add up.

Like I said before, in my case I had 3 desktops and other equipment running, then turned them all off so there was nothing running in my house whatsoever, and there was no difference in waveform.

So it sounds like you're assuming that either the user here or their neighbors have some unnamed equipment running that distorted the waveform into a flat top?

Do we have any proof of that? Like a before & after waveform? 

If you want proof I suggest you find a job in the right industry, from our armchairs we can only speculate but theories add up. So you turned 3 PC's off and that is going to make a difference to what? the entire grid? Everything between you and the generator will have an impact. What you are seeing is being caused by whatever is going on for hundreds of km around you. Your entire street is just the start of it, it's impedance to that of the local substation. The local substation is already getting a distorted supply due to all the other substations that present the same type of load and so on up the chain.

It's the same as me trying to turn my 4-5kW of solar generation on or off to see if that makes a difference. As I am one of 3 or 4 houses in the entire area with solar panels it will probably not even make a dent.

[Simon]

The reason for the flattening is fairly obvious.

And that reason is?

If you do simple rectification into a large reservoir capacitor the rectifier doesn't turn on until the incoming voltage exceeds the voltage across the capacitor plus the drop across the rectifier. It then turns off as the voltage drops below that threshold. So, the current draw is focussed near the the peaks of the voltage waveform. Ohmic losses then mean the load only drags down that voltage near its peaks. The result is flattened peaks.

I agree, hypothetically there are a bunch of reasons that CAN cause a flattened sine wave. But the question is which of those things are ACTUALLY causing the issue, and is it acceptable or is there something bad happening? The utility could be overvoltaging a transformer causing saturation and flattened peaks, but that's not normal or good.

Personally I think it's more useful to identify exactly what is causing it in a particular situation rather than assumptions and generalizations.

The grid monitors conditions far more closely than you realize. If there was a serious issue they should know about it ahead of time. When was the last time you had an outage for mare than 1 hour? in the 10 years I have lived here we have had a few blips that might but often don't even crash my PC. This is usually a "bird strike" as we have cabling over the local nature reserve. And yes they send someone out to find the body of the fried bird that didn't pass their navigation school.

[ocw]

Wow, ocw, that's interesting. By emergency generator do you mean one of those tiny 5kW things, or a big industrial Caterpillar-type thing?

That was from a 40 kVA three phase generator.  Some UPS's still have problems accepting that generator waveform as "clean power" and will not pass it.  Many home generators have a much worse output--not that far from square wave.  The commercial power comes from a sub-station not much more that an kilometer away which translates to clean power.