Generator harmonics question(s)

[et...@whidbey.com]

Years ago I adjusted a neighbor's generator so that it put out 60
cycles at 120 volts. I used a TEK 465B 'scope to look at the
frequency. I saw the major 60 cycle waveform and then a lot of smaller
waveforms at higher frequencies. I was told that these harmonics were
typical of small generators.
Today I am replacing the voltage regulator on a small generator and
adjusting the governor. I expect I will see the same thing as with my
neighbor's generator.
Looking inside the generator all I see is rotating coils inside a
stationary coil. The rotating coils are energized with DC from the
voltage regulator.
Looking at old power plant pictures and video it looks as if those
giant generators were built the same way. And many of them are still
in use today.
But I know that the sine wave from the wall has been very good for
almost since, if not since, the first generator installation at
Niagara Falls.
So how do they get the good clean AC from these big generators? Or
maybe it's not so clean and is instead filtered. If so how is this
done? And can I do this myself without too much effort? Is the effort
even worth it?
Until the voltage regulator failed on my generator it ran my stuff
at home just fine. The fridge, freezer, lights, phone and router. I
didn't try any computers. So maybe I don't need to even consider
filtering the generator output.
But I do see a lot of generators that advertise "pure sine wave
output" from an inverter that is part of the generator electronics. So
maybe a pure sine wave is important for some electric or electronic
stuff. Clocks for example.
Anyway, thanks for reading. And for any answers too.
Cheers,
Eric

[Foxs Mercantile]

On 11/26/2017 1:20 PM, et...@whidbey.com wrote:
> Anyway, thanks for reading. And for any answers too.
> Cheers,

Shameless stolen from a forum about these things:
To begin with, most people, including many equipment manufacturer's,
many electricians, and just about all the average Joe's in the world,
haven't got a damn clue about power quality. And many who do, have no
idea what the effects on power systems, equipment like generators, UPS,
'sensitive electronics equipment', etc., cause. In fact, if the average
person were to take an oscilloscope and look at what they think is a
'pure sine wave' coming from their source, they would be absolutely
horrified; and probably wouldn't be sure if they've made a mistake in
hooking it up. The terms 'dirty power' and 'clean power', generally mean
absolutely nothing in the power industry, and are used mainly by people
who don't know anything about power quality issues; and snake oil
salesmen trying to sell their overpriced products to these very people.
So, let's get some fact's straight.

First, most electronics equipment that people think of with bogus terms
like 'sensitive electronics equipment' aren't sensitive at all. In fact,
for the most part; this equipment cares the least about the quality of
its power input. Bear in mind, that most of this equipment first
converts its power source to DC, before using any of it. It will be the
equipment itself, including its filters and voltage regulators that
determines the quality of its DC power - not the source. Yes, there
could be some ripple transmitted to the DC power; but that is primarily
a function of how crappy its power supply is. Most of the buzz about the
need to 'clean up your power' actually comes from manufacturer's selling
equipment that supposedly does just that. And more often than not, it's
pure BS.

Computers, PCs, and anything with a switch mode power supply (SMPS),
draws its power in two large bursts of current during each cycle of a
sine wave. It basically doesn't give a crap what the source looks like
as long as there is power to draw. And oppositely, SMPSs do more to
damage the quality of the input power from their source, than anything
else. If you want clean power, don't put SMPSs on circuits with other
equipment. This equipment is designed to function as long as its input
source is within the CBEMA (ITIC) curve. Basically, its designed to
function without rebooting with a very crappy source. This means that
some 90% of equipment people think are 'sensitive electronics', in fact
are not; and also damage the quality of the source horrendously by
loading it up with harmonics.

Now, lets talk about UPS. The majority of UPS - especially ones that
cost under $20,000 - simply present a large non-linear power supply to
their source. Just like the SMPSs, they wreak havoc on the power quality
of their source. Someone mentioned 'double-conversion' UPS. Nice; but
clueless. (Sorry.) The output of a double-conversion UPS is almost
always much, much poorer than the source sine wave entering. They use
PWM or some other method to convert the DC back into AC, resulting in
one of the most choppy 'sine waves' one could ever see. Yeah - they
ensure the voltage doesn't drop when their input does; and they filter
transients. However, that's about all they do. Aside from that, the
power quality of their output is horrible. Again, anything under $20,000
and the output is going to look more like a square wave than anything.
(Add capacitance of the circuit, and its much poorer than that.) The
majority of low cost (under $20k) UPS, chop their output into three
squarish looking chunks each half cycle. You have to pay big bucks to
get a 12 or 18 pulse UPS, with adequate input and output filtering. But
its not a problem, because the SMPS equipment that is connected to UPS
systems 99% of the time, doesn't care one bit.

Now lets talk generators. Even cheap, low quality ones. Mostly the power
quality problems from cheap generators fall into two issues - varying
input voltage and frequency; and harmonics. Not big deal, because most
of the equipment you load onto a generator, doesn't care about any of
this. Certainly, none of the equipment with SMPSs cares at all. Take a
look at their nameplate. Typically 90 to 240V, 50 or 60 Hz. They will
work at a frequency of probably 30 to 100. As long as your generator
doesn't drop below 90V for more than a second, they will operate just
fine as well. SMPS's generating harmonics into the source may actually
cause more issues for your generator (overheating), than the other way
around.

Now for the big surprise. What types of loads have the most concern from
harmonics, voltage and frequency? Motor loads. Harmonics causes the
windings of cheap motors (and generator windings) to overheat. Motors
that deal well with harmonics are made to be run on variable frequency
drives. They cost more money, because they have larger windings, and
often special treatment to their bearings. Obviously, dropping voltages
can cause motors not to start. Minor frequency issues won't upset them;
but should the frequency of your source drift severely, your motors will
have problems. Trying to use a UPS to help this, will only make it worse
- mega harmonics.

The generator itself may actually be the most concerned about issues
being generated by its loads. Loads with leading power factor (such as
can happen with loads with high capacitance, large filters or high
harmonics) can wreak havoc on the generator, cause it not to run, and
actually physically damage it. Harmonics being generated by the loads
can add to the harmonics being produced by the generator and cause
over-heating. The only solution for this is to over-size the generator
(or de-rate the one you have). Leading power factor problems are
generally solved by introducing isolation transformers between the
generator and its loads. Isolation transformers go a long way towards
cleaning up the power that a generator or motor is seeing; but to really
work wonders, it needs to be a three-phase system. (3-phase delta-wye
transformers filter out triplen harmonics.) Electrostatic shields
between the windings will improve the transformer's filtering of transients.

In addition to motors, there certainly are some other loads that don't
like AC sources with rich harmonics. Could be televisions; not really
sure. Probably, after the issues I've discussed; the most concerning
issue to loads, especially electronics, is transients. These are much
more prevalent from your utility source, than your generator. Switching
transients from the utility opening and closing circuits; as well as
large industrial loads being switched in and out, are the primary
culprits. Nearby lighting strikes are also to blame. The best way to
protect your equipment from these utility power quality problems, is via
high quality surge suppression. So, if you want to waste money in an
attempt to 'clean up your power'; go ahead - plunk down as much as you
can afford and buy an expensive UPS. Chances are probably 95% or better,
you've just made your power quality worse. (FYI - UPS are for having
'uninterrupted' power - not quality power.) If you are concerned about
power quality, the first thing to spend your money on is high-quality
surge suppression. That is likely all you need. The only thing more, in
addition to cleaning up transients (an uninterrupted power) that an
expensive double-conversion UPS will give you is a steady output voltage
- at the cost of high harmonic content. Most of your equipment doesn't care.

Seriously, what you are trying to do is akin to replacing the carburetor
on a car that is running fine; because you think it may not be running
fine. Don't waste your money trying to fix something that has no
symptoms and is not broken. First, you'll probably make it worse.
Second, you'll be out of a lot of money, that you could have used
usefully for something else. If you really have money burning a hole in
your pocket; hire someone with the equipment and know-how to perform a
power quality study, and provide a proper interpretation. Then spend
money fixing anything that is truly a problem.

--
Jeff-1.0
wa6fwi
http://www.foxsmercantile.com

[mike]

Some classes of electronic devices now require power factor correction.
Would be interesting to see similar comments on the consequences of that.

[Jon Elson]

et...@whidbey.com wrote:


> So how do they get the good clean AC from these big generators?

Those big power house alternators have more slots in the stator. The
winding pattern is set up such that the middle of the magnetic pole has more
turns, due to ovelapping sets of coils, than the ends of the pole. This
creates the proper wave shape. Smaller alternators can't have as many
slots, and thus separate coils, so the induced waveform has some "jumps" anf
"flat spots" in it.

> Until the voltage regulator failed on my generator it ran my stuff
> at home just fine. The fridge, freezer, lights, phone and router. I
> didn't try any computers. So maybe I don't need to even consider
> filtering the generator output.

Right, the inductance of typical motors smooths out the small imperfections
in the voltage wavem and you pretty much get a sine-wave current.

> But I do see a lot of generators that advertise "pure sine wave
> output" from an inverter that is part of the generator electronics. So
> maybe a pure sine wave is important for some electric or electronic
> stuff. Clocks for example.

Well, some inverter generators produce horrible square waves or stepped
approximations of sine waves, that might cause problems for a variety of
equipment.

Running clocks off the typical generator will be a waste of time, the
frequency control will be awful.

Jon

[et...@whidbey.com]

On Mon, 27 Nov 2017 14:31:49 -0600, Jon Elson <jme...@wustl.edu>
wrote:

Thanks for the reply Jon. Do you think Tesla's generators at Niagara
Falls were made the way you describe? I wouldn't be at all surprised.
Eric

[Jon Elson]

et...@whidbey.com wrote:


> Thanks for the reply Jon. Do you think Tesla's generators at Niagara
> Falls were made the way you describe? I wouldn't be at all surprised.
> Eric

Everything made since about 1890 had some scheme to control the harmonic
problem. The first alternators had "salient poles" on the stator, which
were just blocks of stacked steel laminations with coils wound on them.
But, even with these, it was possible to shape the rotor flux field to
minimize the harmonics. Later, they went to slotted stator lamination
sections that made it possible to have coils that spanned several slots, and
this made harmonics much smaller.

Also, in the early days, they ran the alternators off steam engines, and the
AC had a definite pulsation with each stroke of the piston.

Jon

[John-Del]

On Monday, November 27, 2017 at 6:30:13 PM UTC-5, Jon Elson wrote:

> Also, in the early days, they ran the alternators off steam engines, and the
> AC had a definite pulsation with each stroke of the piston.
>
> Jon

I would have thought they'd be using a nearly infinite mass flywheel to stave off power pulses.

[Tim R]

We had a power plant on site with huge diesel generators, low speed (300 rpm). They did have huge flywheels. I never knew how clean the power was. They were built to power an entire block of buildings, but by this time they were peak shaving only, and have since been scrapped. Very impressive, the floor shook when they ran.

[et...@whidbey.com]

On Mon, 27 Nov 2017 17:32:43 -0600, Jon Elson <jme...@wustl.edu>
wrote:

Greetings Jon,
Do you know how the harmonics were measured? Did they use an
oscillograph? I am amazed by how much good science was done with such
relatively simple instruments. And by how much was discovered and
figured out when many measurements took so long and then doing the
math that also took so long.
Eric

[Jon Elson]

The alternator, itself, was a big flywheel. But, with a several hundred HP
steam engine with just one cylinder per pressure stage, the power pulses
would be pretty strong. They did often put an additional flywheel in the
system, but they were NOT infinite.

Jon

[Jon Elson]

et...@whidbey.com wrote:


> Greetings Jon,
> Do you know how the harmonics were measured? Did they use an
> oscillograph? I am amazed by how much good science was done with such
> relatively simple instruments. And by how much was discovered and
> figured out when many measurements took so long and then doing the
> math that also took so long.

No, and I'm wondering if even oscillographs were available back in the
1890's. But, I can easily imagine some improvised methods, like mounting a
small motor on Prony brake-like structure. The motor's stator is mounted on
bearings, and allowed to swivel around the shaft. An arm rests on a scale
to measure motor torque. You could replace the scale with a spring and
watch the arm wiggle.

Anyway, the guys who were DEEP into the physics of this stuff, like
Steinmetz and Tesla, could analyze the magnetic fields and the shapes of the
machine's poles and figure out what the harmonics would be. So, they didn't
do it ad-hoc, they analyzed what they needed to do to get harmonics
(waveform distortion) to manageable levels, and built the machines that way,
I do have some electrical engineering texts from 1910 or so that descibe how
this was done.

Another thing that was well-understood way back then was power factor. It
was typical to have one real big machine in a plant that had a wound-rotor
synchronous motor driving it. By adjusting the rotor field, you cound turn
it into a source of leading power factor, correcting the power factor of the
whole plant. These were called synchronous condensers, ie. a rotary
replacement for capacitor banks. I actually found a pair of these in a
1950's vintage facory built for the Korean war. They had a pair of HUUUUUGE
Ingersoll-Rand air compressors of 50 - 100 Hp each. They had salient-pole
synchronous motors that were about 8 feet diameter and one foot length. The
piston was a foot in diameter and had a stroke of several feet,
horizontally. On the wall, there were power factor meters that adjusted the
rotor field strength to keep the plant power factor near 1.0

Jon