Generating 3 phase power with a small engine

[Iron Beaver]

Hey everyone,

I am contemplating building myself a small 3 phase generator, something big enough to run a 5 hp 3 phase motor. I have a nearly new Kohler CH620 engine that I would use as a power source. Using the rule of thumb that says 2 engine HP per generated KVA , I could run a 9 KVA generator head. With a PF of .8, that gives me a 7.2 KW generator.


But, I'm concerned about the quality of the generated power. Seems like with a V-twin engine spinning the generator, the phases might be pretty unbalanced?

 

[930dreamer]

Will the generator run cheaper than electricity powering the idler motor? You can balance each leg if needed. I use a 10 hp idler for mine.

 

[ez-duzit]

"Seems like with a V-twin engine spinning the generator, the phases might be pretty unbalanced?"

I don't see how the number of cylinders affects it.

There needn't be an idler in this system.

 

[Iron Beaver]

This is contemplated for use in areas with no electrical hookup, so until I figure out how to plug the idler motor into a rock or tree I'll need an engine to power it

My concern is not so much unbalance between the legs of the 3P power. More that a V-Twin engine has timeswhen there is no power applied to the crankshaft, which means that to keep turning the generator, kinetic energy will be spent from the flywheel, causing it to slow down. Then on a power stroke, the flywheel will speed up again.

According to my quick calculation, this should happen about 3600 times per minute. Not noticable for normal applications, but enough that I'm worried about the frequency of generated power oscillating between <60 and >60 Hz within one cycle

 

[Iron Beaver]

This seems like it could then exceed the THD limits for sensitive equipment

 

[like2wheel]

Seems like the mass of the rotating flywheel/rotating assembly would take care of that.

 

[ez-duzit]

The flywheel effect of the generator should keep the power even.

 

[east_tn_emc]

It might be easier to get a MILSURP generator for a few thousand dollars.

My MEP003 can make single phase or 3-phase power.

 

[American Locomotive]

THD isn't great with small generators, but it's fine enough for just about every purpose. I doubt you'd have an issue with anything.

 

[like2wheel]

This is contemplated for use in areas with no electrical hookup, so until I figure out how to plug the idler motor into a rock or tree I'll need an engine to power it

I don't see what the problem is

 

[Iron Beaver]

THD isn't great with small generators, but it's fine enough for just about every purpose. I doubt you'd have an issue with anything.

...Except my mag drill which has a large warning label "Not to be used with power from an engine drive welder", VFDs, and computer equipment.

I may be willing to make that trade off and power the sensitive stuff with a pure sine wave inverter; I just want to be sure I'm on the right track to expect relatively high THD from such a setup due to the unevenness of a 4-stroke V twin engine

 

[Iron Beaver]

I don't see what the problem is

I don't have those kinds of rocks around here...

 

[American Locomotive]

...Except my mag drill which has a large warning label "Not to be used with power from an engine drive welder", VFDs, and computer equipment.

There's a reason why it says engine drive welder.

The #1 reason is that historically, engine-drive welders had 115v DC outlets that ran from power coming off the generator's exciter. That typically works fine with simple brush-motor tools, but not so much with tools that actually have other things going on - like mag drills or modern grinders with electronics. So it's easier to say "Don't run it off a welder, period" rather than than have some bozo try and figure out if his welder's output is DC or AC.

Secondly, welding is inherently a very electrically noisy process. The arc basically creates a giant extremely powerful spark-gap transmitter, emitting very intense broadband noise. Some of that noise is going to end up on the rest of the welder's electrical system, which could cause issues.

Thirdly, if striking the arc causes a momentary voltage drop until the regulating circuitry or engine governor kicks up, it could cause the mag drill to detach from whatever it's attached to.

As far as computer equipment and VFDs go: Most modern consumer electronics do not care at all what the incoming power looks like, neither do VFDs. They all rectify the incoming power into pure DC, filter it with giant capacitors, and then chop it back up into high-frequency AC.

Really the only things that care about crappy generator power are cheap appliances with electronic controls, because they typically have a garbage old-school linear power supply that just passes all of the AC mains noise right into the output, or use a simple capacitive dropper/resistor circuit that does the same.

 

[Iron Beaver]

I am aware of broadband noise from the welding arc, but didn't know about the other points. Thank you @American Locomotive. Also thanks for clarifying about modern consumer electronics power supplies.


There is also the question of whether 3phase power is worth it. I would say yes but my wallet might have an issue with generator heads that start at $1200

 

[American Locomotive]

It's only worth it if you have a need to run 3-phase equipment, otherwise I do not see the point.

 

[Iron Beaver]

Well, my local scrap yard has an almost unlimited supply of 3-phase electric motors in all sizes. Much cheaper than single phase, especially at scrap prices. Plus most machinery I buy comes with a 3-phase motor and it would be nice not to repower everything. I might be moving off grid soon, otherwise I would buy an RPC in a heartbeat and be done with it.

 

[Lassen Forge]

Our home genset has a V-twin power plant, and we've never had an issue with the frequency being off or the sides being "unbalanced"... The power source has nothing do do with it as long as it's regulated and stable... it's the generator head that generates your power and as such, the phases....

You're talking apples and tuna fish.

 

[Iron Beaver]

Yes, but if the generator head has input from an engine with cyclically varying RPM, the frequency and voltage will vary cyclically as well.

Any 4-cycle engine with <4 cylinders has cyclically varying speed. Whether it's enough to matter is quite another question

 

[ez-duzit]

...Any 4-cycle engine with <4 cylinders has cyclically varying speed...

No, that is not typical.

 

[nadogail]

I see no problem with the use of a two cylinder engine to spin a three phase alternator; assuming the engine has enough power to do the job.

 

[Iron Beaver]

No, that is not typical.

Sorry to disagree, but it is definitely typical.

Consider a 4-cycle, 2-cylinder engine: each cylinder fires once every 720 degrees of crankshaft rotation and provides torque over a little less than 180 degrees of rotation. Therefore there are at least 180 contiguous degrees of rotation with no power input. If this engine is hooked to a load (even just internal friction counts), it is putting out power during this time.

This power comes from the flywheel. But by conservation of energy, the flywheel is losing kinetic energy in the process, and therefore speed. Thus the engine slows down until the next power stroke of a piston.

For the case at hand (a 4-cycle, 2-cylinder engine driving a generator at 3600RPM) this could happen over as little as 1/120th of a second. If the flywheel has a sufficient moment of inertia, this effect could also be small. But it will exist, and cause distortion of the waveform. An electric motor driven by this current will spin in the same fashion, as if an unbalanced load were bolted onto it's shaft. The THD of the waveform will be non-zero.

My question was if this effect would be large enough to matter in practice. The consensus seems to be no.

 

[American Locomotive]

Any engine will have varying crankshaft speeds between firing cycles. Even a 4 cylinder engine still has "dead spots" with no power being applied to the crankshaft. You really don't get near seemless power until you get into the 12+ cylinder range.

The phases will be balanced just fine, but you will have some measurable amount of harmonic distortion.

 

[ez-duzit]

Sorry to disagree, but it is definitely typical.

Consider a 4-cycle, 2-cylinder engine: each cylinder fires once every 720 degrees of crankshaft rotation and provides torque over a little less than 180 degrees of rotation. Therefore there are at least 180 contiguous degrees of rotation with no power input. If this engine is hooked to a load (even just internal friction counts), it is putting out power during this time.

This power comes from the flywheel. But by conservation of energy, the flywheel is losing kinetic energy in the process, and therefore speed. Thus the engine slows down until the next power stroke of a piston.

For the case at hand (a 4-cycle, 2-cylinder engine driving a generator at 3600RPM) this could happen over as little as 1/120th of a second. If the flywheel has a sufficient moment of inertia, this effect could also be small. But it will exist, and cause distortion of the waveform. An electric motor driven by this current will spin in the same fashion, as if an unbalanced load were bolted onto it's shaft. The THD of the waveform will be non-zero.

My question was if this effect would be large enough to matter in practice. The consensus seems to be no.

Oh come on!

Always wondered why my Harley kept slowing down and speeding up.

 

[Iron Beaver]

Any engine will have varying crankshaft speeds between firing cycles. Even a 4 cylinder engine still has "dead spots" with no power being applied to the crankshaft. You really don't get near seemless power until you get into the 12+ cylinder range.

The phases will be balanced just fine, but you will have some measurable amount of harmonic distortion.

True, but the problem gets worse as the number of cylinders decreases. I should not have used "unbalanced" to describe the phenomenon under discussion, as that almost always applies to voltages/currents being equal between the phases. That won't be a problem. Not sure about HD yet, but it sounds like it may be fine. Maybe.

 

[Bert_]

The number of cylinders isn't something to worry about. The change in speed over a rotation is too small to matter...

 

[wyliesdiesels]

sounds like you need a 4cyl engine

 

[u3b3rg33k]

worrying about this is stupid. if 3 phase power is more even than single phase, won't the "problem" be "worse" with your single cylinder single phase HF predator generator?

 

[walta]

Why would your output be unbalanced? When I think about unbalanced 3 phase to me that means having a voltage difference compared to the other phases if you have a quality head that should not be a problem. This is a common problem for rotary phase converters when creating the missing phases.

In my mind your big hurtle will be getting your motor to spin at the exact speed necessary to generate 60 cycle per second power and not slow when the generators load increases or over rev when the load decrees.

It seems unlikely you will happen to build the perfect governor for the used motor you happen to have and connect to whatever head that is available to you at an affordable price. Seems to me engineers spend entire careers in labs hoping to get this right.

If you have just one motor to power and have a single phase power on site it seems like a VFD would be the best solution and cost less than a generator head to buy and much less to operated over time.

Walta

 

[Walkers]

I have two generators. One is a 20 horse Onan, the other is my 6-1 Lister slow speed diesel. I notice nothing but smooth power out of my 20 horse. My lister with an 1800 rpm head has what is referred to in the circle as the ‘Lister Flicker’, since my lister is running at 600 rpm and is a 4 cycle diesel. I think you are way overthinking your issues with the unbalanced motor.

 

[u3b3rg33k]

Why would your output be unbalanced? When I think about unbalanced 3 phase to me that means having a voltage difference compared to the other phases if you have a quality head that should not be a problem. This is a common problem for rotary phase converters when creating the missing phases.

In my mind your big hurtle will be getting your motor to spin at the exact speed necessary to generate 60 cycle per second power and not slow when the generators load increases or over rev when the load decrees.

It seems unlikely you will happen to build the perfect governor for the used motor you happen to have and connect to whatever head that is available to you at an affordable price. Seems to me engineers spend entire careers in labs hoping to get this right.

If you have just one motor to power and have a single phase power on site it seems like a VFD would be the best solution and cost less than a generator head to buy and much less to operated over time.

Walta

don't forget to turn on slip compensation and sensorless vector mode on the VFD - get a more constant RPM output that way than on utility power - no slip compensation as the load increases on utility power.

 

[TractorJeff]

If you are running the "stock" governor on that engine, your power distortion will be from the "lag/response time" of the governor. The flywheel of the engine will be "enlarged" when you add the rotor into the equation thus changing the dynamics of your kinetic energy. THD from this setup probably would not be an issue with most of the 3 phase equipment you want to attach to it. If you are building an offsite Server room then I would say you are going to need to run a power analysis test. If you have the resources available, I so go ahead and build it!

 

[Iron Beaver]

Well, I might be running a metal lathe on it. Everyone and their brother says to use 3 phase for a metal lathe motor because 3 phase motors run more smoothly. But I do think this one might not run smoothly if it is electrically coupled to a 2-cylinder engine.

I am aware that the actual solution to that issue is a VFD. But theoretically, I think the above still holds true. I'll probably build this setup anyway because 3 phase power is cool

 

[u3b3rg33k]

Well, I might be running a metal lathe on it. Everyone and their brother says to use 3 phase for a metal lathe motor because 3 phase motors run more smoothly. But I do think this one might not run smoothly if it is electrically coupled to a 2-cylinder engine.

I am aware that the actual solution to that issue is a VFD. But theoretically, I think the above still holds true. I'll probably build this setup anyway because 3 phase power is cool

stop caring about how many cylinders the engine has. it doesn't matter.

3 phase motors can be easily controlled in multiple ways, single phase cannot. all are reversible (swap 2 leads via a drum switch), many can be Y-∆ started, all can technically be run on a VFD/at different Hz for different speeds, some can be externally wired for a slower/faster speed.

 

[txvwnut]

You won’t see an imbalance with the engine at governed speed which will be enough that any “power loss” won’t be noticed. With engine driven generators you set the engine speed by the frequency output of the generator and the governor will/should maintain that setting. If you are planning on running the engine at a low enough rpm to encounter power loss you won’t be able to create any useable power out of the generator.

 

[nadogail]

The engine flywheel and the rotating components of your alternator should, for all practical purposes "smooth out" your power.

 

[u3b3rg33k]

ok gUiSe, this is what the output of an ANCIENT single phase generator looks like with a 3kW load on it:

sadly I don't have a picture, but I have a 45HP 4cyl tractor that can't run the head at 60Hz without flickering, because the governor is ****.

 

[BoostedOne]

Why not just go the VFD route? Im running a 3hp motor on a $185 Teco VFD and its absolutely fantastic.