Idea for a simple high-current brake circuit

[Flux]

Reply to Beaufort ( I don't know how to reply to specific comments)

I spent so long replying that the thing timed out and I lost it and i don't propose to go through it all again.

Briefly for properly designed motors on pure sine supplies ( grid)  star and delta are interchangeable and performance is near enough identical.

With no load or a resistive load on nl alternator there will be no circulating current in delta if the phase voltage contains no harmonics. If it does then the odd harmonics will circulate. The loss caused will not be large but for a machine at cut in then it is a loss you can avoid and you can gain this loss into your battery.

If you are using changeover schemes then the small loss in delta is no big deal when there is significant power in the wind. If you want to optimise the winding for perfect phase waveform then the gain is probably not worthwhile unless you go to full mppt capability ..

In general I tend to agree with Dan, furling works if you get it right, these clever braking control schemes are difficult and unpredictable, I would spend the effort in making the furling work properly. Machines never burn out if you work within their limits. If you want more power in high winds then you can do more with load matching than with all the alternator messing about.

Flux

[ChrisOlson]

With no load or a resistive load on nl alternator there will be no circulating current in delta if the phase voltage contains no harmonics

I assume by harmonics you mean something out of "tune" in the generator - coils not spaced properly, sine waves not exactly 120 degrees, or something similar.  From experimenting with it, it appears to me that the most important thing is phase timing and not having a weird sine wave (like semi-square) for best performance in delta.

But even so, on one older stator that I tested that I know was not all that perfect, I connected 1E to 2S, 2E to 3S and 3E to 1S and tested all three connecting wires with my milliamp meter to see if there was current circulating just giving the generator a good spin by hand (maybe 150 rpm).  That meter is pretty sensitive - it will pick up the "leaks" out of an extension cord laying on a wet cement floor (the meter is design to test for transient currents in dairy barns where people have jumpy cattle because of very tiny currents flowing in water lines, stanchions, etc) - and it didn't even make the zeros flicker on it.

If somebody has ever tried to measure it like I have, and actually gotten a reading, I'd like to see the specs on that generator to get an idea of what caused it.
--
Chris

[Flux]

I don't want to get too involved in this but basically a single turn coil optimally placed will have a virtually rectangular voltage waveform,  the rectangle consists of a fundamental and many odd harmonics. If you mess with the size of the turn in relation to the magnet size you can change the number and phase of the harmonics but you have to do something very special to produce a sine wave.

As you depart from a single turn and distribute the winding about a coil with finite leg width you have different turns having different harmonic components. With the right choice you can balance the in phase and out of phase harmonics to cancel and you end up with a clean sine wave on the phase voltage.

Most axial machines tend to have a phase voltage with a significant in phase harmonic component and the voltage is somewhere between a sine and triangle. When connected star the harmonics cancel in the line voltage and the phase voltage is a clean sine.

If you delta connect a winding with less than a perfect sine wave the harmonics will circulate within the delta, this is a small voltage at best and the current will be small even when circulating within the low resistance of the coils, if you introduce milliammeters in the circuit the resistance of the meter will virtually block any harmonics and you will measure virtually nothing. You would need a transformer type current probe to measure the circulating current so that the winding resistance is not increased.

As well as harmonics there is also an issue of unbalanced phase voltages and it is here that practical imperfections in the windings, spacing, magnet strength etc come into play and also affect delta. Again the IRP method is immune from this effect and has an advantage over delta but unless construction is bad you are not looking at a great effect, just enough to feel the drag when you close the delta winding.

In a 3 phase system the instantaneous sum of the voltages is zero, if the volts are not balanced this can't be true and if the circuit is closed something must circulate somewhere.

Actually arriving at a coil to magnet relationship that produces a pure sine phase voltage is probably luck but unless you do something very bad the things are not often far enough from sine waves to have a serious effect even in delta but if you want delta at cut in it would be worth checking the waveform of your test coil on a scope. Balancing the actual volts is mainly down to careful mechanical construction of the magnet rotors and coils. The issues are very different with slotted iron cores and i don't intend to cover that here.

Flux

[ghurd]

Well, the thing is, what circulating currents in delta?  I asked the fellow (who's been in the business for over 40 years) at the electric motor/generator shop where I get my magnet wire about these circulating currents in a delta generator.  He looked at me over the top of his bifocals and said, "What?  Circulating what?"

Then he said, and these are pretty much his exact words, "Some dimwit looked at a picture of the classic delta connection and decided that since it's all connected in a circle there must be circulating currents in it."

Then he pulled out a piece of paper and scribbled down the KVL calcs as to why there's not. 

I understand what he is saying.
However.
KVL requirements are met no matter what.

Hang a 1K resistor across each phase and KVL is met.  That does not mean there is not current going backwards through the resistors.

Make 2 phases with 1,000 turns and the 3rd with 1 turn, and KVL is met.

Some of that would be pretty obvious with 2 phases with 1,000 turns and the 3rd with 1 turn, then calculate what is doing what at cut in RPM.
And KVL will still be perfectly satisfied.

Pushing amps into a circuit with voltage is different than getting them out at a fixed battery voltage.

If someone was dumb enough to connect a 48V battery bank in parallel with a 24V battery bank in parallel with a 12V battery bank, KVL is still met.
(that's sort of the disclaimer to keep people from doing something silly)
G-

[Ungrounded Lightning Rod]

With no load or a resistive load on nl alternator there will be no circulating current in delta if the phase voltage contains no harmonics

I assume by harmonics you mean something out of "tune" in the generator - coils not spaced properly, sine waves not exactly 120 degrees, or something similar.  From experimenting with it, it appears to me that the most important thing is phase timing and not having a weird sine wave (like semi-square) for best performance in delta.

Harmonics means "the waveshape is not a sine function".  Periodic functions of other shapes than sine can be analyzed (by "fourrier analysis") as a weighted sum of ("component") sine and cosine waves at the frequency of the waveform's period and at integer multiples of it.  The sine-wave component at the basic frequency is the "fundamental" (or "first harmonic") and the sine-wave components at N times the basic frequency are the "Nth harmonic".

Linear circuits (i.e. "the response to the sum of stimuli equals the sum of the responses to the individual stimuli"), when stimulated by periodic functions, can be conveniently analyzed by examining the responses to the individual sine-wave components of the periodic function stimulus.  That's why this technique is handy.  And until you get to the diodes and battery a charging system, a coreless generator is linear (and a genny with a core that is not in saturation comes very close).

= = = =

Symmetrical rectangular and trapezoidal waveforms are very good for battery charging and are what you get when you build an axial flux motor with rectangular or pie-segment magnets and coils.  (You also get them with radial flux machines if you wind each coil in a single slot or with equal numbers of turns in several adjacent slots:  Coils in commercial AC machines are normally wound with carefully-chosen varying numbers of turns in several adjacent slots to end up with something very close to a sine wave.)  But rectangle and trapezoid waves are sums of a fundamental with lots of odd harmonics of nontrivial amplitude.

As I recall it, the third harmonic and every multiple of it (6th, 9th, 12th) ends up with all the coils in-phase, leading to circulating currents limited only by the resistance of the coils.  So with these waveforms the 3rd and 9th harmonics are strong, shorted, and produce nontrivial circulating currents.

Now in a commercial motor winding, with the fancy turn counts that approximate a sine wave and, more importantly, minimize the problematic harmonics, the circulating currents are small.  But that's most of of why they did that funny winding in the first place.

[jlt]

I don't  about all the fancy stuff that been  presented here. but on my machine when i had it wired up delta there was a slight drag on the machine. In irp it went completely away . so that's the way mine will be . the coils may not be exact.but were as close as i could make them. Air core machines are great for start up in low winds. so why add drag if you don't need to.

[SparWeb]

I can concur with JLT, in Delta, the open-circuit torque to turn my motor conversions have always been higher in Delta.  I really doubt you could blame the coil spacing when the stator teeth are made so exactingly.  My magnets aren't perfectly lined up, but we're talking about a change in free-running torque when the wires are hooked up a certian way.  Since you're making Axials I doubt the "drag" is a bother no matter how it's hooked up because it's very low to begin with.

I had to look this up to refresh my memory about what harmonics are.  I just re-read this:  http://www.allaboutcircuits.com/vol_2/chpt_10/7.html  which brings it all back now.  Harmonics pervade these generators because they aren't allowed to have a clean sine wave by the rectifiers.  Only if the load was perfectly resistive would that be true, but batteries and rectifiers work differently.

The example below illustrates the idea - althought they are considering different conditions:



It's showing you clearly these extra currents going 'round and 'round...

In a battery-charging 3-phase rectified genny, it's even worse, because the waveform is not even symmetrical.  It's not just the odd-numbered harmonics that can go around, but some of the even-numbered ones, too.  (well that may be a small factor given the small distortion).

http://www.allaboutcircuits.com/vol_2/chpt_10/7.html


So - why did the guy give you that look?  To him, maybe a bit extra current just heats the motor a bit more (so what if it "runs hot"?  Depends on the person and their attitude I suppose.)

[Flux]

With rotating machines, it's not really possible to have even harmonics the waveform has to be symmetrical but otherwise I agree.

The odd harmonics will always circulate in a delta winding. They can't circulate in star and they can't circulate in IRP below cut in. When the rectifier conducts and murders the waveform there are loads of odd harmonics and they will cause increased loss but they are only an issue during start up and in the very few watts available near cut in.

Machines driving rectifiers into batteries have a lower inherent efficiency but the limits are higher than our common loading method lets us get to so it's no big deal . When you get to mppt then it could be an issue if cost was no object but usually it's not cost effective to aim for extremely high efficiency on wind turbine generators that hardly ever run at full load.

Flux

[Ungrounded Lightning Rod]

 - Even harmonics are not a circulating-current problem because they are not present in symmetrical waveforms, which typical rotating machines (including machines hooked to diode/battery loads) produce (unless something in the machine isn't symmetrical).

 - (3N+1)th harmonics are not a circulating-current problem in three-phase because they have the same phase relationship as the fundamental.
 - (3N+2)th harmonics are not a circulating-current problem in a three-phase system because they have the opposite phase relationship from the fundamental:  They "spin" the other way but are still 120 degrees apart.
 - (3N)th harmonics are a BIG circulating-current problem in a three-phase system because they are all in phase.  Delta shorts them.

So the troublemakers are 3, 9, 15, 21, 27, ...  Of course with typical waveforms the higher you go the less harmonic energy you have.  So it's the bottom few of those that matter.

And the nonlinear load (which clips the tops of the waveforms sharply, producing lots of odd harmonics) "reflects" harmonics into the genny, which results in circulating current if it's wired delta even if its induced-EMF waveform IS sine.  Circulating current in a resistance represents an energy loss and stator heating, even if the lost energy is what "bounced back" from the diodes.  The genny ate it rather than swatting it back to the diodes and battery for another try.

Since we're going to have harmonics anyhow, flat-top generated waveforms throw more energy into the batteries per unit stator heating than sine, and flat-top waveforms are what you tend to get from the handy flat-sided magnets and flat-sided coils (which also "pave" your machine better), why bother going for sine in a genny designed for battery charging.

The main advantage of delta over Jerry/IRP is that you get a bit more current out of the two coils that are less-in-phase at any moment.  (Half that of the currently-in-phase coil.)  But you also get heating in proportion, because they're driving half the current through twice the resistance at I-squared-R.  So why bother.  Size your machine to do the job without this increment, give 'em some time to cool off on each cycle, and avoid losses and heating from circulating currents.

[ChrisOlson]

Size your machine to do the job without this increment, give 'em some time to cool off on each cycle, and avoid losses and heating from circulating currents.

Here's my practical observation; in the past month I've built two wound-for-delta stators and put on two of my turbines that had star-wired stators, plus rewired one that was in IRP to delta.  The one that was rewired - can't tell one bit of difference.  The delta stators in the other two put out slightly more power than the star-wired stators did, run cooler, and have no problems with cut-in, although one of them I'm starting up in star to get the low end voltage, and switching to delta at ~12 mph wind speed, so that observation is only really valid on the 12 volt one that starts in delta.

I ran one of those with my hydraulic motor setup I built to test generators with, using the published torque chart that comes with the motor to determine input to the generator shaft, and using my Sun AVR to measure output.  The delta generator is more efficient across the board, from cut-in to full power than the star one.  So someplace, in all this current racing around in circles, there has to be a practical application and testing applied, and proof that the star or IRP configurations are "better" than delta.  I haven't found that proof.  In fact, what I have found is an efficiency advantage in delta, and no difference to IRP other than reducing the amount of wire running down the tower.

Efficiency is were it's at with wind turbines.  The newest one that I start up in star to get the low end voltage and switch to delta appears to be the most efficient of all of them simply because it has less turns of wire in it than the other two.
--
Chris

[DanB]

Chris -  I've done the same and built several machine wired in delta intentionally, and also rewired stators into delta when the alternator was not a good match for the blades.  Most of the time they worked fine and when they didn't it had more to do with a bad match between alternator and rotor than it did to do with Delta.

If the machine built with any reasonable care then I think it's not too big a deal.

However, I do notice the parasitic currents - I have not measured them but I can feel it immediately as slight drag on the alternator just turning it by hand.

That said, I also notice the same affect whenever I wind coils with multiple strands in hand (especially when I use lots of strands, like 3, 4 or 5 in hand) - as soon as we fit the stator to the alternator, there is very slight drag and currents running between slightly unequal strands within each coil.  It's small I'm sure and very tricky to get away from.

It could be that it's just as well to wire in Delta and use thinner wire/fewer strands in each coil, than it is to go with star and have lots of strands or very thick wire.  Seems for the larger alternators that need to handle lots of current we either have to live with parasitic currents between phases, or parasitic currents within each coil.  And again, in any event it has never been a significant problem for me.  I tend to stick with star now perhaps just because I'm used to doing things that way and don't care to change because it works fine but I know that either way can work out fine.

For me, IRP is pretty much out, just because I can't get my mind around the idea of running 6 wires down teh tower or putting rectifiers at the top.  Perhaps I should get over my hangup there...

[ChrisOlson]

For me, IRP is pretty much out, just because I can't get my mind around the idea of running 6 wires down teh tower or putting rectifiers at the top.  Perhaps I should get over my hangup there...

Hi Dan,

I didn't word my post correctly about the delta/star comparison, but my point is that delta gets a bad rap and it's not deserved.  I too have problems with running six wires all the way to the rectifiers, and I won't put the the rectifier on the turbine and run DC down the tower.

According to what I've tested it appears that the best configuration for all around performance is to use star for cut-in and switch to delta as soon as practical after the turbine gets to putting out a few amps and the wind speed is sufficient to keep it putting out power at the higher rpm's.  There's not many cars that are built with only one gear in the transmission and changing to the higher "delta gear" after cut-in keeps the rotor running in its optimum TSR range over a wider range of output.

The problem is in how to do that.

On the one where I'm doing that presently, I have the relays on the turbine in a weatherproof box and I'm running wires up the tower from a toggle switch in the base junction to turn the relays on to switch it to delta.  The relays draw .5 amps when they're on so that's not the best method either.  Probably a mechanical switch on the machine with a governor or something would be better - I don't know.  It's possible your Lenz Effect switch that you experimented with could be used up there somehow.  One of these days I'm going to tackle that one.
--
Chris

[DanB]

Yes I agree, it could go a long ways towards improving efficiency at the higher end...
I do tend to find that in higher winds with a battery charging system I have usually more power than I need anyhow so I don't worry too much about that stuff, I'm not sure it's worth the trouble - but if a good/simple/bombproof/reliable sort of switch could be made it'd be a nice improvement.  (it has been done, Ed at windstuffnow.com was doing it years ago)

There would have to be some hysteresis in the switch, otherwise I expect it'll find a certain spot and start switching back/fourth quickly.

[ChrisOlson]

However, I do notice the parasitic currents - I have not measured them but I can feel it immediately as slight drag on the alternator just turning it by hand.

This is interesting.  I lowered my tower after supper tonight to close the air gap on my new turbine because it was running a bit fast and the blades were making a lot of noise.  I have been curious about this "drag" that people have described.  So as long as I had it down I came up with a rather crude method to try and measure it.  You'll probably laugh at it.

I took the rotor off the machine and screwed a fine thread stud into the end of the shaft.  I slipped a piece of rubber fuel line over that stud and hose clamped it, then did the same on the shaft of a heater motor out of a truck that I found in my shop.  I ran the heater motor with the battery charger so it would have constant voltage and used 6 volts instead of 12 volts.  I put a DC inline ammeter on the motor and turned it on.  It will spin the generator pretty fast but not to cut-in in star.  I think about 100 rpm and the motor was drawing 5.7 amps.  It was sort of hard to hang on to the motor and read the meter so I had my wife switch the turbine to delta for me while it was spinning.  When she switched it, absolutely nothing happened.  I figured maybe the amps would go up to 5.8 or something if there was drag in delta.  But it didn't.

Not very scientific, and maybe not even very accurate.  But it was fun even if I didn't learn anything from it.
--
Chris

[Beaufort]

However, I do notice the parasitic currents - I have not measured them but I can feel it immediately as slight drag on the alternator just turning it by hand.

I have been curious about this "drag" that people have described.

I did some testing on a single rotor last year and found measurable drag just by having the coils in the stator, not even hooked up.  The stator had pockets where I could replace the coils easily, and was able to get a result from just a single coil.  A full stator of coils hooked up in Star had more drag.  I can't recall if Delta gave more or less drag.  This was done with a prony brake and digital scale.  It was small, but it was there and I'm assuming it was eddy currents.  In fact, you could feel it just by holding a coil in front of the spinning magnets.

[ChrisOlson]

I did some testing on a single rotor last year and found measurable drag just by having the coils in the stator, not even hooked up.

I would assume that in this little "test" I did that my generator also had some drag on it - and it was hooked up to everything as normal.  I didn't think to test the motor for no-load amps but I'm reasonably sure it would be lower than when driving a mass of steel whirling around.

I guess it was just a thing I did in less than a half hour to test if there was more drag in delta.

I guess that I think that even if there is small amounts of drag there, it's no big deal with bigger turbines.  If you're playing with a micro-turbine with 4 or 5 foot blades you would probably want to reduce parasitic drag as much as possible.
--
Chris

[Ungrounded Lightning Rod]

Here's my practical observation; in the past month I've built two wound-for-delta stators and put on two of my turbines that had star-wired stators, plus rewired one that was in IRP to delta.  The one that was rewired - can't tell one bit of difference.  The delta stators in the other two put out slightly more power than the star-wired stators did, run cooler, and have no problems with cut-in, although one of them I'm starting up in star to get the low end voltage, and switching to delta at ~12 mph wind speed, so that observation is only really valid on the 12 volt one that starts in delta.

What is the shape of your magnets and coils?

Delta (in the absence of nontrivial circulating current losses) has less resistive loss than Y for a given volume of copper and frequency.  This is because Y sends all the current through two out-of-phase coils, while Delta only sends a third of it through such an arrangement and the other 2/3 through a single coil.  (And it provides more output current than Jerry/IRP, because the two off-peak in-series coils present more voltage together than they do individually)  So if your waveform keeps the circulating current component down enough that it doesn't eat the surplus, delta is the way to go.

[ChrisOlson]

What is the shape of your magnets and coils?

It has wedge coils and the 8" OD wedge magnets in it.  The generator rotors are 10" diameter, .2" distance between the coil legs, and I set the coil winder pin spacing so the N and S pole magnets precisely wipe their respective coil legs at the same time.  There's a small amount of magnet overlap on the inside of the coils when the magnet passes over the center of the coil.  It is only about .050" on each leg but using the wedge magnets vs bars reduced that overlap to almost nothing.  I would've had to increase the generator diameter slightly to get rid of that with a 24 volt due to the width of the coil legs.

It is a very, very smooth running generator and is considerably smaller diameter than my previous ones.  When it is running I can put my ear up the tower and hear a very slight "hum" in the tower but that is all the noise or vibration it ever makes at any speed.
--
Chris

[ChrisOlson]

What is the shape of your magnets and coils?

For ULR:

This is a photo of the stator when I was building it:


And these are the gen rotors that went in it:


--
Chris

[SparWeb]

I suppose I could add my own measurements - I just did the test so why not?


I guess I'm just contributing to the topic drift.  Delta obviously has more "drag" or parasitic torque that must be overcome to keep up any speed, even when it's running open-circuit.  But Jerry isn't free of it either - somewhere in between Star and Delta.  I didn't really expect that.

None of these power requirements are particularly difficult to deal with when the prop is 8 feet in diameter, and will be even more negligible when I make a 10' rotor.

(Okay the attachment doesn't show in the preview so I don't know if it looks right.  Only if I took the time to upload it to my own website and hyperlink it, can I preview the attached image.  If it doesn't work I'll correct it later this way.)

[ChrisOlson]

I suppose I could add my own measurements - I just did the test so why not?

I guess I don't understand what this is.  It looks like a generic graph.  According to this it takes 200 watts power input to turn this generator at 450 rpm, open circuit?  And only ~78 watts in wye?

That's not what I've found testing axial flux generators with my hydraulic motor setup.  The only parasitic drag I've been able to measure is what it requires to turn the bearings and I can measure torque input accurate to +/- 7 lb-inches with that setup.  It's accurate enough that I can hook up a 25 watt light bulb and it'll detect the load with an increase in the hydraulic pressure required to drive the generator at the same flow using a PFC (Pressure/Flow Compensated) hydraulic pump and a Ross TF-series torquemotor.  I apply the pressure vs flow chart from the torquemotor manufacturer to determine rpm and torque input, calculate hp and convert it to electrical, then measure the output of the generator with my Sun AVR to determine losses due heat.

The axial flux delta generators I've tested using this method are more efficient at converting input power to electricity from cut-in to full rated load than any of the wye-configuration generators I've tested.  The efficiency difference widens the harder you push the generator.

YMMV with these motor conversions.  I've never tested one.
--
Chris

[SparWeb]

Maybe your method and calculations work just fine - I can't tell.  I'm just pointing out that there is a way to measure mechanical power that is an order of magnitude simpler than what you do.  Perhaps the key missing ingredient is the formula for mechanical power:

torque X rpm = power.

I measured 36 inch-pounds of torque on the beam secured to the case of the Baldor motor as the lathe drove the shaft at 320 RPM.  To make the units work you need a factor of 84.52, so:

36 * 320 / 84.52 = 136 Watt

Since we're talking about mechanical work, you might think of horsepower frist, but 0.18 hp doesn't tell you much.  All I need is a bathroom scale and a 2x4!

This much power is required to keep the shaft turning at a constant speed.  The power is turned into heat in the iron.  It's mostly eddy currents and bearing friction in a motor conversion.  The eddy currents are nearly zero in an axial, so that's why they turn so freely.  Can't do much about the bearings, though.

I don't know much about using a hydraulic motor to drive an alternator.  That would work, but trying to deduce what shaft power from the hydraulic pressure and a flowmeter is hardly very accurate, when compared to a simple prony brake set-up, like I used for my test on the lathe.  Sounds like a bunch of comparison and look-up charts are required, too.

[ChrisOlson]

Maybe your method and calculations work just fine - I can't tell.  I'm just pointing out that there is a way to measure mechanical power that is an order of magnitude simpler than what you do.  Perhaps the key missing ingredient is the formula for mechanical power:

torque X rpm = power.

Actually, that formula is torque x rpm / 5252 = SAE calculated hp.

The hydraulic unit I built was built out of stuff I had laying around the shop and is much handier than the lathe because I can run big generators with it on the test stand.  My lathe only has a 7 hp motor in it and that hydraulic motor will make 40 hp.  With my shaft-type turbines it was very easy to rig up with a LoveJoy driving the turbine shaft.

And yes, I do have to apply the dyno chart for the torquemotor to calculate input power to the turbine shaft.  But it has worked reasonably well for me because it allows me to run a generator at a particular output for several hours if want without overheating or even working hard, and has given me a reasonably accurate benchmark to compare changes that I've made in different stator designs.

It seems I get a wild hair and wind a new stator almost once a week with a new idea that I come up.  I got better than 30 of them hanging on my walls and if anybody is building a turbine they could ask me if I got a stator to fit it - and I could probably go to the wall and grab one that will work.  They're all wrote on with permanent marker with the test data that I got out of them so I can remember what they were for, and what they did   

Some I've run on my turbines and some I haven't (yet).  But they're cheap to build (I get Essex mag wire from a local motor shop for $4/lb) and fun to play with in my spare time.  Some I've tested right to burn-up and I've kept those too - that's an even more interesting collection   
--
Chris

[SparWeb]

You don't need a 40HP motor to drive a 2kW generator.

Will this help?

http://www.youtube.com/watch?v=604dgSMzvu8

[ChrisOlson]

You don't need a 40HP motor to drive a 2kW generator.

I'm well aware of how much power it takes to drive a 2 kW generator.  Like I said, I had that motor laying in the shop and Ross Hydraulics was nice enough to send me a dyno chart for it.  I work on hydraulics all the time because all modern equipment is hydraulic.  So I have very accurate test gauges for pressure and flow and have several tractors with PFC pumps that I can hook up to run it.

It was handy and quick - all I had to do is make a bracket to mount the motor on my turbine stand and put a LoveJoy on it to drive the generator.

I've even used it to drive my Winco 12 kW single phase PTO generator.  I have to run a tractor at 2,200 rpm to drive that generator with the PTO @ 540 rpm.  Using the hydraulic motor to run it, I only have to run the tractor engine at 1,200 rpm and the hydraulic motor runs that generator at its 540 rpm shaft input speed without even breathing hard.  It's just slick.  Run the tractor up to 1,200 rpm so the swashplate can put full stroke to the pump pistons, and turn the flow knob in the cab until I get 220 volts from the generator.  With PFC pumps it automatically compensates for load on the motor and it stays dead on 220 volts no matter what the load on the generator.

This is one of those deals where "there's more than one way to skin a cat"
--
Chris