synchronous mod

[joestue]

Prior to tear down this was a 3/4 hp 3 phase Dayton motor with two burned out phases.
220/440 vac motor, 2 x 210 turns of what looks like 26 awg wire per phase. not too efficient it would appear, so i figured rather than rewind it I'd have some fun with this.
It would appear i can just thread rotor coils in there and use it as a generator and see what i can get the efficiency up to.
Pull the slip rings off of an old alternator and press it onto the shaft, with some kind of custom brush holder made up.
Can't be that hard...

Another option is to machine a .25 inch thick slot through the laminate stack and all the way into the motor shaft, leaving about .25 inch left in the motor shaft. (because the motor shaft shorts out the flux)
This would unfortunately weaken the motor shaft significantly, but i think it would hold up.  then, insert (2) .25 inch thick neodymium magnets into the slots.
the laminate stack is a convenient 1.5 inches thick btw.

However, if i try to do that then there's nothing to hold the "pole pieces" onto the shaft, other than the magnets alone.
I'm not sure epoxy can be trusted, not at 3600 rpm anyway, I think i would have to mill a slot in the newly formed pole pieces, cut a bar to fit into the slot, drill holes through the bar, laminate stack, and into the shaft. then thread the shaft and run a bolt down it.

Of course, now that i think about it, if i had not melted the alumimum out of the rotor, i would have been able to cut the slot into the laminate stack, insert the magnet into the stack, leaving the aluminum to hold the rotor together. (tensile strength of .999 Al, anyone ? .. yikes)

For the moment I'm going to try and clean the aluminum out, thread paper down the slots and run about 6 turns of #15 per slot and make 8 coils on each side of the rotor, run them all in series. should get a nice sinusoidal field to it. The Skew of the rotor is a bit steep for a 4 pole conversion, but if this works as well as i think it should, I have more burnt out motors to play with.

[bob g]

what is the diameter of the rotor?

there might be a lundel claw pole type that would fit in there nicely
with the requisite pole count to make it work

they are cheap as dirt, already have the slip rings in place, and brush holders are also cheap and widely available.

however it would appear you are wanting a 4 pole machine, so this probably wouldn't work.

funny this topic came up just now!

cool project

what is the goal?  what output voltage?

bob g

[joestue]

I could really care less about the output voltage right now, this one is practice for a much larger project.
The rotor diameter is 3.096 inches.



My options are to buy an 800$ (plus shipping) 7.5 KVA 1800 rpm generator or build it myself, by converting a 10 or 15 hp induction motor into a synchronous machine. A third option is to use an induction motor as is and throw run capacitors at it... but that is a mess.

I am hoping i can get a "stiff" enough machine that i do not need to play with reactive current in order to get a regulated voltage. also, i would like to get better than 90% efficiency.

The application is a 6 KW 440vac hydro installation, built on the extreme cheap. i'm casting and machining my own turgo spoons btw, and they are sized to run at 1800 rpm for this installation. (*currently on the first iteration on the spoons)

But the reason I need a properly engineered generator is because it needs to be able to start up and run a 5 hp 3 phase 440vac pump, 24/7 without a grid connection.

[SparWeb]

Once the aluminum is out, nothing holds the laminations in place, except the press-fit on the shaft.  Not that the aluminum is any harder than butter...

A solution I've tried (not with much success, mind you) is to drill a number of holes all the way through the laminations and put a bolt through.  Seems to be room for three bolts at 120 degrees on your rotor, about 1/2 way out the radius to be clear of the shaft and not interfere with the windings either.   

When it comes to machining the slots, better run the mill from the outside to about half-way in and stop.  Then come out, go around the other side, and mill that way from outside-in until the slots meet up.  The end-mill will push agains the laminations its cutting through, and if you go straight across, then it will get to the end of the slot and ultimately have just one lamination to push against.  The lam will bend over, rather than be cut, and you would end up with a mess.  It might have been better to leave the aluminum in place to "back up" the laminations, but it sure gums up the teeth of the cutters.  TANSTAAFL.

This will be fun to watch progress!  Keep us updated, thanks.

[joestue]

alright well i just pressed the lamination stack off the rotor... i mean all 58 of them one at a time with a screw driver.

its going to take a wire wheel to get the aluminum off the steel, then i'll just toss the stack in my 30 ton press and flatten them back out (they don't look half bad right now

unfortunately the number of teeth on the rotor is not divisible by 4 so its going to have to stay a 2 pole design, but its going to go back on the rotor without any skew.

[bob g]

the only head i am aware of in the 7.5kva and 800 buck range is an st7.5
they are rarely over 80% efficient in single phase, and not much better than mid 80's in three phase.

starting a 5hp motor with a motor changed to an induction machine with caps, is as  you say a real mess... probably would take a 50hp motor to be able to do that one.

i am impressed with your ambition, and envious about your available hydro source that can provide that sort of power!

yes, this is going to be really fun to read along about

thanks
bob g

[oztules]

Not sure i would discount the induction motor as a generator in this instance.

Because your only starting a centrifical pump, you initial start torque is not great, and I suspect a 10-15hp frame with star delta switching (maybe even some R in series for start up) may just get it over the line..

The pump motor will require some Vars as well, and so extra caps may be required, but for 24/7 operation, brushless would be explored first and formost.

Having wound a few starter motor armatures, squeezing wires up the channels is not much fun, but certainly do-able... and dammit I wouldn't mind trying just to see....

Next trick I would ponder is just milling the rotor for two big wound fields, the same as on a regular brush type genny...


Being naturally slack, I would explore the induction genny first ( at least the rotor resistance is virtually nill ).

And last, make your own pole shoes out of solid ( for the simple winding ), and perhaps cutting grooves in the surface of the shoes.


Nice water resource... drool....




...............oztules

[joestue]

so far the results aren't looking that great.
there are 51 turns stuffed in there and at 6 amps it feels like about half as much flux as what you would feel in a standard dc permanent magnet ceramic motor.

anyhow, i've run the numbers on the motor, the peak flux behind the stator's teeth had to be at least 1.9T during operation, the air gap flux had to be around .86T
~1100 amp turns from the line current alone.

to get 1200 amp turns in the rotor i would need to dump about 50 watts into the coils as they are now, best guess is i can about double, maybe triple the amount of copper in the rotor once it is properly wound.

--i think there's a reason they put fins on the aluminum rotors

[Frank S]

so far the results aren't looking that great.
there are 51 turns stuffed in there and at 6 amps it feels like about half as much flux as what you would feel in a standard dc permanent magnet ceramic motor.

anyhow, i've run the numbers on the motor, the peak flux behind the stator's teeth had to be at least 1.9T during operation, the air gap flux had to be around .86T
~1100 amp turns from the line current alone.

to get 1200 amp turns in the rotor i would need to dump about 50 watts into the coils as they are now, best guess is i can about double, maybe triple the amount of copper in the rotor once it is properly wound.

--i think there's a reason they put fins on the aluminum rotors

 Why not dump the THHN and use Magnet wire sleeve the slots then she-lack  the winding after finished.

[joestue]

the only head i am aware of in the 7.5kva and 800 buck range is an st7.5
they are rarely over 80% efficient in single phase, and not much better than mid 80's in three phase.

That is what I was afraid of.

btw, this isn't my hydro project.
i'm building it for a friend of mine, he has 155 to 165 feet of head and unlimited water, but the water has to travel 1000 feet to get to the turbine location.
initially it will be 3 inch pipe, upgrading later to whatever he can find, probably 6 inch.
--this is the major reason why the efficiency of the generator head needs to be as high as possible.

also, the electricity has to be shipped about 1000 feet to run the pump, which needs to pump water 200 feet up the hill.. looking at a gear pump to do that..
it is about 1500 feet from the turbine to the house, so basically we're looking at a 440 vac system, and some nice 98% efficient toroidal transformers to cut that in half to get 120/240

[joestue]

Why not dump the THHN and use Magnet wire sleeve the slots then she-lack  the winding after finished.

that is the plan but this coil will get me the information i need (assuming it doesn't burn up in the process.)
best case scenario is i can only just barely triple the amount of copper in that rotor by switching to #15 wire.

[joestue]

update:


(5ms/div)

This would appear to be mostly a waste of time.

6 amps flowing in 105 turns of wire makes about 8 volts RMS from a 50 turn coil at 1700 rpm.
said 50 turn coil was stretched across 12 teeth so its the highest voltage you can get out of the core.
interestingly it would be less copper to wind the motor as a toroid, rather than a traditional radial because the stack is so thin.

i only wound one test coil, so there is no harmonic cancellation going on.. you can clearly count the 34 poles of the rotor from that oscope shot!

also, at 630 amp turns at a minimum there is an extra 24 watts consumed by my variable speed 90 v motor driving this alternator.
i'll hook up an amp meter tomorrow on the dc side of the speed control to get a more accurate number (probably only 3/4ths of that after accounting for the efficiency of the scr speed controller and the dc motor) but from the looks of it, i've got about 18 watts of iron loss, and i haven't even hit .2T yet.

the rotor resistance is about 1/6th of an ohm, so i could run the current up to a more reasonable 12 amps for a more reasonable 25 watts of loss.

but, end result is for a motor this small... there is an order of magnitude more ampturns required than what you would expect.

[XeonPony]

you are using way thicker wire then needed for a field (If that is what the goal is) you are going for high megnetic flux thus more windings thinner wire, if it where to be a motor then ya you want higher current = higher rotational speed.

Look at the average generator head it is small guage ,magnet wire!

All so for the pump starting the Altinator will do squate for that, for that issue you need inertial mass to help over come the back emf of the inductive start. Read as BIG fly wheel bigger the better, for my little 2.8Kw synchronous system I was designing towards a 200 pound fly wheel!  Running at 1800 rpm I'd gear it up to run 3600 (The fly wheel that is) and that aut to overcum a pretty massive back surge from the inductive load trying to start with out stalling the altinator!

Very nice winding in either case!

[bob g]

if you are not aware (and that is a broad assumption on my part) of the
"transformer controlled alternator" you might wanna google the term
and read up on it.

they excel in motor starting capability

iirc stamford newage has a good explanation of their use of the system.

bob g

[joestue]

you are using way thicker wire then needed for a field (If that is what the goal is) you are going for high magnetic flux thus more windings thinner wire, if it where to be a motor then ya you want higher current = higher rotational speed.

meh, not really, but i was very wrong about some earlier numbers.

Right now with 6 amps flowing in that winding for 630 amp turns I am actually getting a peak flux of 1.28-1.32 T in the back iron behind one of the thinnest segments of the motor core.
I have 10 turns wrapped around the core, around the weld on the left side of the photo in the first post)
(0.1 volts per turn at 30 hz and .915 square inches makes 1.28T)

back to the Iron loss: I put an amp meter on my 90v motor and it jumps from .606 amps up to .837 amps at 1800 rpm for an apparent increase of 20 watts upon energizing the rotor.

My 38 turn test coil (not 50) which is wrapped across the entire width of the core (12 teeth) puts out 7.83 volts ac and this represents a 1.32T peak field strength. (air gap flux of significantly less, perhaps .2T)
Shorting it out causes 3.8 amps to flow through the 24 awg wire, causing the 10 turn coil to drop from 1.032 volts to .93 volts, a reduction of 10%.
3.8 amps times 38 is 144 amp turns.

Shorting out the 38 turn coil also increases the motor current by 320mA, an apparent increase of 29 watts.
the apparent wattage of 3.8 amps times 7.83 volts is 29 watts.

so i congratulate myself on discovering a source of free energy

next up is trying to skew the stator by about 3 degrees to get rid of that nasty 34th harmonic.

[bob g]

fwiw

about 30years ago, i found a delco generator from an old hd truck
it looked just like the more commonly available delco generator found on general motors cars and pickups and other machinery before about 1962.

it had the 2 pole shoes in the case which was wound with something on the order of 22 gauge wire iirc

the armature was wound in the typical manner using  a commutator

the rotor/armature was about the same diameter of yours, if memory service me.

i stripped the amature and rewound it fully with 15 gauge wire exactly as you have done, i then turned down the commutator, fitted it with fiberglass paper and two slip rings made from some copper drain pipe
1.5" iirc.
i then reworked the original brush holders to accept two half size brushes out of an old mt42 delco starter, and made all the connections

with 12volts applied to the field (within oem specs for the field) and spinning at 3600rpm it would make right at 130vac.

driven by a kohler 8hp engine, i could start a 3/4hp induction motor bench grinder, just as fast as it would start with mains power. it would yank it right to life, right now...

i put a scope on it, and because the rotor had no skew i also had a similar wave form, i too could count exactly how many slots there were in the rotor stack.

i had no way to balance dynamically the rewound rotor so it would vibrate much worse than i would have liked, however it was an experiment that i learned a lot from.

now i don't have any idea how much power it would make, i had no way to accurately measure things like distortion, amps etc. all i could muster was a voltmeter and an old allied knightkit scope. 

it turned out to be a useful machine for intermittent use, which i used it for back in a time and place where having remote 120vac 60hz power was unavailable and small engine driven generators were no where near as common as today.

my second project was a delcotron type 30, which for all the world looked like a miniature ST genhead, having a 4 pole rotor which was the field and a nicely skewed stator. i rewound the stator and was able to get just shy of 100vac from the machine at 1800rpm. i used a multitapped transformer to get me up to 130vac no load, but the droop was down to about 105vac starting the same bench grinder.

the interesting thing was the purity of the wave form, it looked every bit as good as that coming from the mains on the scope. having the skewed stator sure cleaned it right up compared to the former experiment.

also it didn't vibrate at all, of course not having to mess with the oem rotor saved me that hassle.

all in all i learned a lot working on those two units.

bob g

[joestue]

well this isn't dead yet.

I discovered my 1/3rd hp 4 pole single phase motor in the to be scrapped pile (due to a bent shaft and non replaceable bearings) has almost the same inside diameter as the motor i tore apart to do this experiment.
so i cut 6 blocks of aluminum to hold the new stator in the middle of the steel shell of the 3 phase motor.
It just so also happens to have nearly the same depth. However the air gap is significantly less.

the new (old) stator is 36 teeth while the rotor is the same 34 teeth.. i am going to go ahead and rewind the rotor to make 4 poles, and skew it while i am at it.
there will be some even harmonics because the teeth aren't divisible by 4 but supposing i were to cut a big square block in the middle of the stack and make a traditional salient pole alternator... i'd have to spend a week making the measurements and inserting the numbers into FEMM or they woudn't mean a damn thing so we'll see what it gets.
but i'm more concerned about getting a benchmark for the non variable losses.

again, the purpose of such an experiment is to see if the numbers support better than 80% efficiency for converting a 10 or 15 hp 3 phase induction motor to a synchronous generator.
granted the nameplate efficiency is 91% for such a motor, so in theory at least i should be able to get 80% efficiency using capacitors to provide the exciton..... but we all know that has issues regarding changing reactive power draw...

[joestue]

Well this is absolutely ridiculous.

The 4 pole stator saturates at ~32 volts and 60 Hz at 1800 rpm.

No, i'm not joking. The core cross sectional area is 45mm by 10mm, and where the bolt holes go through the core there is less, perhaps 6mm of iron.
i'll have to stuff a steel peg into those holes and see if that changes anything.
I wound a 4 pole rotor onto the 34 tooth rotor (well i messed up, there's a bit of a second harmonic, one of the poles (instead of a pair) has 3 coils instead of 4.)

the waveform doesn't look that bad but i haven't had a chance to load test it yet.

[joestue]

I will get some photos uploaded soon....just had my ssd die after only 5 months of use, running a new installation of ubuntu 12.? right now but i'm probably going to go for something more reliable.

Anyhow, i decided to scrap a 1 hp single phase motor that has the same lamination stack as the 3/4th hp motor in the rest of this thread.
Since both are exactly the same, i am able to fit both lamination stacks into one machine.
This will get me a rotor that is  3.096 inches in diameter and ~4 inches long.

one thing i did find Very interesting: I removed the run winding from the 1 hp motor by hand and fit exactly half of the run winding's copper into a ~150 watt transformer. (with the help of a vice and a hammer, had i taken the time to wind it properly, rather than 4 in hand, it would have fit without issue) so if you're wondering why they can't make 1 hp motors more than 60-70% efficient, well, there's not much copper in those things. i'll weigh it and get a number for you, but it wasn't much really, about 3-400 feet of 18 gauge wire, and that means about 100 watts of copper loss at full load, throw in another 50 watts of rotor loss, and 100 watts of iron loss (or more) and we're up to 250 watts of loss for a 746 watt output.. so that's 75% efficient and that is still too high...

Unless there is something radically different about 1760 rpm 15 hp 3 phase motors that say they are 91% efficient, its not looking good for building a synchronous generator from them if the intent is to make the motor produce full line voltage from the stock induction motor winding.
I will probably have to run it at most 80% of line, and then use a transformer to make up the difference.

[joestue]

photos of the new rotor, the stack is 4 inches long.
the skew is a bit steep, slightly more than one tooth.

[SparWeb]

Hi Joe,
I'm still following this, knowing I'm bound to learn something soon. 
Since you've tried a bunch of different things by now and this looks like iteration number 3, could you go over what the plan is now?  And what the result of the copper wire is now? 

I guess you've cut away the steel from the laminations a little (that must have been long and tedious) leaving those deep slots for the copper, before pushing them onto the shaft stock.  Is that copper varnished any?  Is it not shorting out on the steel?

[joestue]

i cut the teeth off the rotor because i haven't found any wound rotor synchronous machines that have teeth, all they do is short out the flux, and make inserting the coils more difficult.

The purpose of combining two motors onto one shaft is because with the prior stack only 1.5 inches long, the end windings on the motors were actually longer than the wires passing through the core. and with the small size of the motor, and therefore the maximum efficiency so low, the numbers were effectively meaningless.

end result is i might be able to get some real power out of this conversion.

[SparWeb]

Now I see it - there's a colour change looking at the laminations.  1/3 are from one rotor and 2/3 from another.  Hope it works!

[joestue]

about 1/2 pound of 26 awg wire, 7 coils times 70-72 turns per coil. with the much larger core i should be able to get more accurate numbers out of it.