3ph Motor-Alternator Conversion [Archive]

Rob Beckers

11th February 2008, 15:15

Hi Laura,

That power is when used as a motor. I have never tried using these things the other way around, so this is speculation on my part: I would guess the windings can handle 8.8 Ampere continuously, since that was what they were meant to handle as a motor. In terms of voltage, the windings will obviously sustain 230V without arcing. If you can get it to work as an alternator with that voltage and current then yes, it would be delivering 3.5 kW into a resistive load. The problem will be that if you want to direct-drive this alternator, the original windings will not make anywhere near this voltage because the turbine RPM is so much lower. If you have to rewind it anyway, then the current rating will depend on the wiring used and how much cooling you can provide (wind).

I've not dared to take a Scirocco alternator apart yet. They go for close to $8000 to replace if broken! So I can't tell you what size the neos are in there. I know it weights around 75 kg, and with the housing made from aluminum all that weight goes towards copper and magnets. Go figure. It's also rated for a whole lot more than the 6.5 kW that it's used for, probably closer to 10 kW. It is probably a bit larger than what you're looking for.

-RoB-

Rob Beckers

11th February 2008, 15:17

Come to think of it, you may want to contact Mario, and get the contact info for Jerry Ball. He has made a number of motor conversions and would probably be able to tell you what works best.

-RoB-

Laurie Forbes

11th February 2008, 17:28

Hi Laura,.......

Hi Roberta: :) Thank you for the comments - I don't plan on direct drive as, as you point out, the achievable voltage would be too low.

I would like to get in touch with Jerry Ball - is Mario on this forum by chance?

Rob Beckers

11th February 2008, 19:22

Sorry for mangling your name Laurie. Been staring at this screen too long today, the letters just run together at some point... :o

Mario de Castris (http://www.greenpowertalk.org/member.php?u=155) is on this forum, and he used to post quite a bit until last summer. Jerry is a buddy of his, but Jerry does not use a computer (not much, or possibly not at all, I had a discussion with Jerry about it last summer but don't recall the details). Anyway, just send Mario a PM and I'm sure he'll point you in the right direction. While you're at it, tell Mario he needs to post more, we miss him! :D

-RoB-

Laurie Forbes

11th February 2008, 19:36

Mark Parsons

12th February 2008, 20:09

Hi Laurie,

Thought I would jump in. Read your comments on the Fiberglass Blade thread and here. Rob is correct on the 3 phase power formula. Volts * Amps * SQR(3) = Watts. An NEMA design B (most industrial motors are this spec) TEFC induction motor mounted in a wind turbine application with ambient air able to circulate and rotor fan intact will supply at least 1.25 times rated nameplate current continuously without overheating damage.

You could self-excite the induction motor using capacitors if you can get the RPM's up to about 1/2 of nameplate rating. This would eliminate the neo retro plan.

If you are intent on retro with neo magnets then sizing is by magnetic field strength required per expected RPM and voltage. You plan on using N42 neodymium? What RPM and voltage is required?

Alternatively, as Stew pointed out there are options with using surplus / used industrial PM servo motors. These are ready to use as is with maybe some replacement of bearings. A member of this forum has access to used motors and e-bay also has quite a selection. Again assume that in a wind application you can draw 1.25 times rated max current without any overheating damage.

OT - I haven't forgot about the variable pitch project we discussed. It is just behind a couple of others. I am hoping to post a couple of new threads in the next few weeks on a 12' diameter HDPE plastic blade wind machine and a 15" culvert pipe microhydro propeller turbine that are taking shape in my shop's 'government' project area.

Regards,
Mark

Laurie Forbes

12th February 2008, 22:35

Mark:

Thanks for adding to the discussion - it looks like from what you are saying that I could get about 4 1/3 KW out of my motor conversion (230 * 8.8 * 1.73 * 1.25) w/o letting the smoke out! Looks like I may have to make the blades bigger than planned but I'm not sure I want to go much above 16 - 18'.

Re the capacitor excitation, not v good efficiency I would imagine plus need to spin it a lot faster(?) than a neo conversion. In regard to RPM and voltage, I would like to get 230V max at whatever RPM that turns out to be (hopefully quite low to facilitate a single stage "gear-up"- maybe 1/2 of what my current 1400 RPM ceramic mag conversion alternator does?). N42 is probably what I would use (seems to be best bang/buck) but I would consider higher if it would make any big difference (I expect lam saturation would be the limiting factor but I have no good idea at what N value or magnet thickness that would occur or, how to calculate even an approximation). Any help there would be appreciated as I hope to order the mags once the motor gets here and I can measure the rotor dimensions. I expect I will go with 1/2" thick neos in the absence of further information.

Please let us know how your projects are coming along.......

Patrick Trogothzon

13th February 2008, 01:55

Just a small link from *my* city/community/village (Vilhelmina, Lappland (http://www.sodralappland.se/External/default.asp?sURL=%2Fdefault%2Easp), Sweden): http://24volt.eu/hjul_generator.php

A picture of his rotor:
http://24volt.eu/bilder/vattenhjul/rotor300a.jpg

The site owner is working with the English version of the site, but all the pictures is in both Swedish, English, Chinese, Zulu-German and colour! :p

Laurie Forbes

13th February 2008, 11:07

Mark Parsons

13th February 2008, 15:05

Laurie Forbes

13th February 2008, 17:12

Hi Laurie,

You are planning on putting neos in the rotor and re-winding the stator to suit number of new neo poles?
You bought a 4 pole motor. Just putting neos on rotor for 4 pole won't give much voltage out. Likely less voltage than nameplate at rated 1760RPM.:(

Here is a pretty good reference from the Otherpower site on motor conversion:
http://www.fieldlines.com/story/2004/8/13/191519/161
http://www.fieldlines.com/story/2004/8/13/214144/026

Regards,
Mark

Mark, my plan is to do it the same way as my previous conversion which was an eight pole motor (the motor had been rewound b/f I bought it for 900 RPM) with eight poles of magnets (used ceramic mags in that case). It outputs 120 VAC at about 1400 RPM (using the 120 V winding connections). My 300 RPM turbine is therefore "geared up" by a 4.5 to 1 ratio.

For the new motor, it will have, as you mention, 4 poles of magnets. In this case, each pole will have more individual magnets to span approx 90 degrees as opposed to 45 degrees each for the 8 pole motor.

I have looked at the fieldlines site you cited. They are converting a four pole motor as am I and are placing four poles worth on the rotor (using multiple round magnets - I plan though on cutting slots in the rotor and inserting rectangular neos). The web pages did mention possible reconnection of the field windings to give more options as to voltage and current output but I saw no mention of a low particualr voltage problem by using only four poles.

I'm not sure why utilizing only four poles would necessarily not give much voltage out. There would not be as many individual windings in series as would be possible with a many pole alternator but OTOH would not windings in a four pole motor tend to have more turns (simply because they would be wider) and therefore give a similar combined voltage output?

I was as well interested to see they used 3/8" thick neos which addresses the question I originally had as to what thickness might be best.

Mark Parsons

13th February 2008, 20:28

Hi Laurie,

voltage = flux / time

The time component dictates a faster RPM (i.e. shorter time) for a lower number of poles to produce a given voltage. The flux is only changing 4 times per revolution in a 4 pole motor vs 8 times per revolution in a 8 pole motor.

To get 230V out you may want to consider starting with a 460V 4P motor for a lower RPM.

As you found in your 8 pole motor it took more than 900 RPM to create 120V. This would indicate that the flux density from the perm mags was less than the original induced rotor magnetism. Neos will provide more flux density than ceramic magnets. Enough to match the induced? I don't know. N42 has about 13,000 Gauss or 1.3 Tesla flux density. Another culprit is poor magnetic coupling from rotor to stator on retro projects.

IMHO - It is hard to beat the close magnetic coupling designed and built into commercially available motors. I am inclined to use this much higher efficiency and save my creative efforts for the turbine design and construction. You should be able to find PM servo motors inexpensively that will easily provide the power you wish to harvest and not lose potential power harvesting due to poor magnetic coupling from rotor to stator and extra gear ratios.

Regards,
Mark

Laurie Forbes

14th February 2008, 12:35

Hi Laurie,

voltage = flux / time

The time component dictates a faster RPM (i.e. shorter time) for a lower number of poles to produce a given voltage. The flux is only changing 4 times per revolution in a 4 pole motor vs 8 times per revolution in a 8 pole motor.

Mark, I'm not sure the flux changing more times per revolution necessarily equates to a faster rate of change of flux in the windings for any given RPM. As well, would not the flux per pole in a four pole alternator be greater than in an 8 pole because the individual pole magnets would be twice m/l large? Can anyone else chime in here - is there a mathematical relationship that could be used for a comparison?

To get 230V out you may want to consider starting with a 460V 4P motor for a lower RPM.

Yes, I may do that depending on how much voltage is generated. Even if the 460 config is used, I should still be able to get full o/p my 12' turbine is capable of supplying (the alternator at that configuration should be capable of at least 2 KW as I understand it). I won't be able to direct drive it in any case so at least a single stage gear-up will be required. The lower the alternator RPM the better I guess but if the gear-up is needed, may as well make the ratio as high as is needed to match the turbine oputput.

As you found in your 8 pole motor it took more than 900 RPM to create 120V. This would indicate that the flux density from the perm mags was less than the original induced rotor magnetism. Neos will provide more flux density than ceramic magnets. Enough to match the induced? I don't know. N42 has about 13,000 Gauss or 1.3 Tesla flux density. Another culprit is poor magnetic coupling from rotor to stator on retro projects.

Why would a retro project necessarily give poor magnetic coupling? I would take care to add as much magnet as will fit and keep the gap as low as possible.

IMHO - It is hard to beat the close magnetic coupling designed and built into commercially available motors. I am inclined to use this much higher efficiency and save my creative efforts for the turbine design and construction. You should be able to find PM servo motors inexpensively that will easily provide the power you wish to harvest and not lose potential power harvesting due to poor magnetic coupling from rotor to stator and extra gear ratios.

Again I don't quite understand your view point here. Apart from the issue of magnetic coupling, even if the coupling is "loose" does that not only imply that the max power attainable from the alternator at any given RPM would be less and not that the generation efficiency (energy out/energy in) would be impaired? IOWs, it may be necessary to use a somewhat physically larger alternator or gear it up a little higher to compensate but the power output should still be similar(?). Or, does poor coupling also imply energy losses within the alternator such as eddy currents or other heat losses?

I'm further assuming here that a PM servo would also have to be geared up but maybe that's incorrect.

Mark Ebert

16th February 2008, 12:17

I seem to be having trouble posting so if this goes through i'll add my input.

Mark Ebert

16th February 2008, 13:45

Hi again Laurie,
I've converted a 5hp 4pl motor in the past using 16 arc segment mags(N42 4 per pole 1x2x5/16)Skewed the best i could.Cogging was slight but noticable(could turn with 2 fingers)on the press i could get 100v oc@600rpm.The slight lumpiness(cog) revurberated through the blabes and drastically affected startup.All that work and i pulled it in a couple hours.

I concluded that running the neos in such close proximity to the stationary steel(lams)saturated them and the whole thing seemed mushy under load.
Iron loss is a killer with these type conversions in my opinion.Now if you could do it like Hughs breakdrum mill with no slots will be better.

I'm working on a dual rotor radial now.Not easy but fun!

Just my .02 worth.

Laurie Forbes

16th February 2008, 22:45

Mark:

Thanks for passing on you experiences with this stuff. I'm sorry your conversion did not work out for you.

In regards to this, I have posted questions to the Fieldlines forum as suggested by Mark (Parsons) and have received some encouraging feedback as certain individuals there seem to have had considerable success. One question I had in particular was how thick the neos should be with particular regard to lamination saturation. One person who replied ("Flux"), who seems v knowledgeable judging from othere posts I have read, says 1/2" is probably OK and, although saturation may impede startup somewhat, it should be _beneficial_ at higher RPMs. There was another long thread on that forum on the effects of saturation but I was not able to draw any particular conclusions from it. The consensus seemed to be it was bad but reasons were elusive - the question seems to be somewhat an art as well as a science.

I'm a bit disappointed to see you got only 100VAC at 600 RPM but what motor connection configuration were you looking at (230V)? It would be nice if you got that on a 120V config (but I doubt that would be available for a 5hp motor). Also, was it wired star or delta?

What did you use to glue the neos down with? Did you cut slots in the rotor or turn it down?

How large is your radial job? Were you able to salvage any of the neos (hmm, probably not as they were arc segs). Any pics avail?

Matt Frost

16th February 2008, 22:58

Hello Laurie,

I saw that you was looking for some information on motor conversions and thought I might be of some help, heres a post on one I did with voltage output numbers on down the post. http://www.fieldlines.com/story/2007/7/14/41633/0913

One thing I know is the more poles you have the higher the voltage at a lower rpm, I saw you planned on gearing but there are losses involved in doing so and I believe you can achieve better results going direct drive especially in lower winds. Even with lower rpm's you can achieve high voltages if you do a rewind and I think your perfectly capable to do a good rewind from viewing your other posts and work you have done. If you rewind I highly recommend skewing the lams or the magnets, if you use as much magnet as I have without skewing you would need a long pipe wrench just to be able to turn the rotor. Mine stator is skewed 10 degrees and has no cog at all, very easy start up.

You wondered about magnet size, I used 24 N42, 1x2x.5 with center hole for mounting, available on ebay. Many have said, and I believe it to be pretty true that you get 150 watts from every cubic inch of neo magnet and sometimes better, depending on grade used of course. Mine puts out 4Kw, 6 amps per phase at 220 volts across any 2 of the 3 phases, that happening at 55o rpm which is a bit high for a 16 to 18 foot prop but I could have got that at a lower rpm if I had taken more time and packed the slots with a little more wire as I had more space to do so. I don't think you have to worry about iron saturation, my output numbers show that.

If your only after high voltage I would make all connections internal and only bring out the 3 output wires, a little less confusing in the long run but not as many output options.

Hope this helps some and good luck which ever way you go. :)

Matt

Laurie Forbes

17th February 2008, 11:08

Matt:

Fortisimo (and very informative)! I wish I had ready access to a lathe and mill but alas I don't and will probably have to pay some machine shop an exorbitant fee just to turn down the rotor. I'm pretty much resigned to doing a gear-up to get the voltage I want (230) (w/o compromising current capacity which I believe more coils in series, to get higher voltage, would do). So then, I don't think I will rewind the stator, at least with this iteration.

Actually though I don't think a gear-up, while perhaps not being as esthetically pleasing as direct drive, is all that bad and is certainly not that hard to implement. Wear and tear is an issue as well as a bit of lost energy but one other advantage (maybe a small one) is that it allows the alternator to be situated behind the yaw axis which reduces load on the yaw bearing considerably (especially with all the forward weight I have on my turbine from the blade pitch control mechaism).

I'm please to see you are not getting saturation with those large neos. I think I will do it similarly but maybe in smaller peices (1 x 1/2 x 1/2?) to better allow a degree of skew w/o increasing the gap unduly.

You mention 550 RPM being too high for 16 to 18' prop - IMO 550 may be too high for even a 12 footer, based on my own 12 footer experience (mainly a gut feeling watching how it behaves and sounds especially in strong gusty winds). I limit my current rig to about 300 (which produces 120V at about 1KW with a 4.5 gear-up).

You also mention getting 4KW at 220V and 6A per phase. I calculate only 2.3KW with those numbers (220 x 6 x 1.73) or maybe I misunderstand your calculation(?).

Please let us know what you decide on for a prop - I'd be v interested to see what kind of direct drive performace you can derrive from your alternator.

Matt Frost

17th February 2008, 22:22

yeah I guess thats what I get for trying to do too many things at one time, it should have read 10.5 amps and thats only known from the lathe testing we did, I'm pretty sure that gen could never see that from even a 16' prop geared, not positive on that though.

That gen hasn't got a home just yet, still sitting garage since I had a huge tower raising problem that won't be fixed until this spring and even then I'm not sure I am going to put it up as I am currently building a much larger unit. The larger gen has a rotor diameter of 10 inches and will use 40- 2x1x.5 neo's , 20 poles in hopes of higher voltage at lower rpm(150 or so) with about 20' to 24' prop. Any how maybe I'll change it with the one my dad has flying now (a small dual rotor).

I know it has great potential, I hooked a 110v 100w light bulb to it and a hand crank to the shaft and I can hand spin it up to about 90 volts fairly easy just not for a long period of time, I was extremely amazed by that since it was my first conversion, I couldn't hardly believe it.

Your probably right on the gearing it up, I had considered doing that but hadn't seen anyone else do it.

I see what your saying on the weight issue and being able to move some of that weight to better balance things.

How much do you think your pitch control weigh's?

I'm going to be working on a torque controlled pitch mechanism for my larger unit which will also need a variable load but I don't see any problem there since I'll be switching on heating elements as the available power comes on. I think my pitch mechanism will function similar to yours in operation, steep start up angle then less angle, then into stall and if load is lost reverts back to starting point with steep angle to keep speed down to a safe level.

Skewing the magnets will work out fine as you said and allow you to keep the factory winding in place. Glad to see others experimenting with these conversions.

I agree on the prop speed, my dad has an 8 footer running around 550 to 600 rpm and theres no way I would try to push a 12' mill into that, not with me there anyhow :) his seems to do just fine that way though, has all winter even through some 60 plus mph winds we have had here, of course it is furling starting around 25mph.

Just curious, why the 230v? what are you using with that voltage? heating of some sort?

Matt

Dan Lenox

18th February 2008, 08:33

Fortisimo (and very informative)! I wish I had ready access to a lathe and mill but alas I don't and will probably have to pay some machine shop an exorbitant fee just to turn down the rotor. I'm pretty much resigned to doing a gear-up to get the voltage I want (230) (w/o compromising current capacity which I believe more coils in series, to get higher voltage, would do). So then, I don't think I will rewind the stator, at least with this iteration.

Laurie,

I have a good sized metal lathe and milling machine, would be happy to do the job for you at no cost

If interested contact me off-line: dlenox-at-briery-dot-com

Dan Lenox

Dan Lenox

18th February 2008, 08:36

Laurie,

All I ask is that you pay shipping both directions...

Dan Lenox

Laurie Forbes

18th February 2008, 11:53

Laurie,

I have a good sized metal lathe and milling machine, would be happy to do the job for you at no cost

If interested contact me off-line: dlenox-at-briery-dot-com

Dan Lenox

Dan, thanks very much for your kind offer. I have recently made contact with a wind home brewer close to here in Alberta - don't know if he has a lathe yet. If I can't get assistance from him I will definately contact you off list.

Thanks again......

Laurie Forbes

18th February 2008, 22:34

Matt:

Is that 10" diameter rotor a motor conversion? If so, what HP a motor would that be?

The pitch control plus three blade hub weights about 70 lbs. It could be certainly made lighter but OTOH, as it exists I'm sure it could handle a prop larger than 12'. If/when I do a 16 or 18 footer, I will probably use the same mechanism (asuming it holds up till then).

I'm v interested to see you are considering a torque contolled pitch mechanism as I had been considering that myself. It would have the advantage of not requiring fly-weights but I'm not sure how it would work as could you not have a situation (especially if a variable load is employed) where the torque has the pitch set to before stall but the RPM is still too high? Also, if the load is lost altogether, would reverting to the normal low torque coarse pitch be necessarily enough to prevent overspeed in a v strong wind?

How would you translate torque into blade shaft rotation? A central bevel gear on the main shaft connected to the main shaft by a torsion spring and with bevels on the three blade shafts would do it and, serve to keep the blades synced. One possible problem I see though is adjusting the action - it would seem this could only be done by changing the torsion spring to that of a different spring rate or changing the bevel gear ratio (not as readily accomplished as simply adjusting the position of fly-weights).

Re skewing the magnets, what would you think of simply staggering them (assuming rectangular magnets that is)? It would serve to reduce the gap as the magnets would remain straight on the rotor axis(?).

I'm looking for 230V to run heaters which is all I plan for as this is a hobby windmill project and I don't foresee adding batteries, a charger or inverter etc in future (maybe a grid tie-n though if sufficient financial incentive is ever offered where I live).

Matt Frost

19th February 2008, 00:41

Laurie:

This would require quite a bit of time for me to answer in detail and I want to include some pic's which I need to take to help explain some, which I will try to do in the next day or 2. It's pretty late right now.

"Re skewing the magnets, what would you think of simply staggering them (assuming rectangular magnets that is)? It would serve to reduce the gap as the magnets would remain straight on the rotor axis(?)."

Sorry for the confusion, thats what I meant "stagger" and not actually angle them (assuming rectangular magnets), as true skewing them would create too much gap I think and large losses of expensive magnets, somewhere there is a pic showing the magnets staggered (rectangular like were speaking of) but I'm not sure where and the user reported good results in doing so as far as cogging goes.

"I'm looking for 230V to run heaters which is all I plan for as this is a hobby windmill project and I don't foresee adding batteries, a charger or inverter etc in future (maybe a grid tie-n though if sufficient financial incentive is ever offered where I live)."

Same here on that thought.

I will post pics of a torque controlled design and want to hear your thoughts and or anyone else's for improvements or faults with it. I'm not dead set on using it, I have seen a lot of designs similar to yours using weights and I think they work fine as well, I tend to think like you and I'm not afraid to tread where few others have gone even if theres not a huge gain, I enjoy the challenge as long as I can keep the cost reasonable.

I would also be interested in sharing/ comparing electrical control design (variable load controller) if your going to be using any, Turn loads on and off as per wind conditions.

More later

Matt Frost

19th February 2008, 20:56

Ok, more time

The 10" rotor is / will be a motor conversion but basically the only part of the original motor that I am going to use will be the laminated core because the motor has a cast iron housing, (total motor weight estimated 300+ pounds), the new housing is 1/8 sheet (in pic) rolled, the new ends will be aluminum and the new rotor will be of my hollow design like in my other conversion. The new entire assembly will be much shorter in length as well to help keep the weight down.

392

The motor is / was a 25hp Delco Remey, 3 phase 460 volt with 36 slot core, the core will be modified to have 60 slots to fit my needs, a lot of work but I think it will be worth it, more slots = more wire + more poles = more voltage at a lower rpm which is what I need.

Lets move on to this torque pitch hub, like I said I'm not sure this is the way to go but I'm not sure it wouldn't be beneficial once I got it all working either.
I do like the fly weight design for simplicity and adjustability.

As I think about each design (flyweight and torque controlled) I think they both work very similarly in that as load comes on, the blades would slow and the pitch would change to a more favorable condition unless the wind was to get stronger, at which point another load would be added and so on ......

Here are the pic's of the hub,This is not my design, I found these pic's here on the net and resized them for posting purposes. Your gonna really want to study the last one to figure out how it works, the others just show individual pieces to help understand how it goes together. I have to add if I make one like this there will be changes made (stronger built, only a couple pieces, 3 blades instead of 2, the stops need to be repositioned elsewhere I think to allow more travel "pitch angle" and maybe adjustable gas filled springs instead of the conventional coil type, probably relocated as well).

394

395

396

397

Looking at this bottom pic, the side with the springs would face the generator and the side facing down would be the oncoming wind side and the unit would rotate clockwise as seen from the front or wind side. You kind of have to imagine what would happen with the unit not having a load applied, that would be your start up position (as well as dropped or failed load "revert" position). Then imagine some resistance "generator load" being applied to the center shaft, then the blades would pitch to less angle and speed up until the gen was loaded harder or at max load condition the blades would move to a negative angle causing stall, very similar to yours I believe.

That's my theory on operation of this design, feel free to offer your thoughts if they differ from mine as I may be missing something, I have studied this design and run the workings through my mind for quite some time, I hope I have it figured out right.

Rob Beckers

20th February 2008, 07:04

Hi Matt,

That is an interesting hub! It reminds me somewhat of the (much simpler) Jacobs pitch-change hub. I can see how the pitch changes with this design, and how the springs try to push it back. What I don't understand is by what mechanism is the pitch change induced, i.e. what causes that torque-plate to rotate?

-RoB-

Mark Parsons

20th February 2008, 08:44

Greetings Matt and Rob,

I have seen this design before. The blades will pitch based on torque differential between wind supply torque and generator load torque.

A couple of questions came to my mind upon first inspection. 1) - Is that a relevant association (control mechanism) to harvest maximum energy from wind power? 2) - This is not a safety furling mechanism, or is it?

Is there any further info from the builder of this design regarding its operation?

Regards,
Mark

Laurie Forbes

20th February 2008, 11:24

Matt, that's an ambitious motor conversion project! It will be v interesting to see how low you can get (RPM wise that is :)).

I understand (I think) how the described pitch control mechanism works but I must say it seems kind of complicated. I"m still partial to a bevel gear mechanism with a torsion spring on the main shaft. The mechanism you describe seems to be more easily adjustable but on thinking a little more about a bevel gear solution, adjustability may not actually be that difficult if done properly. What I'm thinking is the main bevel gear would be "forward" of the blade shaft bevels and the torsion spring would be forward of that. In that way the torsion spring would be easily switchable to something of a greater or lesser spring rate (just remove and replace it with no removal of the bevels required).

On the topic of torsion springs, on the Otherpower forum it was suggested that piano wire could be readily used to roll (literally) your own and I think I will try that for my flyweight controller project. Similarly it may not be too difficult to make a few main shaft torsion springs and experiment as required. They could even be made from the same diameter wire - just vary the number of turns and fine-tune it as closely as you want.

A particular advantage of a torque control mechanism vs RPM/flyweight control is that, at least with the bevel gear setup, a wider range of operation should be possible as the prop can rotate differentially to the main shaft, under torque, pretty much all it wants (within the limitations of the torsion spring) while with flyweights, the useable range is limited to about 45 degrees. This might allow a greater degree of feather at low speed, or loss or load (I'm still wondering how much feather would be required to prevent overspeed in a loss of load condition and a v strong wind).

Another advantage of torque control may be that the motive force is more linear that with flyweights.

Matt Frost

20th February 2008, 19:54

Hi Mark,

"I have seen this design before. The blades will pitch based on torque differential between wind supply torque and generator load torque."

Correct as I see it also. From what I have can tell this design would be of no use for any type of fixed load, unless the load was "heavy" which would defeat the purpose I think, so that is a downfall with it.

"A couple of questions came to my mind upon first inspection. 1) - Is that a relevant association (control mechanism) to harvest maximum energy from wind power? 2) - This is not a safety furling mechanism, or is it?"

Yeah me too, answer to

1) I think it is, the only other thing I have heard about this design is that it was meant to be used with mppt controller used with heating elements. I was hoping someone else could shed some light on it.

2) I think it could be if the torque of wind was matched to torque required by load enough to put the blades into negative pitch angle causing it to momentarily stall, kind of toggle back and fourth, maybe....maybe not, it is complicated and might require a smaller prototype for testing.

HI Rob,

"That is an interesting hub! It reminds me somewhat of the (much simpler) Jacobs pitch-change hub."

I guess I havn't seen that "Jacobs pitch-change hub", I googled for it and couldn't come up with anything.

Would you happen to have a link on it? or more info?

Laurie,

Yeah some project huh, I got to be nuts. It will be time consuming but I'll post up dates occasionally as I go.

"I understand (I think) how the described pitch control mechanism works but I must say it seems kind of complicated. I"m still partial to a bevel gear mechanism with a torsion spring on the main shaft. The mechanism you describe seems to be more easily adjustable but on thinking a little more about a bevel gear solution, adjustability may not actually be that difficult if done properly. What I'm thinking is the main bevel gear would be "forward" of the blade shaft bevels and the torsion spring would be forward of that. In that way the torsion spring would be easily switchable to something of a greater or lesser spring rate (just remove and replace it with no removal of the bevels required)."

Yes, maybe too complicated for what I really need, maybe if I knew more about it then it might be worth making.

Do you have any pic's of this bevel gear design your talking about?

Are there any easily acquired bevel gears that could be used in this design?

"A particular advantage of a torque control mechanism vs RPM/flyweight control is that, at least with the bevel gear setup, a wider range of operation should be possible as the prop can rotate differentially to the main shaft, under torque, pretty much all it wants (within the limitations of the torsion spring) while with flyweights, the useable range is limited to about 45 degrees. This might allow a greater degree of feather at low speed, or loss or load (I'm still wondering how much feather would be required to prevent overspeed in a loss of load condition and a v strong wind)."

Sounds interesting, I'd like to know more, sounds worthy to build and test to me. I think that with enough heads thinking about a design as such that we could come up up with something pretty trouble free with the ability to use as much of the wind as possible all while protecting itself from destruction.

Laurie Forbes

20th February 2008, 21:32

Matt:

No pics, it's all in my head (stuck there until I get that cranial USB port installed :)).

Actually it's not that hard to visualize. A "large" bevel gear is installed on the main shaft and is keyed to the shaft. One "small" bevel gear for each blade shaft is clamped to the end of the shaft(s) and meshes (at right angles) with the "large" bevel.

The prop hub is free to rotate on the main shaft but is constrained by the torsion spring, also on the main shaft. Torque is transmitted from the prop hub to the large bevel via the torsion spring and thus to the main shaft and alternator i.e. one end of the torsion spring attaches to the blade hub and the other to the large bevel. Since the large bevel is keyed to the main shaft, the torsion spring could also be attached directly to the main shaft if that simplifies construction or access.

As load (torque) builds from increasing wind speed, the torsion spring yields and the prop hub thus rotates relative to the large bevel, this in turn causing the small bevels to rotate relative to the prop hub, this in turn rotating the blade shaft(s) and changing the pitch.

The mechanism could be adjusted by rotating the blade shafts relative to their small bevels to change the starting point pitch or, by switching to a torsion spring with a different spring rate to change the end point (into stall) pitch (the torsion springs could be constructed with the same diameter of piano wire and the rate changed by varying the number of turns).

I think......

As for sourcing bevel gears, one supplier I have looked at is:

http://www.qtcgears.com

Metal bevels ain't all that cheap though.......

Matt Frost

22nd February 2008, 21:19

Laurie:

Would you still want a blade with twist or would it not be necessary with pitch control?

Laurie Forbes

23rd February 2008, 13:45

Laurie:

Would you still want a blade with twist or would it not be necessary with pitch control?

Matt:

I would seem to me that the two are pretty much mutually exclusive - pitch control would be valuable with pretty much any type of blade and twist will always give better performance over non-twist.

BTW, visited a local wind/solar outfit for a tour of their facilities. The guy mentioned they always spec an emergency dump load for their turbines in case of normal load failure as they have seen cases of furl turbines flying apart when the load was lost - another advantage of pitch control as load loss makes no great difference to overspeed control. Even if however the "emergency" dump load is present, if the electrical failure occurs upstream the result is the same.

All this was of particular interest to me as I had wondered what would happen to a furl controlled turbine under no-load conditions (in a strong wind).

Martin Gray

24th February 2008, 14:05

Hi Laurie:
Here is an idea I have used, to build a pitch control unit using gears
This one pitches to feather, as opposed to your pitch to stall.
401
This is the finished unit with shock absorbers and tension spring (hidden from view)
402403
Since your talking gears,I thought you might need yet another gear idea in your head. :)

By the way, nice pitch control you built there!

Martin.

Rob Beckers

24th February 2008, 16:33

Very neat way to keep the blades in sync Martin! Hadn't seen that one before. Minimal part count too. Impressive!

-RoB-

Laurie Forbes

24th February 2008, 21:00

Martin:

Thanks for posting that. It's certainly an interesting way of doing it but it looks to me that one of the blade shafts (the one with two bevels, is turning the opposite direction to the other two). I guess this might still do what you want though as two blades would go to feather and one to stall (but one wonders what aerodynamic issues that may bring up). Even so, this method still requires four bevel gears however they are all the same as opposed to my proposal which requires one large and three small.

What's the function of the triangular plate in the front? I also note the design also employs dampers (shocks) but I wonder if you found them to be a necessity (my pitch control rig doesn't have them and seems to work OK)?

May I ask where you bought the bevels - I've found a couple of sources but they are pretty pricey.

In any case, I think bevels are a better solution to the blade sync requirement rather than a "spider".

Mario De Castris

24th February 2008, 22:10

Hello all,

It's been a while sense I have done any posting, too busy I guess or maybe just too lazy. While on the subject of pitch control devices here is another one for you ambitious guys. This unit is manufactured by a friend that some of you have met, Jerry has been making this hubs for at least 20 years. If you have any questions I am sure Jerry will be more than happy to answer them.

Mario

Matt Frost

24th February 2008, 23:38

Laurie:

"Very neat way to keep the blades in sync Martin! Hadn't seen that one before. Minimal part count too. Impressive!

-RoB-"

I have to agree with RoB on this hub, very simple, low part count, easy to maintain and also pretty robust.

I was a bit confused as well on the gears not turning the right directions so I copied it to paint, drew some arrows and low and behold the design works out perfectly, all blades would move in the same direction. One might also add 2 more of those gears, making 6 all together and they would all work out just fine.

It appears that these gears would be the spider gears from axle differential, if they are not then I think diff. gears would work, not the ring and pinion, but the smaller internal gears which I'm sure would be plenty strong enough for this application.

One last thing, working in a machine shop has led me to another source for bevel gears. Ebay has tons of them, many at pretty low prices. -3- jaw lathe chuck, they have the larger center bevel and 3 smaller bevel gears. My boss told me to open one up and check it out so I did, would have been nice to have had a camera then but I didn't. I think a chuck in the 6" to 8" range would work pretty good. I did look on Ebay and found several missing the chuck jaws for CHEAP, the jaws wouldn't be needed anyhow. Just a thought. I can open a chuck and take some pic's if you like.

I do like Martin's design for simplicity, my question is though, would the centrifugal weight design have enough rotation to start with coarse pitch, say 60 degrees, and be able to pitch to stall?

What negative angle would cause stall? -5 degrees

Wouldn't a total of 90 degrees be possible with Martin's design?

Martin feel free to chime in, what is the total possible movement in degrees of your hub?

Do you have it flying and if so does it work well? Would you make any changes to your design? And are the cylinders, dampers or gas charged shocks?

Martin Gray

25th February 2008, 08:16

Laurie:
Well, they do all turn in the same direction. A little confusing, looking at the pictures and trying to follow the gear movments.
The triangular piece is to stabalize the inside bearing mounts.
The dampers are used to slow down the movement. With out them, they seem to snap back a little fast when going in and out of feather. I'm tring to make the tranision more fluid.

Matt: Yes the gears are from a differential. I don't know what make or model there from. A little squrounging at the wreckers or a drivetrain repair shop should produce some. (cheap)

This unit is designed to pitch to feather, not pitch to stall. It does more of a rotation than Laurie's, which flattens out the blades to stall. I get almost 60 deg. in movement.

The hub has worked well when bench tested between 250-300 rpm, exept for the dampers which were just added, and not tested yet.
It might be a whole different animal when the blades are on and mounted this spring.
I must also say that gears are a bit difficult to set up. First you have to much play, then they are a little tite. When I found the happy spot I welded them to the shaft. So theres no chance of misallignment in the air.

Martin.

Laurie Forbes

25th February 2008, 11:50

Well blow me down, now that I'm a little more lucid after a good night's sleep I see they do turn in the correct direction. I should have realized that because I had designed a somewhat similar four blade hub but with the shafts connected by sprockets (one on each) and roller chain (the shafts being offset as in your design).

Anyhow this has me all the more interested in bevel gear blade sync as it appears the bevels should be pretty inexpensive, either, as mentioned, from auto wreckers of from wacked out lathe chucks.

The geometry is a little had to visualize but it also seems to me that the range of motion should also be greater than my design because of the offset blade shafts.

This design also seems only to allow pitch to feather(?) which precludes starting towards feather for better starting but I'm not sure just how valuable that characteristic is anyhow (in my case, it still seems to require the low RPM unload circuit to start properly).

Matt:

I would like to see pics of the bevels inside a chuck (also diferrential gears if you have'm).

Martin Gray

25th February 2008, 18:24

Matt Frost

25th February 2008, 19:17

Laurie:

Heres a few pic's, the first being the lathe chuck internals, last being automotive differential gears.

408

3 jaw 8" metal lathe chuck (outside case diameter is 8"), I didn't measure the large bevel gear but I think it's outside diameter is about 6", inside diameter roughly 4" and approximate height is about 3/4". Smaller bevel gears are about 1 -1/4" diameter.

409

I think if one was to use these gears or like ones a hub could be made,or one modified, pressed in and welded into the larger bevel gear. I think for the smaller ones a square could be milled or ground on the end of a round shaft, inserted into the end and welded in, I think a good close fit would would help to eliminate alignment problems.

410

These gears may appear small, and thats because they are, just for picture purposes the cover was already off this one (small jeep axle), if you were to try these type gears I would use something from a 1/2 ton truck or the like, 3/4 ton axle might be a bit big or they may work fine.

411

Readily available at all used auto parts, downfall is you'll need to scavenge from 2 of the same or similar axles as each axle has 2 larger and 2 smaller ones. ( '78 and '79 chevy truck and blazer 1/2 ton axles are plentiful in the states)

Good point on these though is they all have holes "through" them, about 3/4" in the smaller ones. Pretty easy to fit to a shaft and weld on like Martin did.

Laurie Forbes

25th February 2008, 20:12

Laurie:
Now I'm confused. Looking at your pictures (on fieldlines). If your rotor turns clockwise viewing from the front (upwind) then it looks like the weights would twist the blades clockwise, viewing from the tip to the hub. How would the blades twist counter clockwise with the weights in that position?
Help me out here Laurie.
407

Martin:

I understand your confusion - it's not readily apparent what's happening. One might suppose the fly-weights would move forward (and thus twist the blades clockwise) but they actually move backward because centrifugal force is trying to force them to a larger rotation radius and that radius increases as the weights move back (that is, the radius as measured from the weight to the main shaft). The radius (relative to the main shaft) increases as the weights move back due to the weight's forward offset from the blade shaft.

If the weights are positioned as far forward as possible, there will be no motivation to move (in either direction). If the weights are back far enough such that they are "above" or in the same rotational plane as the blade shaft(s), the motive force will also be zero. It is necessary therefore to start with the weights positioned somewhat back of full forward and to keep in mind that the rotation force, or torque, (on the blade shaft) diminishes to zero as the weights approach their farthest back position. It would be an interesting calculation to find the weight position that produces the greatest torque on the blade shaft. If the weight forward offset arm is equal in length to the distance of the main shaft to the attachment point of the offset arm to the blade shaft, I expect the maximum torque would be exerted when the weight offset arm is tilted back about 45 degrees. The useable range of motion also seems to be about 45 degrees (eg. start at rest from 22.5 degrees and travel to 67.5 degrees at full RPM).

To pitch to stall, the weights must be positioned to "lean" into the direction of main shaft rotation. If one wants to pitch to feather, the position would be a similar number of degrees leaning "back" (in either case the weights will want to travel "backwards" relative to the main shaft axis, turning the blades counterclockwise in the first instance and clockwise in the second).

It's hard to put this into intelligible form - I hoped that helped.

Laurie Forbes

25th February 2008, 20:31

Matt:

Good Pics!

Man, that set of gears from the lathe chuck would work very well in my design (central large bevel surrounded by three smaller ones with 90 degree change in axis direction). The central bevel is larger than necessary but hey, it would work. As you point out however, some metal work would be required (which is a downer for those of use who are underprivileged in that respect :))

Those differential gears might also work - I wonder if there would be enough room for three gears around a central gear if all were the same size? I also wonder if auto wreckers would have the gears separately or would one have to buy the entire assembly & case?

For Martin's configuration however, since all the bevels are the same, it seems that the bevel angle would have to be 30 degrees only to produce the required 120 degree change in axis (for the gears to mesh properly that is) thus limiting what could be found and used(?).

Mark Parsons

6th March 2008, 19:53

Hi Laurie,

Just ran across one of your otherpower postings. Did you find a source for acceptable torsion springs yet?

If not, have you checked McMaster Carr? A large industrial supply catalog house. They stock everything and ship next day. Their torsion spring selection is catalog page 1194 - http://www.mcmaster.com/
Selection chart shows coil OD, wire dia, leg length, torque, deflection angle, priced in Type 302 stainless and music wire. Prices are US$ ships from Ohio.

Regards,
Mark

Laurie Forbes

9th March 2008, 22:14

Mark:

Haven't found a source of springs but have downloaded s/w that lets you design springs - it looks like what I need is something made with .221" music wire with five coils (on a 1" dia mandrel).

I happen to have a McMaster-Carr catalogue. They do have an extensive spring supply but nothing I could find that looks like it could work (insufficient wire diameter on all their offerings at least as far as I could find). They also have bevel gears but again, the prices here too are not very attractive IMO.

BTW, I'm building an anenometer now using half plastic practice baseballs for cups and the bearings from a fridge fan motor. It has a Hall effect magnet sensor which I have connected to a microprocessor for pulse counting and conversion to wind speed. I plan to use the same setup to also monitor (and log) rotor RPM (from which alternator o/p can be deduced) vs wind speed. I calibrated the anenometer but attaching it to a stick which I held out the car window and using a GPS for accurate determination of vehicle speed.

I plan to attach the unit to the turbine head with a 10' length of EMT sticking out horizontally a couple of feet m/l behind the rotor. Hopefully it will be sufficiently clear of rotor turbulence (rotor is 13' dia) to give a good wind speed reading at the rotor elevation (and will not flex enough to contact the rotor!).

Mark Parsons

10th March 2008, 11:06

Hi Laurie,

Sounds like your on the right track with winding your own torsion springs. McMaster is never the least expensive, but if you want it now, they got it.

The 3-jaw lathe chuck appears to be ideal for 3 blade pitch control. Remove the jaws and use the housing with ring gear and chuck key gears intact. Sweet :)

Annemometer data will help significantly with your experimenting. I need to get or build me one.

Regards,
Mark

Rob Beckers

12th March 2008, 08:09

Laurie Forbes

14th March 2008, 19:44

Laurie, how does the 'design your own springs' work? Is the software intended for fabricating them yourself (how?), or do you then send it off to a manufacturer? Could you give a little more info on the software and the process of realizing the springs. Thanks!

-RoB-

The software only designs it for you. You then send the parameters off or build it yourself I guess. It takes the form of an Excel spreadsheet - you enter the descriptive parameters of the spring you want and it calculates the wire size and number of coil turns required etc. It's available in trial version from:

http://www.mitcalc.com/index.htm

along with a bunch of other useful engineering calculation programs.