 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Windmills : Mini, Dual Rotor, Axial Flux Alternator
Page 1 of 2   |
|||||
| Author | Message | ||||
| BlueSmoke Newbie  Joined: 02/05/2013 Location: United StatesPosts: 25 |
Hi All, Thanks to admin. for the registration, it's good to be aboard. Greetings from "Up Over" (Washington state, US), I'm working on a fun little project that I thought I'd share. I'll post a few details and some wiring diagrams in case anyone is interested. As the title suggests, this project involves building a "scaled down version" of the old standard, dual rotor, 3 phase axial flux alternator but instead of using 3-4" diameter x 1" thick neodymium, hand smashing, flux monsters I am using some tiny 3/4" x 1/4" thick, disc neos that were on sale. No great expectations using such small magnets; it's more a labor of love and the process of going through the steps of creating my first axial flux alternator, albeit a smaller version than most. (I think of this one being somewhere in the range between your standard axial flux build and a stepper motor). :D My goal is to use this and a 5 watt solar panel to trickle charge a 12V gel cell (I'm thinking 9Ah) to power some LED's for deck perimeter lighting. As of now, I'm not worried about a controller but if miracles happen maybe I'll need one. Whether I can achieve 14V with this little alternator is the $100000 question. Sorry I don't have any pictures yet but I'll try to capture some of the process for later posting. On this little project I'm pretty much designing as I go. I know the basics of building an axial flux so I'm just using stuff I have laying around the shop, where possible, to fabricate. I don't weld so pretty much everything I do is bolt together. Each rotor will have 20 magnets (20 polls). Rotor backing plates are go-kart sprockets (60 tooth) so are about 7.25" dia. This places the 20 neos centered on a ~6.5" circle on each rotor. Magnets epoxied to plate and resin ring cast around them to hold them in place. Stator will have 15 coils (5 coils per each of 3 phases). Coils will be trapezoidal with an average I.D. of 3/4" and an O.D. of 1-1/2" (width and height), 20AWG @ 70 turns each. (That's the max. I have room for). Coils will be cast in resin forming the stator (about 3/8" thick) which will be sandwiched between the two rotors. Stator will be approx. 9.25" diameter. I've started winding the coils (10 of 15 so far) and, so far, have one rotor with magnets epoxied ... still a long way to go. Hub and rotors will be held together using 6 bolts on a 3.25" bolt circle and the stator will be mounted to a piece of 1/4" thick plate steel using 3ea. threaded rods. This will insure I can adjust the stator location relative to rotors and distance between rotors to control the air gap. Air gap will be as small as possible, probably about 1mm. Rotors and hub will rotate with a 5/8" diameter shaft mounted to the plate steel with 2 bearings housed in pipe (welded) to the plate. Blades will be some Air-X type blades (60" diameter windswept area). I'll try 3 blades first but have 3 more if I want to add them later to capture low winds. Yaw bearing will be a piece of greased pipe over pipe pole. Tail will be mounted with hinge, angled to allow furl. Haven't tested a coil yet (only one rotor so far) but as I get things assembled I'll test one coil and extrapolate to see if I possibly have a chance to hit 14V or if I'm shootin' blanks. Below are diagrams of my stator wiring. (Star, Star Wiring Connections, Delta and Delta Wiring Connections) I will have 6 connection points, each an end of 5 coils in series, per phase. This will allow me to set up as Star or Delta configuration and test each phase for resistance, voltage amps etc. 
Star - Phase1, Phase2 and Phase3 w/Connection Points 
Star Wiring Connections 
Delta - Phase1, Phase2 and Phase3 w/Connection Points 
Delta Wiring Connections I guess if there were a question here, it would be: Do you think I have a chance of producing enough voltage to charge a 12V battery with this setup?(100-150 RPM) (Only testing will tell but I'm curious what others might guess). Thanks for your interest. BlueSmoke |
||||
fillm Guru  Joined: 10/02/2007 Location: AustraliaPosts: 730 |
Hi Blue, I don't see why it wont work , I would not worry about the Trapezoid coils with round magnets ,and just use round former's , you are adding extra coil resistance for very little gain . Best advice is to get it up to a stage where you can test coils , and all will be revealed then , I would drop the star / delta switch , star will give you the best performance for the least resistance . PhillM ...Oz Wind Engineering..Wind Turbine Kits 500W - 5000W ~ F&P Dual Kits ~ GOE222Blades- Voltage Control Parts ------- Tower kits |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Hi Fillm, Since I've got 10 of 15 coils wound already I'll do the others the same but as you say, testing will reveal what she's capable of. If for any reason I don't get adequate results during testing and have to wind others, I'll make them round as you suggest. Wish me luck, I got my neo's epoxied to my second rotor so now it's just a matter of getting the axle and bearings set up to do some testing. If all pans out then I'll resin around the magnets on the rotors and complete the stator wiring then get it cast as well. I don't plan on a switch for star vs delta, just wanted the connections available, I'll wire it as star, no doubt. Since I may go with 3 or 6 blades, I'll need to figure out how to make my tail angles adjustable (they'll likely need to be different depending on which route I go. Always nice to keep the options open. I'll keep you all posted on progress. Thanks, BS |
||||
Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333 |
As Phill said you are really going about this build backwards. You need to setup the magnet rotors and test a coil before just winding wire to a turn count you pulled out of ..... Errr, thin air. My guess is 70 turns will give over 40 volts (volts are easy to make) at 100-150 rpm, but its the Amp's you need to produce at the required voltage. Volts come from the number of turns per coil, Amps come from the wire size used, so i would expect you could use around 3 times less turns per coil and 3 times the wire strands per turn, to give a lower voltage output to suit your 12v battery and 3 times the current. Again that is just a guess and hence the need to test a coil to see, the idea is to get the turn count right to suit the mill voltage intended for a given cutin speed, then fill the room allowed for each coil with as much copper as you can fit in. 70 turns might give you a low cutin speed of 30-40 rpm for a 12v battery and i doubt the mill would ever get out of stall to actually make useful power with Air-X blades with such a low cutin speed. Pete. PS... BS stands for a different meaning in my neck of the woods, (hint... Bull S--t) Sometimes it just works |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Thanks for the info. Pete. I'm working on the rotors, bearing mounting now, as you both suggest, I need to have that set up and get some testing done. If I can actually go with less turns (that would be great news). I do have some larger dia. wire I can wind another test coil with, for comparison. I wish I had an o-scope, all I've got is a multi-meter but hopefully it will do. I'll need to read up on how to determine (calculate) my "potential" VDC output from testing on one coil but if I remember correctly ... (Please correct me if I'm wrong) For testing purposes, using Star wiring, to produce 14VDC, I need to figure the following: (I'll lose a little through rectification but hopefully this should get me in the ball park). 14VAC / 1.73 = 8.09 VAC per phase of EMF needed 8.09VAC / 5 coils per phase = 1.62 VAC per coil needed at desired RPM I'll get the rotors mounted and report back on any coil testing done. Thanks for the help. BlueSmoke PS ... (pun was intended, but) ... Yes, I would imagine being Downwind, BS would not be favorable. 
|
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
This jives pretty well with my own crystal ball, I don't think getting enough voltage will be a problem for you. Connecting the windings in closed loop delta will certainly work, but it is less than ideal. One way to run in delta without the problems of harmonic circulating currents would be to bring out all six wires without any interconnections on the ac side. Then connect each winding to it's own individual bridge rectifier. You then parallel up the three bridge rectifier outputs on the dc side. This works exactly like the traditional delta connection, and it gives the same output voltage and current. Only difference is, each winding works quite independently on the ac side, and any waveform distortion no longer matters. Cheers, Tony. |
||||
yahoo2 Guru Joined: 05/04/2011 Location: AustraliaPosts: 1166 |
Welcome Bluesmoke, Not sure the air-x blade will bring you much joy. I know they are tempting but they are designed to be a high speed blade. I don't want to bore you to tears so have a look at the video on this page and look at the red line on the graph underneath. It will give you a feel of the rpm that those blades need to start to make some torque (the turbine would be barely making power in the video). silent power page according to the video this was 11 to 15 knots, that converts to 5.6 to 7.7 metres per second on the graph. That is not very fast at all for these blades. specs for the air-x marine with the fancy gold plated blades. Rated Power Output: 400 Watt @ 26.24 knots (13.5 m/s) Start up Wind Speed: 4.27 Knots (2.2 m/s) Start up Charging: 5.83 Knots (3 m/s) Rotor Diamter (No. of Blades): 1.15m (3) Rotation Speed: 480 - 1420 rpm 
P.s. That swooshing noise that you can hear at the start of the video is the blades flexing and stalling in the light wind and turbulence. I'm confused, no wait... maybe I'm not... |
||||
| Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333 |
Here is a example i done for a different post some time back and should help to answer some of your questions. The example was for a 24v system so you will need to adjust it to suit your needs. ****************************************************** With regards to coils and voltage, lets work this backwards from 28 volts DC 28 / 1.4 = 20 Volts AC (divide DC volts by 1.4 to get AC volts) 20 / 1.7 = 11.76 Volts AC per phase (divide total AC volts by 1.7 to get AC volts per phase) 11.76 / 3 = 3.92 Volts AC per coil (as there is 3 coils per phase divide by 3 to get AC volts per coil) So your target voltage you will need for each coil will be 3.92 Volts AC at the RPM you want cutin to start at If for example we wound a test coil of 10 turns and at 60 rpm we got 0.5 volt AC, than at 120 rpm we would expect 1 volts AC from the 10 turn. Now if we wanted cutin to be 150 rpm then we need to increase our turns to give 3.92 volts AC at 150 rpm. 150 / 60 = 2.5 (divide target rpm by test rpm) 0.5 x 2.5 = 1.25 (times test voltage by rpm compensation value to get VAC at target rpm for test coil) 3.92 /1.25 = 3.136 (divide target VAC by test coil voltage for target rpm) (times the number of turns used in the test coil by above factor) 10 turns x 3.136 = 31.36 turns (or 31 turns per coil) ********************************************************** Option #2 Some find it much easier to add a bridge rectifier and a 1000uf cap across the DC output side and just work with straight DC readings of the test coil. Its easy to spin the rotor at 60 rpm by hand for all tests (1 rev per second) This also allows for diode loss of the rectifier as well. Then just calculate DC volts per turn and volts per rpm to workout the number of turns required for the cutin speed intended. Pete. Sometimes it just works |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Interesting information Warpspeed I'll bear this in mind, I can quite easily place connection points around the perimeter of my stator without much (if any) additional work. Perhaps I'll keep the connections available during casting to keep my options open. I completed the basic assembly of the rotor assemblies today, have a small issue with concentricity of the rotor vs. mounting plate for my blades and need to work that out before proceeding. It appears that the holes I drilled in the blade mounting plate are not exactly aligned with the center point of the bolt circles of my rotors. I have a few options open to fixing this but will likely need to make a new blade mounting plate. Unless all is concentric, it will introduce all sorts of unacceptable problems and is a relatively easy fix. Fortunately, the rotors (and their magnets) are matched, aligned and the problem lies only in the blade mounting plate. It's always something (as Gilda Radner would say) ... all a balancing act, the rotors, the mounting plate, the tail, the right wire diameter, the number of turns ... all a balancing act. No one ever said it was easy ... but I'll get there, just not today and perhaps not tomorrow ... but I'll get there and the decks will have their perimeter lighting ... one way or another. Didn't get a coil tested yet due to the concentricity issue but should get that figured out soon. I held a coil in hand between the rotors as they are and saw about 1 VAC but due to circumstances don't think it any reliable indication. I think my biggest problem testing one coil will be my multimeter. It only has a 0-200VAC range and I doubt that will be calibrated close enough to see what's going on. I could feed the coil through a bridge rectifier and read the DC, perhaps that might give me a better indication ... but for now, I'll work out the rotor problems, make a suitable testing stator (single coil) and proceed from there. BlueSmoke out |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Hi Yahoo2, I have a few options for blades, the Air-x blades I mentioned are actually a version that are 30" windswept diameter version (I think actual Air-X are 22?) that are made by others (not sure who) but are carbon fiber blades "similar to" Air-X type. Other than that I have a 3 or 6 blade set of fiberglass blades that will also mount up to my existing rotors ... so at least I have a few options available using both 3 or 6 blades of either type. Your information is worth noting and I appreciate the information, the blades I am trying first were a gift and I had them collecting dust in the shop so I thought I'd give them a try to see how they perform. The other blades may be better or not (purchased off ebay) but between the two sets I at least can try some options. Thanks for the input. BlueSmoke |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Hi again Downwind, Great information, thank you! That's exactly what will help me get this all figured out. Thanks for taking the time to look up that post and bring it back to life here!! As previously noted, I've got some concentriciy issues on the rotor portion of this but as soon as I get that straight I'll be setting up a stable test coil and be able to get some good predictions via your help here. I'll keep ya posted. BlueSmoke |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Yahoo2, Correction, that not 30" diameter windswept ... it's 30" radius or 60" windswept. BlueSmoke |
||||
| fillm Guru Joined: 10/02/2007 Location: AustraliaPosts: 730 |
Go With option 2 for coil testing , it is heaps easier and gives the DC out with the diode loss calculated , from experience with my Axials also subtract 10% for wiring losses to get what you want at the batteries. If you are winding all your coils without testing you are heading for trouble , as caculating your flux density is open ended to how far the magnets are apart, Heres a Link to my last Ax build . I would be looking at min 130~150 Rpm cut in on those blades , any lower and you will stall them. PhillM ...Oz Wind Engineering..Wind Turbine Kits 500W - 5000W ~ F&P Dual Kits ~ GOE222Blades- Voltage Control Parts ------- Tower kits |
||||
| yahoo2 Guru Joined: 05/04/2011 Location: AustraliaPosts: 1166 |
Perhaps I should have made it clearer, I am NOT recommending those particular blades, in my opinion the are just the same blade, painted blue with a turbulence stringer added to cut the noise down. I am just trying to give you some idea as to why they are next to useless under 350RPM The longer ones will certainly be better at lower revs, they might start to work at 200rpm, I don't know, if you already have them on hand and they are easy to fit, there is nothing to lose by having a go. 
I don't think it will matter at cutin revs if you have 3 or 6 blades. If they are still stalling at the point that the generator starts to produce load they will struggle. If you listen carefully to the sound that the blades make in the first half of the video you can clearly hear them go from idling for the first 3 seconds, to stalling for the next 22 seconds then they start to fly (and produce power) The buffeting of the wind turbulence as it comes across the deck of the boat is making it even harder for them to get running. At the videos 30 second mark the generator load would still be less than 4 watts at 450+ rpm. It will be interesting to see what volts / revs you come up with for your machine. cheers Yahoo I'm confused, no wait... maybe I'm not... |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
WOW! That's some very impressive work that you've done there FillM. I'll be reviewing your link(s) to see what more I can learn. This is obviously light years beyond what I'm attempting but, none the less, I'll try to reap some benefits from your, well documented, precision build of the AX320-1MF. As noted, I can quite easily change the blades if needed but I don't really see 130-150 RPM being a real problem here, so will stay with the current blades for now. We do have lots of tall trees around us but we live in a windy gorge area so most wind here is, on an average, gusty, changes direction quite often and is fairly low speed but well disbursed throughout the year. Unfortunately, putting up a tower above the treetops (150+ feet) to harness more steady wind is not a possibility on my rather small lot. The hope is that this little mill can harvest some of the multi-directional, lower wind speeds to trickle charge one battery for LED lighting (I'll also have a small solar panel to supplement). I don't mind if this one needs to furl in higher winds as they do occasionally arrive. I've got another 500W wind turbine up and running coupled with a 30W solar panel that I currently use to keep two 6V, 220Ah AGM batteries topped off for (12VDC) stand-by UPS and/ or running my laptop. (It's a Windynation Windtura 500). Fortunately with the current build, I have the liberty to tailor my battery size (voltage, storage capacity or Ah) and potential load since my goal consists only of powering some (low voltage fairly low draw) LED's. More is always better, in terms of output, of course, (both in voltage and amps) but in this particular case I have some latitude and don't think I should expect too much from the limited amount of flux that these tiny neos can muster. With that said, I don't want this build to be a total waste of time and effort either, so I'll work with what I have and what I can learn here to make the best of this humble project. (Perhaps these 40 little neos have more potential than I give them credit for). I do very much appreciate everyone's input here and I Thank You All for that. It helps in lots of ways, sharing the knowledge helps with the learning curve, your interest helps keep me motivated and sharing the fun makes it all quite worthwhile. Best to you all, BlueSmoke |
||||
| govertical Guru Joined: 11/12/2008 Location: United StatesPosts: 383 |
Hi, does having the magnets so tightly packed together cause a problem? Would it decrease the lines of flux passing through the coils? just because your a GURU or forum administer does not mean your always correct :) |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Hello GV, My thought here is that more rapid polarity changes per revolution are a good thing here, (especially at lower RPMs which are generally my target with this mini mill). Don't really know if flux paths will interfere with each other due to relatively close proximity, as in this configuration, but these are pretty small neo's so I would think it will not be a big problem. Currently getting the rotors dialed-in and testing a few different coils is my goal. I made a new blade hub plate today which seems to align well (concentrically) with the rotors. Now it's just a matter of getting a stable coil testing arm in place to see if I can reach my goals (re: 14VDC with usable amperage to trickle charge a 12V battery). The question that you pose is not really something that I could ever test but I do know that others have made 20 pole rotors with a degree of success. I think having the coils well matched to the magnet array and taking the best advantage of the space available is what is important. Not really sure what you mean by "is the extra effort needed?" ... which extra effort are you referring to? BlueSmoke PS: I rather like working with this smaller scale of wind turbine but as the size reduces, so too do tolerances, they become tighter so it's a bit more difficult without precision, machine tooling available. I envy those of you who have access to machine tools, always enjoy seeing builds where the job is done right and with precision. My shop is pretty basic but it's fun to use what tools and knowledge I've accumulated over the years. Today is Cinco de Mayo so we're enjoying the new deck, having mesquite, slow cooked chicken breasts with dejon mustard on BBQ, some fresh vegies from the garden and good tunes ... life is good! Enjoy, Blue |
||||
| govertical Guru Joined: 11/12/2008 Location: United StatesPosts: 383 |
Hi, if the distance between the magnets on the same rotor is less than the air gap between magnet rotors, is there a decrease of lines of flux passing through the coils? just because your a GURU or forum administer does not mean your always correct :) |
||||
| BlueSmoke Newbie Joined: 02/05/2013 Location: United StatesPosts: 25 |
Hi GV, Ok, I see what you are saying ... what type of data or literature do you have supporting this relationship of distances? I've never read anything published regarding this particular relationship. I would have to guess that it would depend on many factors (rather than just the two dimensions). (e.g. magnet size, overall shape, cross section profile, thickness, strength, backing plates used ... etc.). Since the magnets are polarized top & bottom (N&S) rather than side faces it would seem to me that these faces (N&S) would be dominant and interact more closely with (S&N) on the adjacent rotor as long as distance A and B do not differ a great deal. (On my setup, distances A & B look to be within a mm or two of each other) so I won't be losing any sleep over it here, but ... ... if the relationship of distances A & B is of great concern then I would wonder why Pie shaped magnets that actually touch side to side (where A = 0) are often used on axial flux rotors with great success. It would be interesting to know more but with this humble build it's really a mute point for me since I already have two rotors mounted with magnets epoxied to their backing plates. I'll certainly be keeping my air gap as small as possible, however. BlueSmoke |
||||
| govertical Guru Joined: 11/12/2008 Location: United StatesPosts: 383 |
http://www.femm.info/wiki/HomePage Finite Element Method Magnetics allows you run simulations by spacing magnets and changing air gap. On my own project I increased distance "A" to be greater than distance "B" and noticed a improvement. I understand it is mute point now that you have fabricated your rotors, but I was curious if anyone had a similar experience. I look forward to your results. just because your a GURU or forum administer does not mean your always correct :) |
||||
| Page 1 of 2 |
|||||
 Print this page |
