500 watt eBay wind turbine!

[makenzie71]

FedEx brought me the first two to play with today!



This is the specific model I'm working with first.

First thing I noticed is that all the blades have these marks on them.  I think they're possibly injection points, not entirely sure.  Each blade has them, in the same location, and in the same pattern.





About half the holes on the hub needed to be deburred.



The blades also needed to be deburred in order to fit in the hub properly.



Money shot!



The stator is wound with 22g wire.



There are 36 individual coils.



Output wire is bigger than 12g, smaller than 10g...11g?



There are twelve magnets measuring roughly 18x12x4.





The outside lamination is already trying to peel away :/



Idea of the air gap...the rotor is stuck to the stator on the opposite side so it's roughly half the gap seen here.  The single bearing construction of the generator made it not possible for me to hold the rotor center and get a real measurement.



Also to note...something I don't like is that the only bearing is in the nose.  Rotor on one side, prop on the other.  The hearing is roughly 25mm so it's not an insubstantial load ability, just should be two bearings.

Another thing I do not like...there is no groove for the o-ring sealing the bearing carrier to the rest of the body.  The o-ring also sides on top of the bolt sockets...the bolts just kind of push the ring out of their way.  I don't think it's going to be an effective seal.



That's really all for now, I will follow up when I can actually get this thing on a lathe or something.  I attempted to run it on my drill press and it never once occurred to me that my drill press is only 1hp.  Running this thing through a Honda CBR954RR regulator/rectifier (three phase rectifier rated for 400w@14.3vdc) connected to a battery (showing 12.4vdc when disconnected) I was charging at 15.3vdc (a little high) and 4amps at 122rpm.  I set the press up to run 400rpm and it kicked the amp output up to 8.5amps, still 15.3~15.7vdc, but I never cleared 200rpm...my poor harbor freight press just couldn't do it.  My neighbor has a huge vertical mill...I might pop over there this weekend and see if he'll be okay with letting me chuck it up.

As you can tell by the boxes, the other mill is much larger...I'll get a basic startup like this one for that model tomorrow.

EDIT: fixed image tags

[SparWeb]

Thanks!

The wiring looks messy.  Usually the wires in the coils should be stuck together.  That's because there's a force between the magnet and the field in the coil when current flows.  If the wire is free from the bundle it will "buzz" around.  Has the stator been "dipped" in any kind of laquer or varnish to hold the wires together?

You should look for a plastic vernier caliper to use with those magnets.  I have magnetized too many of my tools, and hopefully won't accidentally do this to my calipers or mic's. 

[DanG]

Ridges on blades are artifacts from rushed CNC 3D mold making - but wait - that's not a flaw that's a feature! ( https://en.wikipedia.org/wiki/Tubercle_effect (way old Fieldlines topic))

Having a jig and liner paper to avoid rotor/mags from contacting laminations would be good, getting something like 2L bottle plastic in/out for insertion & removal would be a whole trial & error battle of will power.

From my 750VDC work days I'm partial to Glyptal Red Enamel 1201 for protecting coils etc. but its expensive ($45 qt (last forever) & $40 aerosol) - on coating the rotor and magnets I'd recommend POR-15, it's thin, penetrates, catalyzes once the carrier naptha has evaporated, and so freaking hard it gives pause to an angle grinder wheel. Yeah, also a little spendy, make sure everything is perfectly degreased and even a little 400 grit wet/dry paper scuffed.

Anyhow - China has a huge market for indigenous peoples and that looks like a perfectly respectable Mongolian Yurt power source - cheap but not really disposable, able to work X days a month for a bunch of seasons, lightweight....  At $100 it almost makes me want to nab one myself   

[makenzie71]

Thanks!

The wiring looks messy.  Usually the wires in the coils should be stuck together.  That's because there's a force between the magnet and the field in the coil when current flows.  If the wire is free from the bundle it will "buzz" around.  Has the stator been "dipped" in any kind of laquer or varnish to hold the wires together?

You should look for a plastic vernier caliper to use with those magnets.  I have magnetized too many of my tools, and hopefully won't accidentally do this to my calipers or mic's.

I don't believe the coils were dipped, if they were the solution was very thin.

I have an aluminum caliper...it didn't even cross my mind until the rotor snatched the thing out of my hand.  I don't get it out much because aluminum but I've gone ahead and set it out next to all my wind stuff.  I can say that these magnets are very robust.  They have a lot more pull than my N42 half inch cubes.

[makenzie71]

Ridges on blades are artifacts from rushed CNC 3D mold making - but wait - that's not a flaw that's a feature! ( https://en.wikipedia.org/wiki/Tubercle_effect (way old Fieldlines topic))

Trailing edge, though no tubercle effect for me

Having a jig and liner paper to avoid rotor/mags from contacting laminations would be good, getting something like 2L bottle plastic in/out for insertion & removal would be a whole trial & error battle of will power.

I'll keep that in mind for the next one.  It wasn't really too big a deal to get the rotor past the coils but the next one has a bigger rotorand bigger coils (assuming).  Might be more prudent to guard things on that one.

[quote)From my 750VDC work days I'm partial to Glyptal Red Enamel 1201 for protecting coils etc. but its expensive ($45 qt (last forever) & $40 aerosol) - on coating the rotor and magnets I'd recommend POR-15, it's thin, penetrates, catalyzes once the carrier naptha has evaporated, and so freaking hard it gives pause to an angle grinder wheel. Yeah, also a little spendy, make sure everything is perfectly degreased and even a little 400 grit wet/dry paper scuffed.

Anyhow - China has a huge market for indigenous peoples and that looks like a perfectly respectable Mongolian Yurt power source - cheap but not really disposable, able to work X days a month for a bunch of seasons, lightweight....  At $100 it almost makes me want to nab one myself   
[/quote]

Big fan of POR15....it's in a lot of bikes' gas tanks and radiators.  Great stuff.

This particular windmill is the one that wentfor best offer.  I got it for $90 shipped peanuts to play with and sinceit probably won't quite produce what they claim it will i'll likely end up with a partial refund...all the chinese sellers ship out of american warehouses but operate out of china.  They don't want a return because they can't actually receive.  They want to settle.  The other turbine, for example, was suppose to come with a charge controller and it came with a bent tail vane...they don't want a return and they can't send me a controller without taking it out of another turbine's box...so they're giving me $125 back to buy another controller from someone else.  Had i paid full price for that one it would have been a $325 turbine.

[Adriaan Kragten]

You said that the bearing is only at one side of the armature. Is only one bearing used? In this case there will be a lot of instability especially if the bearing gets some clearance. Two bearings together will be a lot better. The 12 magnets are curved but have a rectangular shape. This means that you will get 36 preference positions in combination with a stator with 36 slots. Have you measured the peak on the sticking torque from stand still position?

[makenzie71]

You said that the bearing is only at one side of the armature. Is only one bearing used? In this case there will be a lot of instability especially if the bearing gets some clearance. Two bearings together will be a lot better. The 12 magnets are curved but have a rectangular shape. This means that you will get 36 preference positions in combination with a stator with 36 slots. Have you measured the peak on the sticking torque from stand still position?

Yes just one bearing.  I didn't get a to measure the depth but it's at least 30mm and the shaft is not very long.  I think that if the bearing is of sufficient quality it would be plenty of support.  The cap that supports it is very heavy and thickly cast.  My concern is that "high quality" and "chinese roller bearings" aren't typically synonymous.  We'll see.  If the bearing fails it'll be a pretty short, easy thing to replace with a really good one.  The way this turbine is made, though, there is no way to put another bearing at the other end.

[makenzie71]

whoops missed the torque question...from a stand still, sitting on my table top in the kitchen, my scale doesn't even register a pull pulling at blade tips.  There's very little cogging.  If I short out all three wires it registers as less than a pound pulling from the blade tips.  It'll have a pretty easy start up.

[makenzie71]

Alright I haven't really been able to do much more with this...my drill press wasn't able to handle the load for rpm tests and I'm still waiting on a charge controller to do real life tests.  The next one is larger and I'll be able to get it on a lathe to do the rpm tests.

This one has been out in the wind a bit now, though, and I can tell you that I think these particular models have too much swept area and not enough tail.  This one is 48" of swept area.  It does okay in lower wind but the faster the wind gets the less stable it is.  So long as the wind is straight on and blowing hard it stays pointed into it, but the second it lets off or if the wind shifts slightly the thing spins out of it and tries to put the blades down stream.

I got a video of it I might upload.  With only three blades up the thing does really get moving.  I'm really eager to get my controller in so I can see what it's actually producing.  I might go wire it back up to my CBR rectifier and just string a bunch of 12v lights to it to see...

[SparWeb]

Mackenzie,
Welcome to the windwright's club.  Your addiction also begins now!

Counseling sessions are held in the church basement on Saturdays.

[Adriaan Kragten]

Alright I haven't really been able to do much more with this...my drill press wasn't able to handle the load for rpm tests and I'm still waiting on a charge controller to do real life tests.  The next one is larger and I'll be able to get it on a lathe to do the rpm tests.

This one has been out in the wind a bit now, though, and I can tell you that I think these particular models have too much swept area and not enough tail.  This one is 48" of swept area.  It does okay in lower wind but the faster the wind gets the less stable it is.  So long as the wind is straight on and blowing hard it stays pointed into it, but the second it lets off or if the wind shifts slightly the thing spins out of it and tries to put the blades down stream.

I got a video of it I might upload.  With only three blades up the thing does really get moving.  I'm really eager to get my controller in so I can see what it's actually producing.  I might go wire it back up to my CBR rectifier and just string a bunch of 12v lights to it to see...

Once very long ago I have measured the wind speed behind the rotor for a rotor which was tested in the open wind tunnel of the University of Delft. We have found that the wind speed at the center of the rotor was very low and that the wind speed increases as you come more to the outside of the rotor. If a small vane blade is positioned just behind the center of the rotor it is just at the place of the lowest wind speed and therefore it isn't very effective. But it is strange that a fast spinning rotor turns out of the wind if there is some shift of the wind direction because a fast running rotor turns itself in the wind because of the self orientating moment. May be turning out of the wind happens only at low rotational speeds.

[makenzie71]

So i ended up going out and connecting the thing to my CBR R/R and ran it to a 175w light bar for load.  It's like the brakes are on.  We have a good 20-28mph wind going and voltage never excedes 9v.  Marginal amps.  The light bar is lit dimly.

Should i have it wired to a battery first, and then the load connected to the battery?  Doesn't seem like that's how i've seen test setups done but ult8matelyitks how this thing will be arranged anyway.

[DamonHD]

Incandescent lamps have a very low cold resistance.  Not a realistic test, I suggest.

Rgds

Damon

[makenzie71]

i'm wondering if this shunt rectifier/regulator eats up too much current by itself.  I know there's some drop but doesn't seem like a whole lot...the stock stator on the 954 only makes about 200-250 watts, it doesn't have a lot to lose.  But the turbine goes under load just hooking it up the rectifier.  Hooking the light and battery to the rectifier doesn't seem to load the turbine up any more, it just drops the voltage from 19vdc to 9vdc.

And i went and hooked it up to a battery just to see if the battery added some cushion...no change.

[makenzie71]

Vendor says “no no must have six blades to perform”

So i guess i'll toss the other three blades on it just to see.

[makenzie71]

I popped the three other blades on...one of them is all caddywompus...so there's that.  The vendor asked for a video of the setup so I did this.  The thing is wired up to 25ft of 10g wire, to my CBR R/R (shunt style regulator), which then goes to my battery, and my LED light bar is hooked to the battery.  We had gusts over 30mph today and I never saw charging voltage.  A few days ago we had gusts upward of 60mph and got 12~14vdc but never more that 10amps.  I've been running three blades on it, but the vendor says this model has to have six...whatever.  Just waiting to see what suggestions they have.

I can say with six blades its start up is much lower and it holds its rpm better...but still not getting 12vdc.  Just went out and checked and it's holding at 9vdc in a very nice, steady 12mph wind....probably going to kill my poor battery.

[SparWeb]

And i went and hooked it up to a battery just to see if the battery added some cushion...no change.

Did you disconnect the lights while you did that?

I think a light bar would act like a dead short when it's too cold.

[makenzie71]

Battery first, then the light.

[makenzie71]

Just connecting to the rectifier/regulator caused it to drag, connecting it to the battery and/or the light really didn't seem to add any more load to the generator.  It's not acting as a dead short, though.  I shorted all the AC connections to brake it when I climbed up to mess with the blades and there is a significant difference in the dead short load on the rotor and the "charging" load I have on it.

[makenzie71]

Went out and checked...with six blades it's even going now.  We usually have no wind in the morning....it's like 2-5mph.  Sticking turning, but still only making 9.5vdc and vrtuslly no amps.

[DanG]

Sounds like you need a point-to-point breakout of coil ins & outs...

[makenzie71]

Sounds like you need a point-to-point breakout of coil ins & outs...

Can you explain this?

Right now the vendor is probably going to refund and write the turbine off.  This is a pretty common practice when you're dealing with Chinese sellers...they ship from US warehouses but aren't equipped to take anything back.  When you buy a Chinese product if it doesn't perform as described, if you open a claim, they will lose almost by default...and, if you have a claim filed against your account enough times eBay will shut you down.  The sellers do not want negative feedback and they do not want a claim and they can't usually accept a return so they're VERY motivated to settle with you.

This is why I'm not generally worried about buying something risky like a Chinese turbine with bloated specs.  It's either going to perform the way they describe or they're going to lose money on the deal.

I'd be happy enough if this guy would maybe exchange this turbine for another...maybe this one is defective in some fashion.  I mean mean doing poorly is one thing...I didn't expect it to do what the ad said it would do...but this one just "isn't doing".  That suggests, to me, that something up there isn't working quite right.  I doubt the vendor will exchange, though, and will probably just refund and move on.  At that point I'll be up for suggestions for how to make this one better...like maybe ripping all thirty-six 22g hair thin wire out and sticking in 18 thickly wound 14g coils and seeing what happens

[JW]

Quote from: DanG on Today at 01:30:59 PM
Sounds like you need a point-to-point breakout of coil ins & outs...

Can you explain this?

Here are 2 topics that may or may not help you.

https://www.fieldlines.com/index.php?topic=143265.0

https://www.fieldlines.com/index.php?topic=132443.0

This is why I decided to preserve all the database 14 yo topics etc.

JW

[makenzie71]

Here are 2 topics that may or may not help you.

https://www.fieldlines.com/index.php?topic=143265.0

https://www.fieldlines.com/index.php?topic=132443.0

This is why I decided to preserve all the database 14 yo topics etc.

JW

Thanks!  That first one seems relevant...I'm waiting on a final decision from the vendor before I move forward, but I am looking forward to digging around in the coils now that I'm not worried about messing it up.

[makenzie71]

Been up for a while with no real change...it's a little too cold out and the pole is a little too wobbly for me to climbing up there to get it :p

I did try a different rectifier today...I got my nifty Chinese hybrid controller in today (I'll make a new thread about this thing) and hooked it up in place of the CBR954RR rectifier/regulator.  I thought since maybe the CBR engine engine is suppose to be running at may more speed and maybe since there's going to be more loss because it's also trying to regulate voltage...that maybe it wasn't very well suited to the task.  Throw the eBay controller on it and...it's at least a little better.  But not much.





18~24mph wind, fyi.

The battery was only at 8volts when I connected it, though, and I'm reading that this might contribute to the crazy low output reading though I can't fathom how.  Still learning that stuff.  It is up to a whopping 11.8vdc now, though.

[makenzie71]

Okay so the seller opted for refund and keep...which means I keep the turbine and got a refund.

So the fun part starts now: How to actually get 500w out of this thing?  I don't think it's actually going to be a matter of just spinning it faster...this thing has already been spun way the hell up there in 40~50mph winds and still was only producing a handful of watts.  Going to have to get a closer look at the coils and really explore them as suggested above.  Going to start by finding that wonky blade, discarding it, and going back to just three blades.  Didn't start up as easy with three blades but I think it kept rpm up a lot better.

[kitestrings]

Mak,

I haven't followed this real closely, but have you tried a different battery?  Eight volts sounds like a boat anchor to me, if it has been in that state very long.  I assume this is a 12V nominal system.  Is that correct?

[makenzie71]

Mak,

I haven't followed this real closely, but have you tried a different battery?  Eight volts sounds like a boat anchor to me, if it has been in that state very long.  I assume this is a 12V nominal system.  Is that correct?

It is a 12v battery and maybe the battery is shot, but i wouldn't think so.  I'll swap it out for another one later this morning and see what kind of difference it makes.

Last night was pretty blowy...when i last checked it it was still holding at 11.8vdc.  This morn8ng, thouh, down to 10vdc.

[kitestrings]

Something doesn't sound right then.  Anything below nominal voltage at rest is pretty low.  Lead-acid will typically look like this:

[makenzie71]

Something doesn't sound right then.

Something isn't...the fun part is finding out what!

Swapped the batteries out about an hour ago.  The battery voltage showed 12.5v when I took out and still shows 12.5 now.  It's not very breezy at the moment so the turbine is only getting out 40~80rpm.  When I disconnect the battery the thing spools right on up in a hurry and unloaded voltage gets up to about 60vdc.  As soon as I hit connect it back to the battery terminal it's like the brakes are "kind of" on and goes back down to the 40~80rpm and battery voltage.

[DamonHD]

IMHO a good controller, eg MPPT, would match the load impedance to the available wind power and let the device spin up rather more I suspect, than when connected to the previous dead/ill battery boat anchor!

Rgds

Damon

[Adriaan Kragten]

Something doesn't sound right then.

Something isn't...the fun part is finding out what!

Swapped the batteries out about an hour ago.  The battery voltage showed 12.5v when I took out and still shows 12.5 now.  It's not very breezy at the moment so the turbine is only getting out 40~80rpm.  When I disconnect the battery the thing spools right on up in a hurry and unloaded voltage gets up to about 60vdc.  As soon as I hit connect it back to the battery terminal it's like the brakes are "kind of" on and goes back down to the 40~80rpm and battery voltage.

This is the behaviour of a wind turbine which is connected to a battery with a much too low nominal volt level. In my report KD 78 I have tested a generator for different constant voltages. Assume the winding is rectified in star. In figure 4 of KD 78 you see the Q-n curve for short-circuit in star. In Figure 8 of KD 78 you see the Q-n curves for 26 V star, for 52 V star and for 76 V star. You can see that all Q-n curves have about the same shape but that the curve moves to the right if the voltage is higher. Every wind turbine has a certain optimum parabola in the Q-n graph. Assume that the wind turbine is designed for 48 V battery charging. This means that the Q-n curve of the generator for 52 V star is lying close to the optimum parabola of the rotor. But if this wind turbine is connected to a 24 V battery, the Q-n curve of the generator shifts a lot to the left. The point of intersection of both curves will now lay at a very low rotational speed. So a too low battery voltage can almost work as a brake. I think that your wind turbine is designed for 24 V battery charging and that you use it in combination with a 12 V battery.

Another option might be that the generator has an option to connect the two layers of the winding in parallel for a 12 V battery and in series for a 24 V battery and that you didn't connect the layers in parallel. You should look at the generator terminal if something can be changed. Parallel-star connection requires a terminal with nine contacts if all wires coming from the coils are permanently connected to the terminal.

[makenzie71]

This is the behaviour of a wind turbine which is connected to a battery with a much too low nominal volt level. In my report KD 78 I have tested a generator for different constant voltages. Assume the winding is rectified in star. In figure 4 of KD 78 you see the Q-n curve for short-circuit in star. In Figure 8 of KD 78 you see the Q-n curves for 26 V star, for 52 V star and for 76 V star. You can see that all Q-n curves have about the same shape but that the curve moves to the right if the voltage is higher. Every wind turbine has a certain optimum parabola in the Q-n graph. Assume that the wind turbine is designed for 48 V battery charging. This means that the Q-n curve of the generator for 52 V star is lying close to the optimum parabola of the rotor. But if this wind turbine is connected to a 24 V battery, the Q-n curve of the generator shifts a lot to the left. The point of intersection of both curves will now lay at a very low rotational speed. So a too low battery voltage can almost work as a brake. I think that your wind turbine is designed for 24 V battery charging and that you use it in combination with a 12 V battery.

Another option might be that the generator has an option to connect the two layers of the winding in parallel for a 12 V battery and in series for a 24 V battery and that you didn't connect the layers in parallel. You should look at the generator terminal if something can be changed. Parallel-star connection requires a terminal with nine contacts if all wires coming from the coils are permanently connected to the terminal.

Quite possible.  The eBay listing said 48v, the box says 24, but the paperwork that came with it says 12~96 haha.  I can hook it up to a 24v setup and see what happens...nothing wrong with experimenting.  The voltage has been holding at 12v in the light breeze today but virtually no amps.