Previously he had asked me for photos/permission for a davit crane for a book that they are writing. Apparently there are not a lot of them on small wind turbines.

In his email back he had asked what size steel that I used for the telescoping tower stub. When I replied that I used 4-1/2" Sched 80 pipe he seemed pleased and remarked that using Sched 40 may have been a little on the light side.

A little confirmation every now and again is good to know that you made good decisions in your design...

With this milestone I am finishing out the build log portion of this and going into generation/data acquisition mode.

This information will be available on my Results page.

I suspected that something was not right as the output from the turbine appeared to be on the low side.

After posting my situation on the OtherPower discussion board I got my answer - I wound the coils incorrectly - they were wound with 80 turns of #12 wire (single strand) and they should have been would with 59 turns of #14 wire (two strands). I think that at one point in time in my project I suspected this - as looking in my garage I have two spools of #14 wire. I think in the rush to get it up and working last October this got overlooked.

At this point in time it appears that I have to remake my stator. Which means that the turbine has to come down to the ground in order to perform this task. Since the outer rotor has to be removed to remove the stator I am not willing to try to do this while it is on the tower, I want a more controlled and easier environment to perform this task.

And of course this means that the blades and tail have to come off as well... I may possibly be able to lash them to the sides of the tower to save bringing them all the way down and back up again - we'll see.

Over the past couple of days I have to admit that I have sat on a bench in my front yard just watching the blades of the wind turbine rotate, it is a bit of a hypnotic thing.

I have realized recently that I have a bit of a handicap with terminology, the difference between leg and phase is a bit obscure to me while it is perfectly clear to others that have considerably more wind turbine experience.

I have been in communication with numerous people from OtherPower.com discussing the facts at hand. Most people had stated that I probably have at least a 'measurement' problem. I don't dispute this fact! Some of the advice goers (from my perspective) are very close to electronic diagnostic techniques than I am. But it has been a slow road to learn that just stating the fact does not help in diagnosing/fixing the problem what I needed was some instruction on how to proceed.

Luckily I was personally contacted by Stew Corman about my dilemma, and he shared with me some testing procedures that I might employ to if nothing else figure out exactly what I have, and it's capabilities.

Within a couple of short emails I realized that I did not even know where or how to take voltage measurements! Stew summed it up in a couple of sentences: "The voltage between any two wires is no different than the house wiring, except the utility uses 2 phases 180 degrees apart and a ground. Your three phases are 120 degrees apart ..no ground You can measure the frequency between any two legs..."

My comment to Stew is that I don't believe in all my research that I have heard this. I had always been under the impression that each 'phase' voltage was measured from the center of the star (buried inside my stator) to each output, but clearly this is not the truth!

Well finally it was like the clouds and seas parted and I understood! On a 'typical' axial flux wind turbine you measure the voltage of one 'leg' between any two of the three phases that go to your rectifier! It has nothing to do with the center of the star connection inside the stator.

Immediately I knew that my wind turbine was achieving 'cut-in' speed as the voltage I measured across one leg was well above 48 volts (35v RMS * 1.4), so I know that my turbine is getting up to speed.

I am hoping that within the next week I can do some of the test procedures that Stew outlined. Once done I will report back.

Initially my plan was to duplicate the good work that has been achieved by the guys at OtherPower.com in the building and design of their 17' wind turbine. I did make a glaring error in the winding of the coils, but exactly at this time do not know the width and breadth of my error, my first task is to know exactly what it is that I have as well as the potential output possibilities that are present.

He suggested using multiple 230volt 4500watt water heater elements as my dump load (and for testing). With the final resistance of around 4-5 ohms. The elements should measure out to be about 11.7ohms each.

I went to my local electrical store and found four older elements that they let me have for $5/each. Getting them home I measured them out to be 13ohms each, so three of these in parallel will give me 4.33ohms.

One issue that came up is that these elements need to be in water, if heated in air then eventually they would crack.

So what I devised was to use a metal 5-gallon bucket, mounting these elements into the top cover so that they could be suspended in the bucket full of water. In addition I put in an extra hole in the top of the bucket so that I could periodically check the level of the water and easily re-fill it when necessary.

Here is a photo of the elements mounted to the bottom of the metal lid:

I have done some calculations to see what the resistance is for my stator, to reiterate: I have 12 coils, wound with #12 wire, 80 turns each, using a total of about 950' wire. From there I have 150' of #8 wire to garage, then 150' of #2 to the rectifier.

Resistance values:
#12 : 1.5883 ohms per 1000'
#8 : 0.6282 ohms per 1000'
#2 : 0.1563 ohms per 1000'

Resistance for single 'spoke' of star:
each coil = 0.125 ohms : Total resistance per 'spoke' of the star = 0.503 ohms (4 coils per 'spoke')

Resistance for wiring from tower to rectifier (each line):
#8 wire = 0.094 ohms
#2 wire = 0.023 ohms : Total resistance per line = 0.117 ohms

I then disconnected the wiring to the rectifier block and measured across each leg and got 1.3 ohms which would be the resistance of one complete leg (two 'spokes' of the star), and it works out correctly. (each leg is meant as between two of the 3-phase wires at the rectifier block).

The resistance per 'spoke' of my coils is a bit less than when winding two-in-hand, 59 turns of #14 wire (.596 ohms). My stator with more turns per coil will mean a higher voltage, and a lower cut-in speed, but it appears that it is certainly workable.

Initially when I made my blades I basically copied the numbers from the 17' diameter cedar blades that S.D. Scott made up for the OtherPower turbine, 3 degrees at the tip, 6 degrees mid-point, etc. Stew Corman has been helping and he thought that the TSR was closer to 6.

According to Stew they should rotate at about 2x the loaded speed. A little over a week ago I let my blades free spin in a very low wind speed, and within short order it sounded like a helicopter landing in my field! A bit frightening - and was able to stop it using my brake actuator (go kart brake).

I checked both the bladedesign.xls posted on OtherPower, as well as purchased the Blade Designer software from WindStuffNow, and both of the basically said that the TSR was about 5.5, optimally with a 17' turbine I have found out that you would like to have a TSR of about 4.25. This would give better low speed torque, and sacrifice a little high-end output.

It is interesting to note that I reread the last page on the original 17' OtherPower turbine and they state: "So far the machine works quite well. It has such a huge swept area compared to our previous machines that it seems to start up in practically no wind, and it's making a little power by the time the anemometer says 5 mph. At 10 mph it's doing around 400 watts and at 16 mph it's up around 1.5KW. Above that I believe the blades are overpowered a bit by the alternator. I do see 2KW from it frequently and I've seen about 3800 watts from it a couple times in very high winds, but overall I believe the blades are held back a bit in higher winds by the alternator. I can improve it by adding a bit of resistance to the line - this would allow it to speed up in higher winds and the blades would run more efficiently - but as it is it seems very slow and peaceful, and it rarely goes over 200 rpm"

Even with 59 turn coils in the stator they mention that the turbine overpowers the blades. So I believe that I have a bit more of a mismatch between my blades and my turbine.

To help prove the point further, yesterday a storm blew through my area bringing about 16-18mph winds, according to my inverter it was putting about 1.1kw back into the grid. In addition I stood out in the rain and noticed that the tail was about 30% furled.

Some good data! 1) my turbine is capable of good output 2) the tail furls properly (possibly a little bit soon) with tail of about 10 sq/ft 3) using a stop watch I calculated my blade rpm to be just under 120rpm 4) my brake actuator can stop the turbine

There is also a certain amount of understanding that you have to get, typically by measuring and observation.

I am still learning on this project and I think that in spite of it being a bit painful that it is good to know in the long run how each component in your system works and the interaction/dependencies that occur between them.

If you go a similar step that I did - in making your own turbine - build into the system the ability to perform measurements on: wind speed (near top of tower), DC amps, voltages and Hz. If you do then confirming that your turbine is outputting what you believe that it 'should output', and diagnosing problems/issues is easier.

In addition I don't know anyone else with a 'grid-tie' system, and I am finding that I have a fair amount of control over many variables, which need to be tuned to *your* system. After playing around with many combinations I have finally came up with some values that are some-what compensating for my mismatch blade TSR/generator.

By upping the system voltage to 56.8v before the inverter goes into 'sell' mode I find that I can generate power in as little as 5mph winds.

(btw: LaCrosse makes it), it is FAR better than the Taylor unit, as it updates every 5 seconds and I believe that now I have a good wind speed indicator. I can even get reception in my house (about 200' from tower) through the walls!

The Taylor unit was way off, right now we have 8-10mph winds @ tower and I see the tree tops rustling with a little swaying. the Taylor was saying 6mph, so I have to believe that the other day when the storm blew through here that the winds indeed did kick up to 25+mph when the tail was furling!

I decided that this time I am not going to bother putting in "plugs" into the center of all of the coils, rather taking the simpler approach of a solid stator, and no air circulation inside of the coil centers.

The design is all 1/8" thick steel, and borrows heavily from the coil winder that is sold by Ed at WindStuffNow.com. If you are not up to the task of making your own coil winder I would highly advise purchasing the one sold by them, it is a bargain at $14.95. Add on their universal triangular winding head for another $12,95 and you are good to go!

Here are a couple of pictures of the new coil winder:

To help out even further every 10 windings I put a dab of super-glue across the coil and kept on winding the coil. I got all 12 coils wound correctly this time, using #14 wire 2-in-hand, and each coil has 59 turns! I used every bit of the two 800' spools in making up my coils with about 50' to spare.

With all coils wound, I placed them into the mold, I got the 4 coils for each of the three phases soldered up. Next I soldered the leads of the output to some 3/8" phosphor/bronze bolts, this time I will be putting the bolts into the stator. I also am bringing the center of the star outside this time, it will have a 1/4" bolt. Currently I do not have the tower wired for this but maybe in the future will make changes to accommodate it.

This time I lashed the blades to the side of the tower about 15' up off the ground so the goats/sheep could not fool with them. It was easier than carrying them into the garage again.

Today I went to my local ACE hardware store and purchased a couple of v-belt sheaves, I purchased two 8" OD and one 4" OD sheaves. I intend to attach one of the 8" pulleys to the turbine and put one of the other pulleys in my chuck of my 2hp metal lathe. With two different sized pulleys I can attach a v-belt and test out the turbine at the following speeds: 64, 78, 105, 128, 150, 165, 210 and 330 rpm's.

I intend on changing my rotor-gap to achieve a cut-in speed of between 75-80 rpm. In addition I will check the open circuit volts at the various speeds to plot the voltage output. Next I will put the output through my dump load with an ampmeter connected so that I can see exactly what the output power will be. Once I do these two set of tests I can compare the output power against a graph of the prop performance to see how closely they match and determine how compatible they are.

I also had some 7" OD bar stock that I sliced off about a 1/2" thick piece, drilled out the six 5-1/2" bolt holes to mount onto the turbine. I drilled out a 3/4" hole in the center of this piece, inserted a length of 3/4" bar stock and welded it from the back side.

This bar stock has to be long enough so that the pulley extends past the all-thread on the end of the turbine.

However I used a technique similar to lost-wax casting. I made some center-blocks out of urethane foam, and then wrapped the outside of the plugs with electrical tape, finally I dipped them all into a jar of PVA. More about them at the end of this section. Here are what they look like.

First thing we did was to re-wax the mold and then Paul applied a thin layer of PVA to help the mold release. Next was to put wax plugs into the holes where the stator output bolts go, this is so that there was no chance of getting any resin into them and adhere around the bolts. Then the threads on the bolts were also waxed for same reason. The black lines drawn inside the mold are to help to position the coils properly.

I had some 9oz glass cloth that we used and I cut out one piece about 30" diameter (8" larger than the outside mold). The center was cut and the outside diameter also cut towards the center. The second large piece was a donut shape that would be placed on top of all the coils. We also used some fiberglass cloth tape and wrapped the outside edges of the coils, the thought here was to help bond to the cloth strips around the outside edges.

We wetted out the mold and placed in the large cloth piece, on top of that we placed the coils and pushed the bolts through the waxed holes, positioned the coils appropriately and then inserted the coil 'plugs' into the centers of the coils.

Paul mized up a batch of vynlester resin (3-4 qts) and mixed in a whole bunch of chopped fibers into the resin and mixed in about 1% MEK Peroxide hardner. This is about the least amount of hardner you want to use, but I wanted to keep the heat down to a minimum so that cracks would not form.

This resin (consistancy of pancake batter) was poured over everything and we used brushes to help distribute it all over the inside of the mold. Then folded over the extra lengths (inside and out) of the bottom cloth back over the coils, and put a bunch of cloth strips all around the outside edges of the mold and the finally the donut shaped piece of cloth.

Lastly we put on the mold top and tightened down the 1/2" bolt to start compressing the mold contents from the center towards the outside, then using lots of C clamps around the outside edges of the mold and compressed it more to push out the excess resin.

Ok - getting back to the coil plugs... Once I release the stator out of the mold, I should be able to easily 'punch' through the outside layer of cloth and with my fingers easily remove all the foam from the center of the coils. The electrical tape will help show me where I should be careful so that I do not nick the enamel insulation. When done the coils will be able to have significant air flow through their centers, while making it easy to remove the stator from the mold.

The stator came out of the mold ok this time and did not damage the mold. The vynlester resin did 'eat' into the layers of paint that were in the mold, it worked out well casting the bolts into the stator. Once cleaned up I painted the stator white again for UV protection. Here is what the new stator looks like un-painted:

I started out using the new stator with the exact same air gap and spun it up using my lathe, here are open circuit voltage measurements at various rpm's

First the generator with the old stator

Next the generator with the new stator installed

After this last set of tests I slightly changed the air gap and reduced it by about .010" (now about .930" between rotors) which should slightly lower the cut-in speed to about 90 rpm. You can see that the original stator produced way too much voltage, the new one is far superior. By weight almost the same amount of copper in both stators, the difference is how the coils are configured...

While looking over the generator I noticed that the resin at the very outside edges around the rotor magnets had began to separate from the rotor plate, probably since it sat on top of the tower all winter (not running) and some moisture got into the edges, froze and began separating the resin from the rotor. So I mixed up some epoxy cement and filled all the small gaps.

While down I welded on a bracket to help hold the wires in place it is made from 1/2" rebar, I also bent the pick-up point towards the rotor to make it more balanced when picking it up:

This new wire support worked out great! I have a short length of cable running from this support down to the kellem grip which remains inside the yaw bearing. I used two cable clamps to connect the end of the cable around the support. If I were to do this over again I would have made the support about 2" shorter as it got tangled up a bit in the electric winch when I lifted the generator back onto the top of the tower making the installation a bit tougher than it could have been.

I also added a reinforcement to the diagonal brace for the davit crane and touched up the paint. The support 'wraps' the lower section of the diagonal brace, allowing it to slide back/forth within it. This proved necessary because with out it the bouncing of raising and lowering the generator started to create a slight bend on one of the acme threads:

Today I got everything back up onto the tower, blades, tail and got the telescoping stub raised. Things went well, this is only the 2nd time that everything was raised and the procedure is starting to become a known effort.

Of course it took a couple of hours for the wind to come up enough to start the blades rotating. It seems that after tightening up the bearing it needed an 8mph wind to start the blades rotating. Once they started they continued to spin even in 3mph winds. It seems that it was at cut-in speed at around 7-8 mph. The blade speed appears to be around the projected cut-in speed and putting out about 13hz at just over 80 rpm. The slight noise coming off the blades and low growling off the generator is different from the last stator, and I have to believe that it is proper and normal as it is generating more useful power than before.

The other day we recorded a 59mph wind gust from the anemometer at the top of the tower, unfortunately I was not home to observe this, however wind turbine appears to be ok.

In addition, during the middle of last night we also got a 44mph wind blow through.

I am starting to look into data logging capabilities, it would be nice to know what the turbine output is without having to be present with a meter to measure...

I researched this on the FieldLines.com discussion board and here are to discussion threads that explain this:
Matching the Alternator to the Blades by DanB
Matching the Load by Flux

The circumstances that I have fit these explainations, and is actually preferred. To remedy it all I have to do is add some resistance to the line. Most of the discussions I've read talked about inserting a high power wire wound resistor to the positive wire coming from the rectifier. I also put out the question about changing the wild AC lines and this was the answer given by Flux:

"Yes you can add the resistance in the ac cables, there is no real difference in the value whether you add it in 2 dc cables or 3 ac cables, the current in any 2 ac cables at any instant is effectively that in the dc line, that is how the rectifier works.

Unless you can use the heat it seems a far simpler solution to just add the loss in cables rather than obtain awkward value heavy duty resistors.

Just try adding some cable to find what you need then your wire tables will show the prospects of substituting the aluminium.

I am sure that for these higher voltage units people use heavier cable than necessary in the belief that it helps output and it doesn't always work. Just as using a lower than necessary cut in speed is very detrimental to output but they all do it."

(Side Note: Flux's comment about too low cut-in speed was exactly the condition that I previously had with the old stator. Now I understand that with a very little wind (4-6mph) the blades are unloaded and will increase in speed. When they increase in speed they will probably get the voltage up to cut-in speed, the inverter will kick in putting out power which will shortly slow the turbine down enough to drop below cut-in speed and then the cycle begins all over again. So do not base your cut-in speed against wind speed, but against the rpm that the blade profile can support the amount of power that the wind turbine will be putting out!)

Currently for me the simplist solution would be to use a power resister, and currently another person on FieldLines.com has one to sell me pretty cheap, it is a .6ohm 6-8kw resistor, total weight is about 12lbs. I researched this and if I had to purchase one new from a manufacturer it easily could have ran $300-400. So I will probably start out with this as I can pick it up for a song.

So I connected them in parallel to get .5ohm total and put them onto the DC side after the rectifier. As long as the total current output from the wind turbine stays below 30amps I will not run into problems.