Dear Diary,
I'm just so excited! Today I finally finished the conversion of 'Blue Boy', a 500 W motor. Let me tell you the full story, dear diary.
This is what it looked like as I got it:
A small 500 W, 2 pole (3000 RPM) single phase 240Vac motor used as a swimming pool pump. Admittedly, 2 poles are not ideal for conversion due to the geometry of their stator (long and thin, as opposed to short and stubby for slow running multi-pole generators). Nevertheless I decided to convert it. I had made a few more conversions before but had never rewound a motor. As the next project in the pipeline involves rewinding a 10 hp motor I thought I'd better practice a little first. The Blue Boy was the ideal victim and so, encouraged by Zubbly's 'go for it!' on IRC, I started my first motor rewind project. (the shaft that is lying in the foreground of the photo was leftover from a previous conversion project, the 3 hp motor (http://www.fieldlines.com/story/2006/12/17/22270/167). The shaft formed the raw material to machine a new shaft and rotor out of one piece)
...You gotta tear down before you can build up...
The motor taken apart and the original windings removed. I clipped the ends of each coil off with a sidecutter and then carefully hammered the remaining bits of copper out of the stator slots. This must be done carefully as you run the risk of damaging stator laminations (http://www.anotherpower.com/gallery/album68/blue_monster_damaged_laminations). It doesn't work for all motors, but in this case, the copper cores came out in one piece with relatively little effort. Still, all in all there's a couple of hours of work involved in removing the original windings.
Some of the copper 'cores' were saved for future reference. They are needed to determine the slotfill. By counting the number of windings per slot and measuring the diameter, we can determine the amount of mm^2 of copper per slot. This is information we need when rewinding the stator. Slotfill was 14 mm^2 copper.
In the above picture also can be seen the new front end-bell. I happened to have an identical motor (different brand, but obviously the same Chinese factory) with a more suitable endbell than the original one. The original motor that the spare endbell came from I ruined as I tried removing the windings from the stator and damaged it too much.
There are 24 slots in there so an 8 pole 3 phase conversion turned out to be one possible option. After some sketching, calculating and thinking it proved to be possible to make an 8 pole conversion using 12x6 mm round magnets (N42 grade).
The original rotor can be seen here (which was also slightly rusted on one side; chlorinated swimming pool water will do that to steel). A new shaft and rotor were machined out of one piece of steel (the original shaft from the 3hp conversion). The shaft was made a little longer than the original as well. Above picture shows the steel rotor before the aluminium sleeve that holds the magnet was fitted.
A sleeve was then made from aluminium. Holes were drilled to hold the magnets in position. It's hard to see in this picture but the poles are skewed. The skew angle for this particular motor was 3.9 degrees (for more info on how to determine skew angle, see this file: http://www.otherpower.com/images/scimages/3538/decogging_tutorial_V1.pdf)
To mark the holes that needed to be drilled I used some CAD software to draw an outlay, scale 1:1. This was then taped around the blank aluminium sleeve and used to centerpunch the holes. This picture shows that method used on another conversion: http://www.anotherpower.com/gallery/album49/130W_rotor_marking)
After the centerpunching the holes can be pilot-drilled, pre-drilled and drilled to final size. In this case it was a tough job as there was very little room for error, as the magnets nearly touched eachother at the bottom. Drilling took a few hours of work with the necessary concentration so as not to screw up.
After the aluminium sleeve was drilled it was deburred and Loctited (using Loctite 638) on the steel rotor.
After the Loctite had cured the rotor assembly was put in the lathe again and turned down to final size. This ensured that the outside of the rotor is aligned with the centerline of the shaft.
The 48 N42 12x6 mm magnets (~1/2"x1/4") were then installed using epoxy adhesive. Notice the use of epoxy, not polyester or 'fibreglass resin', as it's commonly called. For anyone wanting to know the reason for using epoxy and not polyester, read this story by Zubbly: http://www.fieldlines.com/story/2004/11/7/132725/326
The picture above shows the rotor with magnets installed being inserted into the stator. Just to see whether it fits 
At this time I guesstimate there's about 20-25 hours of work invested into it, so it would be, ehm, un-nice, if things didn't fit.
Now the rotor is taken care of we can start work on the stator:
The first thing to do is to determine how many turns per coil are needed. This is done by installing a single test turn (wave winding) into the stator. (This is also where I went initially wrong...). After consulting with this board (http://www.fieldlines.com/story/2007/10/7/72559/4261) I had settled upon a cut-in speed of 250 RPM in star for a 12V system. Driving the generator with a temporary single-turn (I thought...) winding installed with a cordless drill gave me the RPM/V for this particular genny. It turned out I needed 7 turns per coil, so with a slotfill of 14 mm^2 I decided on 7 turns of 2-in-hand 1 mm wire. The testwinding was then removed. In the picture can also be seen the borrowed coil forms (the white nylon thingies) that will be used for winding the coils. A single test loop of wire is installed to measure the distance needed between both coil forms during winding of the coils.
Slotliner was cut to size and installed. I'm in the lucky situation that there is a motor rewind shop in my small village and that they are interested in my experiments and don't mind me asking for some material (as copper wire, slot liner, slot cover, coil winding forms, lacing tape, old 10 hp motors, ...) at times. It's also interesting to have a look around in the shop and see how the professionals rewind). In the above picture the slotliners can be seen. They're there to prevent the windings from shorting to the stator over time. If you want to know why insulation and varnishing is needed, then check out this high-speed video of the windings in a motor: http://www.youtube.com/watch?v=lPNU2vniUIc&feature=related
The next step was the actual winding of the coils. I used the borrowed coil forms and installed them in a make-shift coil winder. The distance between the coil forms is determined with the help of the test loop as shown in one of the pictures. The coils were 7 turns each, 2-in-hand 1 mm wire. Coils were wound individually and later soldered together. I did it like this because I thought it would be easier for a first rewind.
Inserting the first coil into its slot took about an hour. I then removed this first coil and inspected the wire for any possible damage (scratches) to the varnish. It was fine. The next coil was inserted and this time it went much faster due to having developed the proper routine. The next coils took only a few minutes each. I found that the trick was using properly sized bits of wood to prevent the coils from jumping back out of their slots as you're dealing with the next loops to be inserted. This was a little trick learned from the visit to the motor shop.
Yay! All windings are in! Next step: solder all the individual coils together, making sure to wire them up correctly, to their correct phase, and in the right direction...
Hard to see in this picture, but the coils have been soldered and the joints insulated and bent in position. The coils were also bent backwards so as to leave the stator bore clear. Each phase was measured with the milli-ohm meter to check for possible faults. Each phase measured between 110 and 116 milli-ohm so I figured it was fine. This test, however, does not show up possible inverted connections.
Next the rotor was inserted for a testrun. A cordless drill rotating at 1000 RPM was used to drive the shaft. The good news was, the genny worked fine! Except that cut-in speed was not the expected 250 RPM but 1000 RPM...
This had me scratching my head for a few hours (there was -exactly- a factor 4.0 difference in cut-in RPM) till I realized my mistake: when I had installed the single-turn wave winding I had calculated how many turns I needed per coil. But, in an 8 pole generator (with 4 coils), one loop in wave winding is actually equal to 4 coils of 1 turn, not just one coil of one turn, as I had used in my calculations. Hence the factor 4 difference in cut-in RPM. ARGH!
Putting the sidecutters into the preciously rewound motor did hurt. A lot. Sigh. Evil tongues say tears welled up in my eyes. I deny this though.
...back to square one...
So, instead of using 7 turns per coil I needed 28 turns per coil. This time I used .7 mm wire which gave me a slotfill of 10.5 mm^2 (there was room for 14 mm^2 but I didn't want to fully stuff the slot with wire as this was my first rewind; no need to make it too difficult just yet). This time the 4 coils of each phase were wound in one piece. This would make inserting the coils a little harder but save a lot of work and risk of mistakes when soldering the coils together. No joints also means a more reliable system.
Winding the new coils was relatively uneventful.
The first phase is inserted. Inserting the new coils went much faster than the first time. Guess I'm still on the steep initial part of the learning curve where progress happens quickly
This picture also shows the slot covers that were shoved over the wires to help them keep them inside the slot. The slotcover is made from the same material as the slotliner insulation but is pre-formed by the manufacturer for ease of installing.
All phases are in but not yet laced. The rotor was inserted for another testrun to see if this time it would cut-in at the right RPM. And it did. It was spot on. The motor (actually, it can now rightly be called a generator) was disassembled and the wiring was cleaned up, routed to the right places and everything was bound together using lacing tape.
It's surprising how solid the windings feel when they're tightly bound up with lacing tape. They're already beginning to feel like one mass of copper. It could be just me but it looks much nicer as well when it's laced together.
The next step was varnishing the rotor to make the windings really solid. This is something I wanted to do myself but, lacking the proper varnish, I once more bothered the good people at the rewinding shop who put my little Blue Boy into a drying oven for a few hours at 140 deg. C, dipped it into their varnish tank and then put it in the curing oven. It got varnished and baked along with one of the big boys in the shop. Interesting tidbit: the varnish and the wire are actually part of a system, they're meant to be used together. Apparently one can't simply switch from one type/brand of varnish to another without also using another brand/type of wire. The lacquered copper wire and varnish have to be compatible with eachother.
The windings are now rock-solid. I just hope I'll never again have to take it apart for a rewind...
A last look at the finished motor, err, generator before it is closed again and the endbells are installed.
Next step was determing the generator curve. The generator was wired in star and connected to a rectifier and a small 12V / 7 Ah battery. The lathe was used to spin it up. Voltage and current were then measured for various RPMs. The graph below shows the charge curve for the generator:
As can be seen, it cuts in exactly at 250 RPM. [in the voice of Hannibal: 'I just love it when a plan works']. It puts out about 120 W maximum at 1000 RPM so it should probably be marketed as a 100W genny.
There you have it. One finished generator, ready to be used for a small windgenny. Well, actually, it still needs a new coat of paint. During the building of the generator I also started working on the yaw mount but there's still quite a bit work left. Also it still needs a hub and blades. It'll be a little longer before she flies just yet. But the most important thing is that it was a succesful trial in making a motorconversion that included a rewind. I'm glad I learned the hard lessons on this small cheap generator and not on the 10 hp conversion that work is now slowly starting on.
To be continued...
If you have read this entire longwinded dull technical story you must obviously have been very bored. But, for anyone still awake and interested, more and high-resolution pictures can be found here: http://www.anotherpower.com/gallery/album69?page=1
A final word of thanks goes to the motor winding company Stolk, especially André and Albert, this board, the people in IRC and especially Zubbly, who was the one who convinced me that rewinding a motor wasn't a hard thing to do at all. If it hadn't been for his advice and support this project probably never would have even started.
Peter.
