Hard Disk Generator...

Part # 3 (Jan 11th 06)...

This installment covers a lot of ground, and as such is quite lengthy. The main issues that I wanted to address were to try a number of suggestions from others via forums and newsgroups to finalize the rotor magnet layout, resolve a minor issue with how I was winding coils, set the coils and Magnets into resin, and take some measurements with all the coils into place.

Alternative Magnet Placement(s)

This arrangement is of 16 alternating single poles spaced to accommodate a "Leg" of the coil to straddle on either side of a pole.

Although the coil shown does not have a core, both cored & un-cored were tried, the best measurement was with the cored coil at 1.2Vac.

Note that there are 3 (three) layers of magnets, this equates to an equal mass of magnetic material as 2 layers of 24 poles.

This second arrangement is of 2 layers of 24 poles, but precariously arranged to "Double" the poles such that they are viewed as "N-N-S-S-N-N..."

The repulsive forces of the magnets saw a few fly off the rotor plate during the testing. But again the peak voltage I could read was in the 1.5 to 1.6Vac range.

The above are included as I do want to demonstrate "Due-Diligence" given that the folks that offered the advice certainly put considerable effort into their replies to my questions.

A third alternative was not investigated as it would have required a much larger rotor plate, and one of my "self-imposed" design criteria is to stick with the 7" saw blades that I have a reasonable supply of on hand.

Resin Casing the Rotor

I cut-out a wood spacer to ensure that the magnets are oriented in a consistent manner along the edge of the plate.

Each magnet pair is glued down onto the plate with a small dab of "Metal-Epoxy". The 2-part glue is formulated specifically for bonding metals.

The rotor mold is a 1/2" plywood form that is screwed down onto another board.

A sheet of paper is inserted between the rotor and the base of the mold in case the Vaseline doesn't act as a releasing agent the way that I hope.

The magnets are faced down, the resin is mixed and poured into the 5/8" center hole until it appears out between the magnets.

The edges were poured to just above the level of the back plate. 

In the future I think that I'll use a steel sheet as the base of the mold, to ensure that the rotor assembly doesn't float-up and allow resin over the face of the magnets.

The rotor required a measure of clean-up, though the removal of the resin from the faces of the magnets was slower going as I didn't want to damage the face of the poles.

Pre-Formed Coil Cores

One of the problems I had with the Coil Winder was in removing the hub from the center of the coil when I finished winding. Of the various test coils that were made, all tended to lose their form and  lacked consistency in their shape.

Above to the left is a mold that I routed out on the mill for a trial batch of cores that would be cast in resin and Magnetite. To the right is the mold with 2 coats of acrylic enamel that was intended to act as a vapour barrier between the catalyzed resin and the foam, which I know don't get along. 

I had under estimated the volume of resin and black sand that I would  need and ended up with a couple of empty molds.

Also I mixed the sand to resin in a much lower ratio in hopes that the material would "Flow" and fill the molds completely.

The mold was liberally coated with Vaseline to ensure that the cores could be removed once cured.

Once removed and cleaned-up, they were drilled to accommodate the winder crank and a drive pin.

A 3rd hole was drilled (barely visible - even if you know where to look) that is for the starting lead of the coil.

To the left is a spool of AWG#24, as I simply gave-up on un-winding motor stators... Certainly in a "Post Apocalyptic Era", recycling on that level would be the order of the day, but for the present, I feel that I have such a significant lack of basic knowledge to overcome scrimping on such a fundamental component as the wire is not furthering the progress of this project.

Above to the right is a completed coil still on the winder.

I'm real pleased with the quality of the coils that are coming out now as they remain flat, and they all look the same.

The lower density of black sand in the resin is quite noticeable based solely on how "attracted" the coil is to the rotor.

Also there is a measurable difference in the output as 1.7Vac per coil vs the 2.0Vac with the higher density cores. Both still have 200 winds.

Resin Casting the Coils

Since I had the exact shape and size of the coil cores already on the computer, it was a simple matter to scan an image of the resin encased rotor to make a template for the placement of the coils. 

In the course of cleaning-up the scanned image, I opted to place a white dot consistently on one of the poles for reference. And it was when I started to push the outline of the coils into place that I finally understood why I was being told that my original design would NOT Work...

Originally I had posted that I thought 8 Coils would lay over the 24 poles in a single phase configuration. And in a split second of enlightenment I knew that it was flawed.

For a single phase config, the count would have to be either 6 or 12. For a 12 coil config I'd have to revise both the core shape/size and start experimenting all over again with the winding count... 

So 6 Coils it is.

A second mold is made to accommodate the stator, and it's a similar arrangement with 1/2" plywood screwed down onto another board.

The coil template was printed to scale, centered and placed between the mold and the backing.

Yet again, all contact surfaces are coated with Vaseline to ensure that the resin casting can be removed.

Of note is that once the coils were placed in position I made doubly sure that the leads were pulled clear of the mold area out to the edge, all this work would be lost if even one of the 12 leads was buried in the resin casing.

The cured stator was easily removed from the mold and is pictured above. The 4 tabs out the sides are for positioning the coils as close to the rotor as possible.

This is yet another fixture that needed to be made, but will hopefully serve me well into the future. 4 lengths of 3/8" #16 all-thread are positioned to mount the stator in place over the rotor assembly with the capability of leveling and setting the gap between the two.

The steel jig is held securely to the drill press table with (what else but...) Hard Disk Magnets.

It took surprisingly little fiddling to get the rig assembled into place and semi lined-up.

Notice the shiny new Meter... This was a "Princess Auto Special" at $29.95 Canadian. The reason I got it was that it has support for a "K-Type" thermocouple for the foundry. But it also has a handy "Hold" feature that allows images like this to be caught without having to scramble for the camera while everything is running.

The diagram above is the setup that I used to compile the data listed below, of note is that the Battery was initially sitting at 12.4Vdc at the start of the test, and was at 13.3Vdc at the end. Also the Voltage readings were all Open Circuit, meaning that the battery was NOT connected as a load. A 100uF capacitor was added to smooth the DC voltage this is purely optional (it happened to be sitting beside the bridge rectifier). The generator was tested with a drill press, and the Wattage values are calculated between the Battery Voltage and the current being sourced.

On one level I'm less than happy with the results, but overall, I think that all the work to this point has been worthwhile.

Effectively I've built a generator that has very close performance characteristics to the Permanent Magnet Generator that is currently on the pole. I had thought of stringing 2 batteries in series, creating a 24V load, but opted to pass on that as even if the current that this unit could source stayed the same at the higher voltage (which it wouldn't...) the gen would still only be approx 25watts at the highest speed (which equates to a wind speed that we rarely see up here, with the proven blades that I will be using).

My objective is to hit 100Watts output into a 12V load.

To that end I feel that the coils need to be altered. Firstly the DC resistance of the 6 coils in series reads 9.0 ohms (I realize that this reading is misleading as inductors are not purely DC devices and have unique impedance properties), but from a baseline of reference I think that value has to come down.

This can be accomplished by using a heavier gauge of wire, as I just bought a large spool of #24 I'll double-up the windings by wrapping 2 strands in parallel, this doesn't equate to an AWG #12 but I think closer to an AWG #20 or 21. The additional copper in each coil should translate into higher current output.

Also with regard to the coils, the pre-molded cores are a success, but they require a slight modification in that a higher ratio of magnetite to resin should be used and the center of the cores could be 1/8" larger, and slightly shallower resulting in flatter coils. This will require more testing to achieve a balance between the double conductors and the reduced # of turns.

Also the voltages I'm reading are somewhat higher than I need, so I may be able to drop the # of turns down to 150 or 125. This may effectively increase the cut-in speed more than I want, as I do know that the Volts per coil are directly related to the number of turns.

The diagram to the left is a modification of my original design, with the allowance now made for the "Out of Phase" coil configuration. This system is essentially a 2-Phase design, similar to the Stepper Motor output that I started all of this with almost a year ago. This addresses the reduction of windings per coil with the addition of 2 more coils. The 2 separate output sections (a & b) can be bridged to DC and either connected in series at low speed or parallel in higher winds (this last statement is highly speculative, and only based on what I understand to be one of the advantages of 3 phase).

So in closing, the existing magnet config will move forward, the stator is scrap, a test jig and molds are done. But I have a nagging suspicion that this will end up as a 10" rotor, 3 phase config before it actually hit the tower.

Hard Disk Generator , 2, 3, 4, 5

Back to Energy

Disclaimer (an unfortunate necessity)
All Rights Reserved theworkshop.ca © May 25, 2008