Hard Disk Generator...

Part # 4 (Jan 31st 06)...

This episode is more on the Coils and Cores, and finally the addition of a 2nd Magnetic Rotor assembly.

Certainly I don't want to loose readers through their frustration at reading the incremental failures that seemingly punctuate my attempts at building a wind generator system, but rather that you will share in the triumph that will be mine when (not if) this project reaches completion. 

Similarly, I know that there are numerous sites that already provide proven plans for predictable output if properly followed, but I felt that there would be room for just one more with the twist of illustrating the not-so obvious road that lead to that end-point.

Higher Current via 2-Strand Coil Winding

The simple double spool holder to the right could have been a web-site unto itself as it took me an entire day to get to the point shown...

In all fairness, it was a Sunday and I was nursing a mild hang-over.

The spools are wound from the factory spool (pictured on the previous page) with the aid of an electric drill.

At this point I'm still clinging to the idea that this will be a Single Rotor generator, and opt to continue on with the pre-formed "Backed" cores used earlier. (Pictured above left)

To the right is the first of 3 coils that were wound with 2 strands of wire.

As the coils came off the winder, I used nail polish to fix the windings into place.

The pocket knife was just the right weight to hold it flat while the polish set-up.

The three test coils pictured are 125, 100 & 75 turns of 2 strands AWG#24.

If nothing else I'm pleased with the uniformity of coils.

The individual strands tips are heated and cleaned to expose bare copper, pig-tailed and soldered to ensure a secure electrical connection for testing.

As these coils seem more fragile than the last set, I opted to use a thin length of plastic to simulate a consistent air-gap between the coils and the magnetic rotor.

The plastic is 5/32" thick, this may be a bit on the conservative side as I'd planned on eventually building the final generator to tight enough tolerances that the air-gap would be between 1/16th" and 1/8th".

The results were less than stunning 

*note - the original 200 turn single strand coil was re-tested at the same time for reference.

*note - the original 200 turn single strand coil was re-tested at the same time for reference.

So the results are less than encouraging and even somewhat confusing, in that the 75 turn & 100 turn coils are engaged some kind of a dance swapping places for "Top Current Producer" over the range of test RPM.

I honestly expected a far more linear relationship between the coils, but there must be some non-linear component at work over the speed ranges. The "Drop-off" of the 125 turn coil is likely a result of the higher internal coil resistance, while the 200 turn single strand is left in the dust.

*note - the original 200 turn single strand coil was re-calculated at the same time for reference.

The chart above is the calculated stator output based on a 2 phase series wired configuration. The basic formula used to derive the above is;

Vdc(output)=single coil Vac X #of coils per phase X Vrms constant - bridge voltage drop X # of phases wired in series.

using the 75 turn 2 strand coil at the 280RPM reading...

Vdc=((0.56Vac rms X 4 coils per phase X 1.414) - 1.2V) X 2 phases in series

If there is a glaring error in the above please don't hesitate to set me on the right track...

But assuming that the above is correct the graph below is the result.

Given what I'm seeing above, for a 12V charging system, somewhere between 400 to 450 RPM would be the cut-in speed for either the 100 or 125 turn coils... Bear in mind that the 200 turn single strand coils is not an option as it's internal resistance is so high that little usable current is available. This was put to rest in the previous section.

I can feel my lower lip start to quiver and my eyes are welling up with tears as one of my thoughts was to utilize a broader range of wind speeds by actively configuring the 2-Phace coil arrangement between Serial and Parallel. The calculations point to a near permanently Serial configuration with a very slight operating range for Parallel above the 560 RPM mark. 

The benefits of Parallel are double the current, but at half the voltage. This would require 24Vdc or greater output from the stator prior to switching from Series.

And Again with 3 (Three) Strands per coil...

I let the results posted above fester for a couple of days, posted a few questions on Forums, tried to understand the replies and waited for some inspiration... 

One of the possible solutions that I could try to implement is a 2nd Rotor with an equal number of magnets over the initial rotor. The idea as I understand it is that the 2 rotors would be aligned with opposing poles facing each other. This would increase the density of the "Flux" or magnet field that passes through the coils as the rotor in in motion.

One of the recurring suggestions that I'd received was that due to the rather tight magnet spacing that I have on the rotor I was getting less than optimal flux density as the magnet lines of force were being "Shunted" to the next closest opposing pole (immediately beside it).

The illustration to the right certainly isn't to scale or even based on anything beyond the general concept stated above...

Upon careful review of my vintage 1970's text on Electricity, it would appear that the lines of flux bend as they encounter the copper wire by a secondary force resulting from the induced electromotive force within the wire itself.

No attempt was made to incorporate that into the diagram, but the orange circles are supposed to represent the cross-sections of the coil passing thru the magnetic field...

The +ve and -ve designation onto the copper was arbitrary and has a 50/50 chance of being correct... though is irrelevant in the big scheme of things...

this space blank.

The previous "Backed Coil Cores" would have a negative effect on the magnet forces in a 2 rotor system so I opted to create another core mold and cast new ones. While I was doing the layout for the mold I increased the size of the core's diameter to fully cover the face of a single pole as it is measured on the rotor relative to the circular motion of the rotor, rather than the straight line of the magnet itself used earlier.

The mold was machined into EPS foam and painted as before.

These are the cores just out of the mold.

I've found that the cores are easily removed within 45 min to an hour after being poured.

The catalyst or polymerizing reaction is still underway as they are warm to the touch.

(Vaseline is used as the releasing agent)

Here are cores drilled and marginally cleaned-up, ready for winding.

The 2 images above are of the 3 (three) strands being spooled out to the coil as it was wound. 

The coil winder required a pair of larger side disks cut and drilled to hold the coil in place.

This is one of the 2 test coils that I wound.

Again the coils are tinned for a secure connection.

The coils are 75 and 100 turns of 3 strands of AWG#24 wire.

I opted to drop the 125 turn coil as it was exhibiting less than promising results at 2 strands.

The coils were taped onto the plastic strip for easier placement between the 2 rotors.

This is the second magnetic rotor just pulled out of the mold.

I did drill 3 holes in the saw blade backing disk to accommodate 3 jacking bolts.

Not shown are 3 nuts embedded in the resin on the opposite side. The bolts were carefully coated in Vaseline to ensure that they could be easily removed.

A spacer or support was machine on the lathe from a section of 2" PVC drain pipe (sitting on the rightmost rotor).

The rotor on the right is securely clamped into a vise, while the rotor on the left has the "Jacking bolt" heads hooked into the T-Slots of the mill table to ensure that there are no jumping and fusing of rotors.

It may seem a bit on the extreme side, but prior to taking this shot, I was playing about with the rotors and the attractive forces between them are very great.

The Jacking screws worked like  a charm as the rotors were eased together in a controlled and orderly fashion.

The PVC spacer appears to be just the ticket as the lathe ensured an accurate and parallel cut on the supporting edges.

The rotors are held in place by nuts and washers, attractive forces and held apart by the PVC spacer.

The spacer was sized to 0.600" height, sets the coils approx in the center of the 2 rotors when using the 5/32" plastic fixture.

The dual rotor is chucked into the drill press and spun over a range of speeds and recorded for the 2 coils under test.

As an aside, I'm amazed at the capabilities of digital cameras as the disk is spinning @ 450RPM while the photo is taken and appears to have been completely stopped.

Note - The xxx-2 refers to the previous single coil 2 strand data, while xxx-3 refers to the current dual rotor 3 strand coil data.

I don't need to graph this to see that there is no significant improvement in the Open Circuit coil Voltage... if anything a few values are marginally lower??? Although the meter readings do tend to scatter while testing I've made every effort to hit a consistent reading 3 times before recording it.

The Short circuit current values are marginally better, but I have a hard time accepting the additional work and materials required for the second rotor for the sake of a couple 100mAmp gain in output.

Again, there is something amiss that I'm just not getting... I know conclusively that I made a mistake using AWG#24 wire as it's far too light for carrying any appreciable current due to it's relatively high resistance. But the current gain posted above should have been considerably greater given the 3rd strand reducing the internal resistance of the coil, as well as the benefits of the second rotor.

The change in core diameter may have had an effect and this will be investigated by testing the previous 2 strand coils within the dual rotor setup.

The Air-Gap between the two rotors may also be a concern that can be verified by reducing it by 1/16th" increments and retesting against a know data set.

There may be a MAJOR FLAW in that the to rotors should be positioned such that they have Similar poles in alignment ie N over N & S over S, as opposed to the current N-S, N-S config (this may be tougher to get as the 2 rotors really don't want to face that way... a 2nd set of hands will be required...)

And to top it all off, I scanned the surface of the 75 turn 3 strand coil and placed an outline of it over the rotor to see how much room I had to deal with...

From what I can tell 75 turns is the absolute limit with the 7 1/2" rotor.

This gives me lots to consider as I close this installment...

As always I'm grateful for any advice based on experience... I certainly don't want to be rude, but email that starts with "I've never actually done this but..." isn't overly helpful at this point.

Hard Disk Generator , 2, 3, 4, 5

Back to Energy

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