n-c alternator modifications: discussion and testing

[wcorey]

The standard four coil testing is over for the time being, I hope I covered enough bases there so to be adequate for some conclusions to be reached. I feel like I should go through and pick out the newer, more consistently done test data and move it over to the other thread, 23 pages is getting cumbersome to wade through looking for various results. I've gone back through and edited in some corrections, notating the edits and without changing the mistakes so as to keep resulting responses in context. So if you've copied data to your posts, don't necessarily rely on it's accuracy, go back and check the original source post.

What's next?
Bruce and Dad have been very kind as to provide me with a pair of extra coils to set up a six coil stator to further 'the cause' and I'm working on it as I type this. Logistically it makes the most sense to do this first, then try the modified series arrangement after, in both the four and six coils, of course.


Bill

[ecurbruce]

I must appologize, guys , I've made a critical mistake in the configuration of the stator as a whole...
Here is the corrected scheme of things, as far as I can tell;

new-coils.jpg

All coils in circuit (1-4) are on the outside, alternating counterclockwise, clockwise,
All coils in circuit (2-3) share coinciding spools and lugs,also alternating counterclockwise,clockwise, and all are underneath their coinciding coil, so A-B-C-D are on top, E-F-G-H are underneath.

Then, though all wire on a given spool is wound a certain direction in it's entirety, the approach and exit of the wires is different one circuit to the other. For instance,
continuity of coil (A) is wound outside to inside, as is all the coils in the red circuit.
While continuity of coil (E) is wound inside to outside, as is all coils in the blue circuit.
All coils have the same amount of turns, same direction of winding, and same layers as reported before.Larger ones-4 layers of 22 turns,, smaller ones 4 layers of 14 turns.

If you energize isolated circuit (1-4) with DC current coil (A) will attract negative compass and alternate polarity from one coil to the next.
If you energize isolated circuit (2-3) with DC current, coil (E) will attract positive( the oposite of (A) which it shares spool with) and alternate polarity from one coil to the next.
I checked this twice to be sure, cause,well I'm still a little a little confused...

Then when you add a jumper wire to (4) to (3), as in series both sides configuration(drawing 'D' page 10), and energize (1) and (2) with DC current, all the coils in circuit (2-3) reverse polarity, and each lug has the same polarity throught it's entire coil.

Disreguard my proposal for a rewire, till I figure this one out.

Bruce.

[DewCatTea-Bob]

" Here is the corrected scheme of things, as far as I can tell; "

____ Thanks for the updated & improved nicely well done depiction of the layout for the coil-spool arrangement, Bruce !


" If you energize isolated circuit (1-4) with DC current coil (A) will attract negative compass and alternate polarity from one coil to the next.
If you energize isolated circuit (2-3) with DC current, coil (E) will attract positive( the oposite of (A) which it shares spool with) and alternate polarity from one coil to the next.
I checked this twice to be sure, cause,well I'm still a little a little confused... "

____ This all seems to make proper sense Bruce but, we can't be sure since you haven't confirmed which ends of the stator-windings received the Positive or Negative outputs of your DC-source. ...
Let's say you connected your DC-source's Pos.lead to '1' , and it's Neg.lead to '4' , which in turn caused coil-A's outside-end to produce a North-field (thus attracting the South-pointer of your compose), and then of course coil-C is same, while coils B & D produce the opposite/South-field.
Then I assume you next may have connected your Pos.lead to '2' , and your Neg.lead to '3' , which in turn gave you just the opposite North/South-arrangement of what you had obtained with the 1-4 winding - (that's to say: S;N;S;N , instead of N;S;N;S as with the 1-4 winding).
In any case, I'm left to "assume" because you didn't make that very relevant detail clear - (that's to say that I'm left having to guess which winding-ends were connected to which polarities of your DC-source). _ As you (no-doubt) ought understand, it shouldn't be left to the no better than 50/50-chance, that your readers will happen to assume correctly on exactly-how you did your testing-connections, just as you actually had them.


" Then when you add a jumper wire to (4) to (3), as in series both sides configuration(drawing 'D' page 10), and energize (1) and (2) with DC current, all the coils in circuit (2-3) reverse polarity, and each lug has the same polarity throught it's entire coil.
Disreguard my proposal for a rewire, till I figure this one out. "

____ Again, we readers are left in the dark,, this time, as to whether or not stator-connection '1' remained connected to the Pos.lead (of your DC-source), and your Neg.lead moved from '4' over-to '2' ,, as I'm just happening to assume, (or perhaps some-other way-around ?).
__ I gather that what you're saying is that with '3' & '4' jump-connected, and your DC-source connected to '1' & '2',, all-four core-coils -(lugs & spools) then become ALL magnetic-North (or all South),, is that not what you meant to point-out ?
Assuming I've gathered that outcome correctly, I can understand how such a disordered result could make ya scratch your head. _ However, with the use of a mere hand-held magnetic-compass, such an indicated result is fairly understandable...
__ Not knowing for sure how you had the polarity of your DC-source connected to each stator-winding during your 1-4 and 2-3 single-winding tests, it's therefore left uncontested that this seemingly strange result may actually be fully sensible. _ In other-words, your single-winding DC pass-through tests, may or may not directly-correlate with your similar testing of both windings in series,, we just can't tell for sure ether way, cuz you didn't mention all the various polarity-connection relationships during each of your three tests.
But assuming that I've assumed every assumption correctly, then your mag.compass can't detect any polarity difference from one core-coil to the very-next, because the two (now more-obviously unequal !) stator-windings are producing both North AND South magnetic-fields, simultaneously ! _ And one of the two fields is just slightly stronger than the other (in all four core-coils), because the two opposite fields can't evenly cancel-out (since one winding is slightly longer).
__ Now if I've assumed something or another incorrectly, then please straighten-out the things I've had to assume, and then we can get to the bottom of any possible remaining contention.


Hopeful-Cheers,
-Bob

[wcorey]

Looking at the individual coils, I get the impression that the inner and outer windings are wound in opposite directions.
With the newly revealed 'all in/all out' configuration, when run in dual parallel 'stock' setup, the direction the current runs in would have two windings on the same coil/bobbin/core as opposite polarity. Run in series with the center leads ungrounded would result in the two being the same polarity. I'm getting a bit bleary-eyed but assuming this is even correct it could account for some of the performance differences we're seeing...

Update...
Preliminary results for the six coil mod seem to be about what was expected, a few head scratchers but more or less mirrors the four coil numbers plus the appropriate percentage of gain. I need to sleep so I'm not crunching the numbers right now but can throw out some rough numbers; 5ohm 3450rpm series was about 80w, 1ohm about 24w.
(odd results with the peak wattage at 3 ohm, will have to re-test) 5ohm 'Dual half wave' 'stock parallel' about 63w, 3ohm about 76w, 1ohm about 60w.
'One side' series 1ohm about 56w, 3ohm about about 72w, 5ohm about 60w.


Bill

[DewCatTea-Bob]

" Looking at the individual coils, I get the impression that the inner and outer windings are wound in opposite directions. "

____ Well not actually 'wound' in "opposite directions",
(as I believe Bruce, that none of the coil-turns, [of either winding on any single coil-spool], are turned in countering-directions),
but rather, 'ROUTE' their current-juice in 'opposite-directions' !
__ Seeing the latest depiction of the 8 coil-windings, now allows for correct diagnosis of the well-noted/strange hinder & help effects, noticed with the series & parallel arrangements.


" With the newly revealed 'all in/all out' configuration, when run in dual parallel 'stock' setup, the direction the current runs in would have two windings on the same coil/bobbin/core as opposite polarity. Run in series with the center leads ungrounded would result in the two being the same polarity. I'm getting a bit bleary-eyed but assuming this is even correct it could account for some of the performance differences we're seeing... "

____ I think you may have a valid thought in there Bill, but it seems your current chosen wording somewhat fails to get it properly conveyed, (just as you admit may be so).


" Update...
Preliminary results for the six coil mod seem to be about what was expected, "

____ I think this topic deserved it's very-own post, Bill !


" 5ohm 3450rpm series was about 80w, 1ohm about 24w. "

____ So just as before, the series-configuration/arrangement will supply fairly high wattage when the load's power-consumption is low,, but when the power-load demand gets high, then the series-connected alternator can't produce very much wattage to feed a demanding load.


" a few head scratchers
(odd results with the peak wattage at 3 ohm, will have to re-test) "

____ Actually Bill, it well stands to reason, cuz (as I've pointed-out before), the load-resistance must not fall below that of the (inductor-type)- power-source, or else the inductor-windings effectively become as part of the load.
Before, with the four coil-spools, the 1-ohm load was probably quite close to the stator's impedance,, but now that you've increased it's impedance (with the addition of the two extra coil-spools), the 1-ohm load must now be less (than that of the six-spool stator) !
So it now seems that a 2-ohm load test is called-for (so as to better determine max.power-output).
__ Another thing to become concerned with, is what the intended bike's load-resistance actually is,, because if it's TOO low, then the stator-winding will have to start dissipating it's own power, (as with a 'short'), which means it could then overheat !
So such concern could make-the-case for retaining more than just one stator-winding.


" 5ohm 'Dual half wave' 'stock parallel' about 63w, 3ohm about 76w, 1ohm about 60w. "

____ The lower power-output of the 1-ohm test, further confirms that the stator's new -(extended winding) impedance is higher than the 1-ohm.


" 'One side' series 1ohm about 56w, 3ohm about about 72w, 5ohm about 60w. "

____ How can just one side be "series" ?
__ I wonder just what we actually have here, as it doesn't seem right that one alt.winding by itself would produce almost as much as the (stock like) 'DUAL half-wave' setup.
I gather that this test is with full-wave rectification, rather than merely half-wave, right ?
You'll have to keep us filled-in on each of your particular rectification test-setups.

____ Thanks for the additional testing reports Bill !


Happy-Cheers,
-Bob

[ecurbruce]

Just a couple quick notes this AM, as I haven't had time to digest all that you wrote back yet...
I connected the DC positive to (1), neg. to (4). Then seperately, positive to (2) and neg. to (3).With jumper on (4) to (3), DC positive connected to (1) and DC neg. connected to (2). This configuration changes the direction of current flow in circuit (2-3) and changes the polarity of those coils to match their coinciding 91-40 outer coils. It's not hard to detect with a compass how dramatically it changes.
Also when circuit is energized, when I checked for DC voltage with meter at each junction of coils, in and out, the voltage showed a greater reading progressively towards the end of circuit.

More later,
Bruce.

[DewCatTea-Bob]

" I connected the DC positive to (1), neg. to (4). Then seperately, positive to (2) and neg. to (3).With jumper on (4) to (3), DC positive connected to (1) and DC neg. connected to (2). This configuration changes the direction of current flow in circuit (2-3) and changes the polarity of those coils to match their coinciding 91-40 outer coils. "

____ It's quite understandable that circuit '2-3' got it's polarity flipped-around oppositely in your series-test, (as compared to it's test alone)... as IT's polarity WOULD'VE remained unchanged in your series-test, (while the polarity of circuit '1-4' would've been the one to get flipped),, IF you had instead chosen to make the '1-4' circuit be the one to get tested backwards while in the std.series-arrangement, (as you had actually done with the '2-3' circuit, instead).
So no revelation with that particular observation.


YengYang-Cheers,
-Bob

[MotoMike]

Bruce

thanks much for that re-draw of the coil arrangement. Based on it, I would think your plan to re-wire each coil so that each set of coils sharing a common core could be placed in series and then putting all of the reconfigured coils in series, could be done to replicate a simple series wound alternator stator. On it's face unless I'm missing something, I would think you could connect the outer most connection of the inner coil to the inner most connection of the outer coil on each core. since the coils sharing a core are wound in the same direction, each two coil set on a core should then be the same as one continuously wound coil on a core. then the inner most connection on the inner most coil would go to the outer most connection on the coil on the next core. I am expecting that in this configuration you would have greater output of voltage, current and cosequently power than all the combinations we have tried so far.

It occurs to me that in this arrangement, we might not need the additional two sets of coils to make an adequit 12 volts system with a comfortable surplus. I know I'm going out on a limb here, but what the heck?

[ecurbruce]

Mike said- "I would think you could connect the outer most connection of the inner coil to the inner most connection of the outer coil on each core. since the coils sharing a core are wound in the same direction, each two coil set on a core should then be the same as one continuously wound coil on a core. then the inner most connection on the inner most coil would go to the outer most connection on the coil on the next core. I am expecting that in this configuration you would have greater output of voltage, current and cosequently power than all the combinations we have tried so far."

Mike, yes, that's what I was getting to in the proposal for a rewire (in that screwed up diagram), thanks for the support. I think it would work well. Now who's going to build and test the theory?

Mike said-"It occurs to me that in this arrangement, we might not need the additional two sets of coils to make an adequit 12 volts system with a comfortable surplus. I know I'm going out on a limb here, but what the heck?"

Yea, MIke sounds right, however, for my case, I already have the 6 coil spools set up, so guess I'll keep it and enjoy some excess.

Bruce.

[ecurbruce]

Bill,
I also would like to thank you for the 6 spool testing, it sounds promising right out of the gate!
Could I make a request right off the bat- a test of Mike's drawing 'E' from page 14? On your own time, of course, we're like a bunch of mad scientists late night in the basement (Grandpa Munster) lately, so just when you feel like it!
thanks,
Bruce.