Alternator forensics

[wcorey]

I really hesitate to stick my toe back into these waters for fear of again getting in over my head. I’ve forgotten many of the details of what I learned then and going back to that confusing thread to dig up particular info is not an easy or straightforward task…

I don’t believe we (collectively) ever arrived at any conclusion or consensus in regards to either the physical layout of this stator, or why it is the way it is. Only two of the four separate coils were actually dissected (one large and one small) and some of the results got a bit confused and sketchy. I know I was never convinced that any of our various drawings of such were completely accurate, specifically in regard to winding direction.
Coincidentally I seem to recall there being some sort of wiring difference/aberration in the second large coil which so happens to be the one with the ‘error’ in the drawing here, not sure if it applies to this particular issue or not. It makes logical sense that the coils should be consistent with each other but not everything about this alt ended up being so obvious.


On the ‘pull/pull’ effect, I think at one point early on the data was suggesting/supporting this but as I got my act together and the testing got more consistent the opposite eventually proved to be true (with the stock configuration). Bob, if you go back and read some related posts (starting near the bottom of page 14 and conditioning into the next couple pages or so), you’ll find that running a single winding set produced more output than both run together in the ‘stock parallel’ configuration (divided by 2).
As I said, my memory of the details of all this is somewhat vague and maybe at some later time we again found otherwise (or it could have been due to the impedance matching issue) but not that I can recall.

The following is a small snippet excerpted from ‘the alt mod thread’…

Note; ‘Full wave’, ‘stock parallel’, can be otherwise/also known as 'Dual half wave', 'stock parallel'...

Full wave, ‘stock parallel’, 5ohm 3450rpm-----------2.4a, 30.4vac, 12vdc,-----28.8w
Full wave, Series, 5ohm 3450rpm---------------------3.4a, 21.2vac, 17vdc------57.8w
Full wave, Single winding, 5ohm 3450rpm-----------2.3a, 14.7vac, 11.4vdc----26.2w



Bill

[DewCatTea-Bob]

." I really hesitate to stick my toe back into these waters "

____ Even-so Bill, you are certainly not unwelcome to contribute (as you wish) further !


" I don’t believe we (collectively) ever arrived at any conclusion or consensus "

____ I'm sure that if we -(the main posters of that extended thread) had all been in person together during the time, I would've then been able to get us all to fully realize what I-myself had concluded,, (although I must admit that I had to 'put 2 & 2 together' in order to reach my assumption).


" some of the results got a bit confused and sketchy. "

____ Earlier-on & within the thread, that kind of shakiness was of-course to be expected along the way, as that (much extended) thread was a post-as-ya-go & raw/unfully-baked type of thread, which never (yet) got sifted-through and condensed-down (into a more 'finished' thread, as has been intended).


" Coincidentally I seem to recall there being some sort of wiring difference/aberration in the second large coil which so happens to be the one with the ‘error’ in the drawing here, not sure if it applies to this particular issue or not. "

____ Well, of that which I've concluded must be an "error", has to do with the power-coil's INTERNAL coil-wiring, where-as (I assume) the "aberration" you refer-to was concerning the rather un-correlated location-point of an EXTERNAL wiring-connection/joint. - (Right ?) _ And if that's the correct case, then indeed it's just a (probable) coincidence (that both are concerning the same power-coil).


" On the ‘pull/pull’ effect, I think at one point early on the data was suggesting/supporting this "

____ Yes, (I'm fairly sure), as that was no-doubt when your testing was set-up to exactly reflect circuit-connections which were as the STOCK 'dual half-wave' setup.


" as I got my act together and the testing got more consistent the opposite eventually proved to be true (with the stock configuration). "

____ Well later, (when you got more serious with your testing), that was when you also began trying-out alternate circuit-types, and-so I'm thinking that what you're recalling MUST be when you were trying-out FULL-wave rectification of the two stator-circuits (rather than further-maintaining the 'dual half-wave' arrangement)...
Then at THAT time, you tried full-wave rectifying either single winding-circuit alone, and then with both connected-together in series,, and found (rather confusingly) that the TOTAL-output of the two winding-circuits combined-together, was actually LESS than the (naturally expected) double-amount of either winding-circuit working separately alone. _ In other-words, instead of X+X equaling '2X', your discovered-result was instead ONLY '1.5X', (or actually roughly the like).
And such a strange shortcoming probably stands to good-reason. _ Cuz (as I've recently indicated), when both windings are flowing current in BOTH AC-directions (due to FULL-wave rectification), THEN (due to the oppositely wound pair of stator-windings), OPPOSING flux-fields are concurrently created, (both at the very-SAME time) ! _ And-so then not only is the positive-aspect of the Pull/Pull-effect unable to become achieved, the effect may also tend to dampen otherwise normal current-flow... Cuz as the (regularly-occurring) collapsing flux-field (of the positive P/P-effect) TRIES to induce positive current-flow in the other-winding, that other-winding is THEN producing a negative current-flow (due to the F-W.rectification),, so thus-then some actual power-cancellation likely occurs.
And-ALSO, (due to the full-wave rectification), now that the negative-half of the alternator's available AC is also allowed to flow, there's then a (irregularly occurring) NEGATIVE Pull/Pull-effect as well ! _ And-so IT of-course also tends to cancel-out a portion of the (regularly occurring) positive current-flow (within both stator-windings).
So all this undesired opposition (to both AC directions) naturally leads to a (cancellation)- loss (of maybe up to 33% *) of the total power-output which should otherwise be available (if the two stator-windings were both coil-turned/wound in the SAME direction).
(* The ratio-relationship of power produced by the Pull/Pull-effect compared to the power produced by the alt.rotor, is likely a strange & un-obviously occurring varied ratio,, cuz the normally understood magnetic-rotor produced power-production is dependent on RPM, while the P/P-effect more directly depends on load-current to determine IT's particular amount of power-contribution.)
__ (I hope my added flow of info hasn't stolen away your thoughts too far from my suggestion that there's a probability that you've possibly confused the various test-results [between your 'half-wave' & 'FULL-wave' testing-trials of either & both stator-windings] ).


" if you go back and read some related posts,
you’ll find that running a single winding set produced more output than both run together in the ‘stock parallel’ configuration (divided by 2). "

____ I'm afraid I'll have to check-into what's-what with that, later.
If there was such an extreme discrepancy, I'd first suspect that it had more to do with unequaled impedance-matching issues.


" Note; ‘Full wave’, ‘stock parallel’, can be otherwise/also known as 'Dual half wave', 'stock parallel'... "

____ It's too bad that such extra term-wording got introduced, leading to confusion.
While "dual half-wave" may not be commonly-established terminology, it most clearly describes the particular case.
And it's a shame that someone who's been conventionally-trained would've felt the need to insert that the resulting combined-output of D.H-F is "full wave" and thus allow further confusion here at this w.site (just because the 'convention' doesn't bother to differentiate the actual-difference between virtual full-wave & REAL full-wave DC-output). _ Cuz the majority of members here have not been formally conventionally-trained and-so would naturally assume that 'full-wave output' (as seen on an oscilloscope) must logically be the result of 'full-wave rectification' (which provides both the positive AND negative halves of the AC-cycle [which is "REAL full-wave"] , rather-than combining-together two positive [180-degree out-of-phase] halves from separate half-wave outputs to create a "virtual full-wave" output.
So to avoid such non-logical confusion, it's best to call a 'spade' a "spade", and logically refer to the stock-system as "dual half-wave" !

Full wave, ‘stock parallel’, 5ohm 3450rpm-----------2.4a, 30.4vac, 12vdc,-----28.8w
Full wave, Series, 5ohm 3450rpm---------------------3.4a, 21.2vac, 17vdc------57.8w
Full wave, Single winding, 5ohm 3450rpm-----------2.3a, 14.7vac, 11.4vdc----26.2w

____ I'm left not realizing exactly why these particular test-results have been posted.
So can you explain what you had meant to be noticed here ? _ Cuz, as left as is, alone,, it's too confusing to be sure of what's-what, (cuz as-is, I can figure more than just one circuit-setup for each of the stated wording-labels).


Dukaddy-DUKEs,
-Bob

[Jordan]

I started to look at the anatomy of the alternator from a widecase, a Ducati one (not Motoplat), and not CDI. It's one with a damaged - almost completely missing - mounting plate, so a good candidate for a post-mortem.
It has 6 laminated core projections, all with windings on them. There are alternating 3 large and 3 small plastic bobbins, on which the wires are wound. Looks like the different sizes were to give extra room, by avoiding the adjacent bobbin flanges getting in each others' way.
One of the small bobbins was removed, by bending up a retaining tab, which is actually one of the steel laminations. I couldn't avoid cracking the plastic bobbin lower flange, when levering it up with a screwdriver. The bobbin isn't particularly tight, but the bent tab could only be straightened so much, and made it difficult to remove the plastic bobbin. This is not an item made for easy servicing, but at least there's no potting resin on the Ducati made alternators. The missing mounting plate made access easier. More damage might possibly have happened, if removing bobbin from a complete undamaged plate.
The are 2 separate windings, both in the same direction (I unwound them in a clockwise direction, looking from the top or outer edge), and connected at soldered terminations on an extension of the bottom bobbin flange. The windings are joined as series-adding, that is in a logical sense one joins up to the other and the windings keep going the same way. At the join is also connected the red covered wire that exits the alternator. The other ends of the windings connect to an adjacent coil.
Each winding is 24 turns around the core (48 total turns), of 1mm diam wire. In all, near 3.5 metres length. There was some surprisingly strong paper tape between the 2 windings. Although it doesn't look like it, there is some lacquer all over the wires, which were quite secure. The laminated core is a rectangle, 18mm x 11mm. The overall height of the bobbin is 15mm. I looked for some replacement bobbins (transformers use similar, in a limited range of sizes), but no luck yet.
Stay tuned for another exciting report, when I dismember the next coil!

Jordan

[DewCatTea-Bob]

" The are 2 separate windings, both in the same direction (I unwound them in a clockwise direction, looking from the top or outer edge), "

____ It would've been nice to have pictures to look at of this.
__ I would not expect a w-c.stator to have it's coil-turns wound any differently than that of a n-c.stator, so perhaps I was wrong to recall learning that the two separate windings (of a n-c.stator) were turn-wrapped oppositely from one-another.


" The windings are joined as series-adding, that is in a logical sense one joins up to the other and the windings keep going the same way. At the join is also connected the red covered wire that exits the alternator. "

____ If the two windings (of this particular dual power coil), are both wound in the same turns-direction with one of each of their ends connected to the red alt.wire-lead, then that's indeed same as a 'center-tap' type of connection,, however even-though that triple connection-point has the two separate coil-winding ends connected together, (thus APPEARING as if "series-adding"),, considering the way in which Ducati actually intended for current to flow, that triple connection-point is merely just a 'common' connection-point' that's convenient for the connection of the red 'common' alt.wire-lead. - (The red wire-lead is a "common" because it serves a dual-purpose, as it provides the same function of two separate wire-lead/current-flow paths [much the same as a 'ground' provides multiple current-paths].)
So this power-coil which you've chosen to examine first, is the one that happens to be the dual-coil which includes the particular terminal-ends (of both stator-windings) that were chosen to be terminated-together simply to make-use of a common-lead -(the red-wire). _ So-therefore, this particular power-coil's two windings weren't meant to be providing any "series-adding" current-flow,, rather, they are two separate winding-circuits which both happen to need to be connected to the red common-output wire-lead (so that both circuits can each provide their-own separate power-outputs for the charging-system).


" The other ends of the windings connect to an adjacent coil. "

____ I think your stated-wording must actually mean that each of the other/OPPOSITE-ends (of the two windings), are each separately connected to other separate adjacently-located/neighboring power-coils, (and not both connected to the very-same "adjacent coil"), correct ?


" Stay tuned for another exciting report, when I dismember the next coil! "

____ I, for one, am looking-forward to ANOTHER quite-nice/elaborately-detailed report of the sort !
I have hopes however, that you'll then also provide a picture of the particular power-coil, and make note of it's original-location on it's stator-plate.


Enlightened-Cheers,
-Bob

[Jordan]

__ I would not expect a w-c.stator to have it's coil-turns wound any differently than that of a n-c.stator, so perhaps I was wrong to recall learning that the two separate windings (of a n-c.stator) were turn-wrapped oppositely from one-another.

The bobbins would have been wound on a machine that spins them while wire is fed on. I don't see a need to change direction of rotation, which would be more work and complication, compared to just swapping over the wire ends if you want to change the current direction. Or am I missing something?

I dissected the next coil, a large one - same cross section but height = 22mm. The bobbin was easier to remove this time - minimal damage. This also has two separate windings, and 4 wire ends. One has 38 turns, the other 39 (a mistake?). They aren't joined to each other at the bobbin. Curiously, I unwound them both in the opposite direction from the first coil's - counter-clockwise this time.

Jordan

[ducwiz]

Hi jordan,

this http://minus.com/ltVZiOuvlCVZv is what I just found, stored deeply hidden on my PC, source unknown, sorry. Obviously, someone had analyzed the wdg. scheme and rewound the stator. Does this depict the situation you found?

Full wave, ‘stock parallel’, 5ohm 3450rpm-----------2.4a, 30.4vac, 12vdc,-----28.8w
Full wave, Series, 5ohm 3450rpm---------------------3.4a, 21.2vac, 17vdc------57.8w
Full wave, Single winding, 5ohm 3450rpm-----------2.3a, 14.7vac, 11.4vdc----26.2w

I made such measurements ~25 years ago, and from my more or less similar results designed and built my 12 Volt conversion for the w/c alternator/regulator http://ducwiz.minus.com/m1pfYbm8J/.

Some comments:
- I don't believe in any pull/pull effect, because there is no "collapse" of field or flux (like that happening in a battery ignition coil when the points open), roughly sinusoidal current/flux changes only.
- I do agree your assumption: the different bobbin sizes had been chosen to give extra room
- the reason why Ducati chose the dual half-wave rectification with common center in favour of the full wave bridge may be that simple: in those days when the system was designed - mid/end 60ies - silicon rectifier diodes as well as SCRs were still very expensive, but wire, cable and labour in particular much less. The amount of copper wire on the stator would have been (roughly) the same for both schemes, so their decision seems reasonable to me, cause they saved 2 power diodes, and an associated isolated mounting/heat sink.

Hans

[DewCatTea-Bob]

____ Well Jordan, I hope you realized that I was not done completing my previous-post when you submitted yours...
(My PC often likes to freeze-up when I'm trying to type my intended-posts, so never knowing just how long the typing-freeze will last, I click on the 'Submit' button [to post what I have thus-far], and soon-after go-through a reboot, before next-then getting-back to my post to try & get it finished.)



By: Jordan...
" I don't see a need to change direction of rotation, which would be more work and complication, compared to just swapping over the wire ends if you want to change the current direction. Or am I missing something? "

____ Right-now, I suppose all the various lead-ends of the individual coil-windings could certainly be arranged so that the circuit-connections between them will still allow positive-current to flow in opposite-directions -(that's in respect to the two yellow alt.wire-lead outputs), at the same time,, even-though the winding-turns are wrap-wound in the same direction. _ It would just make the physical-locations of the connection-points (between neighboring power-coils), a little less conveniently located (I think).


" I dissected the next coil, a large one -
This also has two separate windings, and 4 wire ends.
They aren't joined to each other at the bobbin. "

____ That's correct (in this power-coil's case), as it's two circuits need to be kept separate,, (except where their power-circuits finally come-together [down-stream] for common-connection to the red alt.wire-lead).


" Curiously, I unwound them both in the opposite direction from the first coil's - counter-clockwise this time. "

____ That's as it indeed has to be, so that the stator's neighboring power-coils will have their produced-current flow all in the very-same direction, as the neighboring magnets in the alt.rotor are of opposite polarity.

____ Thanks for this next w-c.stator-report of yours !
Do you think you could add a picture of exactly what you're working with though ?


Enlightened-Cheers,
-Bob

[DewCatTea-Bob]

By: ducwiz...
" - I don't believe in any pull/pull effect, because there is no "collapse" of field or flux (like that happening in a battery ignition coil when the points open), roughly sinusoidal current/flux changes only. "

____ Well Hans, I realize that you are one of us -(members here) who has an exceptional understanding of Ducati charging-systems and electrical-workings in general,, and the reason I had given the example of an ign.coil's functionality for helping to explain the "Pull/Pull-effect", is because I figured that most others would already be familiar with the induction-process of an ign.coil , (rather than that of transformers), since IT is taught to mechanics concerning ign.systems. _ And I agree that the term "collapse" is much more descriptive for what actually happens to the flux-field within an ign.coil,, HOWEVER, a sine-wave's half-cycle also has an eventual collapse as well, (all be it at a slower rate) !
And for the likes of yourself -(as having been trained in the field of electrics), I submit to you that the Pull/Pull-effect MUST occur, just as similarly as does the induction-process must do within a transformer !
So then, while the first-winding is passing a half-cycle of positive current-flow, it's 180-degrees of building & falling lines of flux (MUST) 'induce' a similar potential positive-power within the OTHER/second-winding. - (I know I've before only claimed just 90-degrees of resulting added power from the Pull/Pull-effect, but that's because that's all I could really expect from it, however actually since the P/P.process is more like that of a transformer (or an inductor-type circuit-isolator), then 180-degrees of additional P/P.generated power is what's likely actually made available (by the Pull/Pull-effect).
__ YOU, Hans, (if you understand transformer-coil function, [as I expect] ), should really understand that !
__ When I get time later, I'll then explain the whole-story so that most anyone can understand it.


" the reason why Ducati chose the dual half-wave rectification with common center in favour of the full wave bridge may be that simple: in those days when the system was designed - mid/end 60ies - silicon rectifier diodes as well as SCRs were still very expensive,
so their decision seems reasonable to me, cause they saved 2 power diodes, "

____ I really don't buy-into that explanation much at all, cuz for one fact, on the oldest DUCATIs, Ducati employed full-wave rectification with the employed charging-systems (apparently designed & built by CEV), and the associated finned selenium-rectifiers were not too expensive to employ (compared to the black-box later employed with their-own dual half-wave rect.setup).
__ I-myself think that Ducati was merely interested in getting the equivalent of 'full-wave' for the price of half-wave (so to speak), without concern for having to regulate raw full-wave power-output from the alternator, (as the contributing P/P.power doesn't need to be regulated since IT's power-output doesn't exist unless the load-system allows it's creation.
Isn't that totally-cool !? _ (If ya don't think so, then you're not fully-realizing that which I've been attempting to convey.)
____ I'll post a posting to convince you of the validity of the workings of the P/P.theory, later.


Dukaddy-DUKEs,
-Bob

[ecurbruce]

Bob,
__ I would not expect a w-c.stator to have it's coil-turns wound any differently than that of a n-c.stator, so perhaps I was wrong to recall learning that the two separate windings (of a n-c.stator) were turn-wrapped oppositely from one-another.

From N-C alternator modifications: discussion and testing, by Wcory currently on page 9 of ducati singles talk,
Windings direction of the narrow case alternator bobbins is discussed as being wound the same direction within each bobbin, and each consecutive bobbin wound opposite of the one next to it- just as they are in the wide case alternator (page 13, june 02, 2011).

Just to clarify, thanks, Bruce.

[DewCatTea-Bob]

" Windings direction of the narrow case alternator bobbins is discussed as being wound the same direction within each bobbin, and each consecutive bobbin wound opposite of the one next to it- "

____ OKAY, thanks for that Bruce !
__ It's now looking obvious as to what my 1-yearold memory had become confused with. _ But that mix-up is an understandable confusion-issue, as the posted-drawing of the stator-windings had strongly influenced my confused-conclusion ! _ (I still hate to realize that that drawing's depicted coil-windings are out-rightly about 50% wrong. _ I feel the reluctant duty to fix all it's drawn errors.)
I guess now, that the "error" I thought I had found (in the originally posted drawing), was actually the only equivalent-part which was really correct, (and all the rest of the OTHER indications of opposingly turned coils were wrong).
__ However that difference doesn't have any alteration/effect concerning the Pull/Pull-effect,
cuz so long as the two stator-windings are on the same alt.stator core-finger, (or even just the bobbin-coil itself), the induction-sharing can-NOT be non-existent !
____ So, the bottom-line here now is,
when I've previously thus-far/before indicated that the two sets of winding-coils were "wound oppositely", what I really should've been indicating, is that the tip/lead-ends of each individual coil-winding (of either stator-winding) are connected-together in the required-matter so as to cause the current-flow direction to be OPPOSED (respectfully between the two stator-windings).
So-therefore, the end-result is still the SAME (as if the two stator-windings had actually been WOUND oppositely, [rather than particularly-connected so as to create the same result] ).


Enlightened-Cheers,
DCT-Bob