Getting additional advance from a stock advance unit

[wcorey]

The subject of how much advance is needed to to get the best performance from various Singles has come up a number of times. There seems to be some consensus that the 'AA359' unit with 28 degrees of advance can in some instances fall a little short, particularly on the 450 with a start point at TDC (0 degrees). Even much more so with the 'AA367' 18 degree high dwell units that came on some battery-less models like the R/T. It can be a compromise between optimal top end and not breaking your foot from kickbacks. Someone (Bob?) was wondering if the stock unit could be manipulated to get more, I seem to recall 32 degrees as a good number to shoot for.

Since I have a 450 I have even more reason to need this, so I played around with removing material from the advance weights on an ‘undesirable’ 18 degree unit so I didn’t feel guilty about potentially messing it up, I also have a 28 degree unit for comparison.

When I initially put a degree wheel on them to get an idea of where I needed to get to I saw there was no way to get an additional 14+ degrees out of the 18. I was a bit confused when I then measured their existing travels as I only saw an indicated 9 and 14 degrees, respectively. Then I remembered that as they’re run off the valve train, the actual movement in degrees on the advance unit is half, the 9 and 14 translates to 18 and 28 at the crank. Now only having to obtain 6 or 7 (actual degrees on the advance unit) seemed much more within the realm of possibility…

Got it to 34 (crank) degrees without much problem and if needed could likely squeek another degree or two out of it. The limitation is mostly in the relationship of the weight pins to how they engage the forks on the points cam, go much further and the fork ends go past the mid point of the pin radius. By then the pins are also moving more away from the forks rather than across them, kind of a diminishing returns sort of thing. At one point the weights can also contact the timing cover unless you remove still more material.

Took the majority of material from the weights on the outside edge where they hit the stops but ended up at only a tiny bit more than 14 degrees (28-29 crank degrees). Was able to get the needed couple more degrees when I saw that ‘at rest’, the pins didn’t quite bottom out in the forks. The limiting factor was from the inside edge of the weights where they hit the base of the shaft the cam assembly rotates on and also each other where they touch nose to tail. Brought them in by removing some weight material in both spots, so I got a slightly earlier start point in addition to moving the end point forward.

I noticed that the 18 degree unit weights appear to be substantially heavier than the 28 degree ones. The weights are made in two layers of material, the 18 has both layers of material present all the way to the ends where the 28 has only about 2/3’s of the top layer and the void is on the outer ends where it really counts. Leads me to assume the 18 would start opening sooner. Maybe the heavier weights are intended to start the advance sooner due to starting at 0 degrees?

The springs look to be identical on both so I think the loss of inertial material on the 18 will just even things up and result in a similar advance rate as the 28. I may mount both up on a drill or something and see if I can determine relative to each other when they begin opening. The start point of the spring tension is less after the mod and I guess I could also stretch the springs a bit to compensate but I suspect I’m splitting hairs.
In hindsight I should have weighed the weights before the modification…



The 28 degree is on the left, modified 18 on the right (I know, no clips...). You can see that the springs on the unmodified 28 are slightly extended where the 18's are fully retracted after having brought the weights closer together. If you really look closely you can see that on the 18, the cam pivot pins on the weight line up to the pins the weights pivot on, where on the 28 the pins don't quite line up (they start later).





Modified 18 fully opened...




Side view where the difference in the weight thickness can be seen...




Unfortunately I seem to have misplaced the ‘before’ photo’s but the places to remove material are fairly obvious…


Bill

[DewCatTea-Bob]

____ Very fine report Bill ! _ Thanks very much for your great effort !
But wonder why you referred to the two units as '28' & "14", when they're actually claimed to be '28' & '18' !?
If you were just mistaking, then please replace all your "14"s with 18s, (or allow me to do so for you). _ Also, I'd like to see your pix be fully see-able, so if you'll give me permission, I'll then take care of that for you.

____ I very much agree that even the '28-degree' unit provides less than optimum advance-range ! _ However I know that there are those who be concerned that without weakening the springs to go along with the lightened fly-weights, that the stock 'advance-curve' & it's RPM-range will be considerably changed by this modification.
To that concern, I have a bit to say...
__ It's my opinion that the exact 'curve' is not important for more than a small-fraction of a horse-power within the advancer-range. _ I think the main-reason for an advancer is to provide ease of kick-starting,, and once the engine has been started, the ign.timing then might as well jump to full-advance. _ So thus the 'rate' of the ign.advance is really of no concern.
__ Next, (after the concern of advance-rate), is the concern of at what RPM the max.advance is reached...
If the stock advancer reaches full-advance at 3000-RPM - (it's fairly doubtful that it indeed does so at exactly that point!), then with the lightened fly-weights (& stock springs), we understand that the full-advance point will then be at some higher RPM than 3000 ! ...
__ I'll next refer to the 450 for an example here, because of it's convenient (for thinking) 0-degree starting-point.
The 450's best max.advance point (for best power-output) is 32-degrees BTDC, yet as stock, it only gets up to 28-degrees BTDC (around 3000-RPM),, so with this modified advancer-unit's (extended) 34-degree range, and still starting-out at 0-degrees, we should reach 23-degrees BTDC around the 3000-RPM and the 28-degree BTDC point at (calculated somewhere around) 3650-RPM, and reach the 34-degrees BTDC point, at about 4400-RPM.
While the first two points are obviously retarded from stock-figures and may indeed cause a slight loss in power under 3650-RPM, the needed & obtained additional degrees of advance will conveniently come into play after 3650-RPM !
__ I would not choose to use this modified AAUnit on a 450 cuz I was told by a much trusted & highly knowledgeable source that the 450s produce their best power with their static-timing set at 4-degrees BTDC, (as opposed to my own belief that 9-degrees BTDC ought to be best). _ I'm sure that Ducati recommends the "0" degree static point pretty-much just for starting purposes.
However for a 350 or 250, the extra 6-degrees of total advance-range is especially helpful, cuz then a 350 engine could have a 37-degree* max.advance with a static-setting of just 3-degrees BTDC, and a 250 could have a 39-degree* max.advance with a 5-degree BTDC static-setting. - (*Optimum figures for best power.)


Dukaddy-DUKEs,
-Bob

[wcorey]

That's funny, I wrote the whole thing up as 18, then was looking something up in another post and it was referred to as a 14.
I thought, 'well considering I just did it the other day I must be having a brain meltdown then' and changed the whole thing, I guess I'm still sane after all...

Viola! done. I shouldn't be so quick to second guess myself.

If the stock advancer reaches full-advance at 3000-RPM - (it's fairly doubtful that it indeed does so at exactly that point!), then with the lightened fly-weights (& stock springs), we understand that the full-advance point will then be at some higher RPM than 3000 ! ...

As I mentioned in the original post, the weights on the 18 (and also the proportional placement of such) appear to be heavier than the 28'. Unfortunately I neglected to weigh them before the modification but you've now got me digging out my digital reloading scale. The 28 weights are 9.94 grams, the modified 18 weights are 10.04 grams.
With both the springs and the geometry of each being identical it seems the curve will now be the same whereas before the mod the 18 (with the heavier weights) would have opened sooner. I also conjectured in my previous ramblings that maybe this was done intentionally to compensate for the later 0 degree start point on the 450's and that I could stretch the springs (or bend the supports inward) to make up for it on the modified version.
All of which may be moot in my case as I may end up using a 're purposed' Crane ignition module with built in adjustable advance curve (originally from a Harley, lol).

__ I would not choose to use this modified AAUnit on a 450...

Assuming I end up using it, I've already planned on making it 'adjustable' with the addition of a set screw in each weight that will allow the weights to be moved to they're previous locations or beyond. The little things to do just keep piling up...

[Bevel bob]

Very good, for my 250 it seems that about 5 degrees static and 36 advanced would be ideal so I'm looking for about 3 degrees ( one and a half on the unit) , I'm encouraged by your efforts to give it a go.

[wcorey]

...so I'm looking for about 3 degrees ( one and a half on the unit)...

Bevel bob,

Here you go...

In looking closer at my 28 degree unit, it appears that not much more can be had from opening the weights further, it’s already at it’s optimal point and has reached the aforementioned ‘point of diminishing returns’. Even worse, my particular unit is shy by one degree, reading in at 27.

There is still the other direction to go in, however that too is not as extremely placed as the 18 unit originally was so there is also a smaller amount to be gained there. Luckily you only require about that much because that’s about all you may get….

So I decided to go for it on my 28 as it is such a tiny modification and I got overly anal about the implementation so it’s virtually unnoticeable anyway. Did the material removal only on the 'nose to tail' interface(on the 'tail' end of the weight), and just barely got 1 ½ more degrees, probably closer to a 1 1/3.

The unadulterated unit minus cam…




The mod-ed unit…




The weight modification… It would be much easier, recommended and equally effective to grind the cutaway all the way through both layers instead of halfway as seen here. On mine, this is where I got carried away with senseless cosmetic details...



It was necessary to bend the spring holding tabs a bit (clockwise as viewed from the top) to hold the weights into they’re newly angled position.

Pictured here in a side view…


I’d guess that another ½ degree could be obtained by relieving the area on the weights where they contact around the central shaft (that the cam rotates on), then a small area on the bottom of the cam unit itself were the weight spring holding pins hit the protruding ledge.
To get still more then the cam fork slots could be deepened and weight pins that sit in (and move) the cam forks would need a flat on the inside of the top shoulder were they would then contact the cam. All simply done but I’m just not going there with this one.
Hope my terminology isn’t too confusing, trying not to use whole sentences to describe the individual bits…

I’m not sure why the 18 was easier to get more out of, considering it starts with less, I have too much time into this already and am not planning on spending more to figure that out. Could be by design or just random variances in the manufacturing process, we should all be used to that, lol.

I put the cam from the 28 onto the 18 and still retained the better degree number gotten there, so if you have both by all means modify the otherwise useless 18 and borrow the better cam from the 28.

My next write up will be on using HEI ignition modules, where points are one option for triggering. In that instance dwell time is meaningless so the ‘smaller’ cam on the 18 is actually a slight advantage as the points are open for less time so the contact heel spends less time dragging across the cam than on the 28’s ‘larger’ cam, reducing 'gap altering' wear.

[DewCatTea-Bob]

" As I mentioned in the original post, the weights on the 18 (and also the proportional placement of such) appear to be heavier than the 28'.
With the both the springs and the geometry of each being identical it seems the curve will now be the same whereas before the mod the 18 (with the heavier weights) would have opened sooner. I also conjectured in my previous ramblings that maybe this was done intentionally to compensate for the later 0 degree start point on the 450's and that I could stretch the springs (or bend the supports inward) to make up for it on the modified version. "

____ I had planned on getting-around to commenting on those "previous ramblings", but since you've brought them up again here, I'll just go-ahead & respond here.....
__ I'm sure the fly-weights on the 18-degree AAU are heavier but, not for the reason you assume...
I believe that Ducati was more interested in cutting costs than producing AAUs that are DUKE-model specific...
I assume that your '18' AAU came from your 450R/T and that's why you think it has something to do with the '0-degree' static-timing figure (which is actually the correct factory-spec figure for the battery-powered 450-models with the 28-degree AAU).
What you don't seem aware of is that the 18-AAU was originally designed for the n-c 250 Scr & Mk3 models which have 21-degree BTDC static-timing settings and thus need the heavier fly-weights to get the ign.advance fully extended within the narrower range-span by the 40-degree max.advance point of those 250-models.
And I believe the static-advance for the 450R/T is to be set at 9 to 12 degrees BTDC, (not zer0 like the other 450-models).
__ It works-out cool that your modified hybrid AAU fortunately starts-out with the heavier fly-weights, so that after they're lightened, they end-up closer to the weight of those of the 28-AAU !_ (So you don't have to be concerned with altering the spring-strength!).


Cool-Cheers,
-Bob

[Bevel bob]

This thread has got me thinking(danger,danger!) the problems i would like to solve are , the cough and stop at "tickover" and kickback .I intend to retain the SSi which is part of the problem.I suspect that the AAU does not hold the timing back at what i call tickover (joke) and if anything, less slack on the springs (bend the posts?) may help.Even better would be a pointless system with a programmable (by me) advance range.But back to reality ,an AAU with a wider range and slightly slower start could be the answer.

[DewCatTea-Bob]

By: wcorey...
" In looking closer at my 28 degree unit, it appears that not much more can be had from opening the weights further, it’s already at it’s optimal point and has reached the aforementioned ‘point of diminishing returns’. "

____ When I last wondered about this, I just assumed that fixing it so that the fly-weights could swing-out further, would provide the desired additional degree-range,, and I didn't think of also increasing the overall advancer-range from the other-end (by removing materials that would allow the weights to further close-inward), thus obtaining a further retarded starting-point, so as to better obtain an increased-range and in a more functional manor as well.
__ Before reading your finely-detailed report, I wondered why not consider flattening-out the fly-weight stop-tabs and creating new stop-tabs on top of the tip-ends of the flattened-out tabs.
But I now gather that you'd point-out that the "diminishing returns" situation would make that method unreasonable,, is that correct ?


Fine-Cheers,
-Bob

[wcorey]

__ Before reading your finely-detailed report, I wondered why not consider flattening-out the fly-weight stop-tabs and creating new stop-tabs on top of the tip-ends of the flattened-out tabs.
But I now gather that you'd point-out that the "diminishing returns" situation would make that method unreasonable,, is that correct ?

Yes, correct... You could just bend them out a bit, being mindful that at one point they will contact the timing cover.

What you don't seem aware of is that the 18-AAU was originally designed for the n-c 250 Scr & Mk3 models which have 21-degree BTDC static-timing settings and thus need the heavier fly-weights to get the ign.advance fully extended within the narrower range-span by the 40-degree max.advance point of those 250-models.

I should have figured as much, but most of my firsthand knowledge of Ducati singles is confined to the 450 R/T that I own. I read a ton about other models but don't tend to retain a lot without first putting it to practical use. I'm much more in my depth with Triumphs/Nortons or Rubberband Ducati's...

And I believe the static-advance for the 450R/T is to be set at 9 to 12 degrees BTDC, (not zer0 like the other 450-models).

Is it correct to assume though that with a different ignition setup it would go back to the zero setting?

[MotoMike]

what a great thread. It lays out in very well detailed language and with very good photos all the questions I had myself pondered regarding these units. I had not however considered removing material to back up the initial starting point. That seems like the way to go to me if not employing all the mods. Is there concern that the locating pins in the forks are about to slip clear if we go any futher out towards the stop tabs?

Thanks Bill for a thread that I have really enjoyed and Bob's thanks too for the additional input and photos. This is really quality stuff.

Regards,
Mike