The Triumph Dolomite Club - Discussion Forum

Here's what I have on the advance curves for the three types of 44D4 distributor used on the Sprint: 41402 (Type A), 41589 (Type B), and 41655 (Type C).



The minimum advance values are extracted from the 1980 edition of the Repair Operation Manual, as far as 2000 rpm at the crankshaft, and extrapolated to the limits set by the stop on the beak inside the centrifugal advance mechanism. The value for the knee and the reduced slope above that in the 41655's curve is estimated from a new set of springs in a good distributor, but may not be exact. The maximum advance values are also from the ROM up to 2000, and the estimates for the wear in the stop are mine. So, again, the upper limits may not be exact.

The limit on the advance for the early one, the 41402, came as a bit of a surprise. It's been suggested that the change to the next, 41589, was because of the loss of availability of 5 star leaded petrol at about the time the Sprint was launched. So I was expecting the 41402 to have a higher limit than the 41589, if a slightly lower rate of advance. But as can be seen, it's by far the lowest maximum value, and only the last, 41655, has a lower slope for advance at the lower end of the rpm range.

I'd certainly appreciate any views on what that means to performance and risks with lower octane fuel, e.g. 97 RON super-unleaded fuel.


I was also surprised at the 41402 being a bit different from the other two models, which, except for the stop value and spring rates, are the same. The method for holding the baseplate is using a sort of wedge-lock with a split in the plate with a screw in it; where the others use two screws into the body, like the 45D uses. I've seen this wedge lock before in a Lucas distributor, but it's not the 45D or the 25D. So I think I must have had one of these 41402's apart before and forgotten about it (there's a reason why my memory is so bad, but I've forgotten what it is).

Here are pictures of the baseplates of a 41402 and a 41589:





The baseplate mounting on the 41402 allows it to be fitted 90 degrees round, with the vacuum advance turned over, and so working the other way. I guess that's to allow it to be used where rotation is clockwise, instead of anticlockwise like the Sprint. It also makes the bolts that hold the distributor to the mounting plate much easier to get at from above.

Apart from the baseplate and advance limit value, there don't seem to be any obvious other differences to the mechanisms. So the later spring sets should fit, with a limited effect from such a small stop value. It should also be possible to swap the main shaft from a later one into the 41402 body, to give the same curves as either of the other two. I was considering doing that with one from a 41655, to get the higher stop with lower initial rate and much lower rate above about 3100 rpm. But seeing the set of curves, I wonder now if that's the right way to go.

Graham

Unless I have misunderstood/misinterpreted the advance curve for the 41402 you show is incorrect. I have attached the "calibration" sheet that I received with my reconditioned 41402D from the Distributor Doctor. It clearly shows a 3 stage advance topping out at a max of 10 degrees at 2500rpm (Distributor not crank figures) NOTE: rpm on bottom scale is distributor rpm so half crank rpm and degrees is distributor which I assume again is half crankshaft advance. (this does tie up with my copy of the R.O.M. AKM3629 2nd edition)




EDIT: just realised the number on the DD paperwork is 41202 have just checked and the distributor is definitely a 41402

marshman wrote: Unless I have misunderstood/misinterpreted the advance curve for the 41402 you show is incorrect. I have attached the "calibration" sheet that I received with my reconditioned 41402D from the Distributor Doctor. It clearly shows a 3 stage advance topping out at a max of 10 degrees at 2500rpm (Distributor not crank figures) NOTE: rpm on bottom scale is distributor rpm so half crank rpm and degrees is distributor which I assume again is half crankshaft advance. (this does tie up with my copy of the R.O.M. AKM3629 2nd edition)




EDIT: just realised the number on the DD paperwork is 41202 have just checked and the distributor is definitely a 41402

In some way I can't see all being correct either. The curve on the Distributor Doctor is a correct curve. The two stage are the two different springs in the centrifugal what every (Lucas) dizzy has. The small weaker spring already under tension from the start for the first rise and at a certain revs the loose sitting stronger spring gets under tension by the moving weights and slows down the advancing rise. Max advance is in the average around 3000 rpm and will be at a sprint around 10 degrees dizzy. Making that 20 crankshaft and in total adding 10 btdc giving figures around the 32 degrees max a Sprint must have before the high speed knock can appear.

The blue and the red line can't be advancing figures as no dizzy would have such a curve. The max advancing is too early at those and the red one would have a way too low maximum advance.
The green line does show a correct shape curve but the actual degrees would be totally off.

Jeroen

Interesting.

The curve I drew is based on the figures in the 1980 ROM, which given minimum and maximum at 1000 rpm as 6 and 8 - all for distributor. I added the limit from the 41402 I have, which has a stop marked as being 7 degrees. I inferred the lower value limit as being that, but I admit I could be wrong. However, if the mechanical stop in the mechanism is 7 degrees, there's no way the tolerance limit on the minimum for the advance curve could be any higher value - there must be tolerances in the shape of the weights and positions of the posts they turn on, etc., that would let them reach the stop before the cam has turned 7 degrees. So the 8 degrees for the limit on the minimum curve on the calibration sheet you got seems wrong to me. Saying that, I would think that seeing a real distributor that stopped at less than 7 degrees would surely be a surprise.

But it's possible that the 41402 that I have has had the top bit on the main shaft - the bit with the cam and the beak with the stop on it - replaced, or even the whole main shaft replaced. Though I don't know what it could have come from, as the 41589s I have are stamped 9 and the 41655s I have are stamped 12. So if anyone else has a 41402 they can look in (you have to take the baseplate out of my 41402 to see), it would be interesting to know what those have stamped on the beak.


As to the knee in the curves: There's an excellent and well worth reading article explaining how the springs manage the shape of the curve in "Tuning the Lucas Distributor" by Marcel Chichak of TDC Engineering. It only covers the 23/25D and 43/45D distributors, but the latter is close enough to the 44D for the descriptions to apply.

The 41402 I have has symmetric springs. In which case, there's no way it could give a knee. That should only happen if one of the springs is long enough that it's slack at 0 rpm - when they become asymmetric springs. Even if one was stretched to give a bit less preload than the other, it would only affect where the advance started from on the rpm axis, not give a knee in the curve, as both springs are still applied to their posts all the time. And the issue of stretched springs shouldn't apply to the minimum advance curve, only the maximum.

So I drew the minimum advance curve as going straight from where the force from the bob weights overcomes the preload in the springs, up to the limit from the stop in the mechanism. However, that is following what the Marcel Chichak paper says. So it's possible there's some bend in the curve from the geometry of the springs and weights that he's not allowed for. But it would be a continuous curve, not one with anything like an abrupt change of slope, like everybody shows an asymmetric set of springs giving.

Below is an asymmetric spring advance curve from the Marcel Chichak paper. I assume that it's okay to include this the under the fair dealing exception as ‘no more than is required by the specific purpose for which it is used’.


As you can see, he shows that curve as made up from what are essentially two straight slopes from the two springs, with a flat top, which comes from the limit of the stop in the mechanism. As the tiniest bit of criticism, he marks the upper slope as being from the secondary spring alone, though it is in fact from the combination of the primary and secondary. However, it is probably fair to say that the secondary normally dominates the primary in that region.

It's also possible, however, that the 41402 that I have could have been re-sprung, and had asymmetric springs before that. Or, the Doctor could have fitted yours with and allowed for asymmetric springs, if the wear in the stop on yours warranted: It is possible to weld up the stop and file back to the right value. But fitting a primary that is on its own equal to the two symmetric springs and a very strong secondary is nearly as good and a lot less expensive an option. That's more or less what I've done with the 41458 I have where the stop has been modified from 9 to 13 degrees. That should have symmetric springs, but I've fitted the asymmetric set for a 41655. As a result, it should run linear (at a slightly lower slope) up to about where the 41458's limit was, and then only advance slowly with increasing rpm to where a slightly worn 41655 would stop.

I guess we could only ever know if the 41402 originally had symmetric or asymmetric springs if we could find one that we know has never been reconditioned or modified, or got a set of springs to that original spec. But I do know that the springs the Distributor Doctor sells for the 41458 spec are indeed symmetric ones. So that can't have a knee in the curve. But also, the 41655 spec ones are very asymmetric. So that one does.

Graham

I've just taken one of the springs out of the 41402 I have and compared it with one of the 41458 springs the Distributor Doctor sold me. They are fairly similar, except that the 41402 spring is 0.83mm wire and the 41458 one is 0.75 mm.

Assuming the 41402 spring is original, that would give an advance with a lower slope than the 41458, as given by the data from the 1980 ROM, and since they are clearly a symmetric pair of springs, no knee in the curve; just as the set of springs for the 41458 will give.



Graham

And here's a pic showing the 7 on the beak stop and the symmetricality of the springs in my 41402. I even measured the two new 41458 springs from the Distributor Doctor to see they are also symmetrical, and they are. So there's no actual primary and secondary. So both apply from when the bob weights overcome the preload of the springs to where the bob weights meet the stop. And for both sets, there's will be no knee in the curve as the secondary comes tight - because there isn't one; just two primaries.

It also shows there's a bit of wear in the stop, but I will have to clean it a bit before I see if I can measure that.

I also need to get rid of the bath sealer that was adding to the hold on the baseplate. I do wonder how well that bath sealer isolated the baseplate from earth, and stopped the car from running. Not the best solution to the problem, methinks.



Graham

I had a word with Martin Jay at the Distributor Doctor, and he confirms that the two springs in a set for a 41402 are the same, i.e. symmetric; just as they are for the 41589. In which case will be no knee in the advance curves they have, and they will go, more or less straight to the stop at, nominally, 9 and 12 degrees respectively.

It's seems the springs in mine are a bit stronger than spec, being only 7 not 7.5 turns. But given there's some wear in the stop, I think I'll keep them for now.

Martin also described the secondary in the 41655 in a way that made me put the effort into checking the slope it gives. So I put together a spreadsheet to calculate rates and compare them assuming the same material for both springs. It turns out that secondary is about 10 times the strength of the primary, not the 2 or 3 times I thought. I hadn't realized that the spring rate goes with the 4th power of the wire diameter, not the square!

I also looked again at the differences between minimum and maximum, which is a nearly constant 4 degrees at the crank – 2 at the distributor. So I've re-jigged the curves, incorporating that stronger secondary for the 41655 and the nearly constant 4 degree difference between minimum and maximum into the allowance for the wear in the stop. The 41655 is still based on my estimate that the secondary comes tight at 9 degrees of advance, so it's worth allowing some leeway on that. But with that big a secondary spring and coming in that late, the 41655 won't get to the stop before the rev limit. But Martin did remark on the exceptional springing in the 41655.



Graham