Sherwood,

You still have it backward. Static the rotor points at the wire. At say 5000RPM the computer fires the spark at 30 degrees advanced. If the rotor didn't advance as well it would be short of the wire by 15 degrees.

-Andy

Yes.

The distributor rotates clockwise and weights move the rotor additionally clockwise as well. The 3.2 distributor is a mechanically retarding distributor (when viewed as a conventional distributor) , thus keeping the rotor lined up to the terminal as timing advance increases with rpm.

Have more faith in Porsche designs ! :)


EDIT: Nevermind !


.

Sherwood, the advance mechanism doesn't move the rotor tip away from the contact. It moves it TOWARD it at the point in time in which the DME fires the coil.

Bosch distributors in Porsches, with one brief exception, rotate clockwise. When the distributor shaft speeds up, the advance weights move the rotor tip clockwise. The rotor tip ends up being "ahead" of the position it would be in were there no advance. This is a good thing because the ECU will fire the coil when the advanced rotor tip passes under the coil contact.

How does the rotor move toward the contact if it started out opposite the contact in the first place? If the rotor "advances" or moves due to the centrifugal weight mechanism, where does the rotor tip go if not a place other than where it started?

If the ECU recognizes the rotor location at a specific rpm, the ECU must calculate the position of the independently moving rotor as the weights move outward to an advanced position, while still discharging the spark to compensate for the rotor's relative movement. IOW, the centrifugal advance mechanism AFAIK, is superfluous for DME-controlled ignition.

Unless an advancing rotor serves some other purpose, it seems to me it could simply be locked in a static position, then allow the ECU to control the spark per the relationship between the piston and TDC just as modern ignition systems have evolved. To identify TDC for no. 1 piston (reference cylinder), there's a crank position sensor for the flywheel to accomplish this.

Sherwood

The distributor does not know or affect at what time the sparks travel through it.

At static timing the rotor is lined up with the terminal.

As the rpm increase the dme sends spark 30 degrees sooner (15 at the dist). If the rotor was fixed it would be 15 degrees off (before) the terminal at that time.

The advance weights move the rotor 15 degrees further in the direction of rotation so that when the earlier spark is sent the rotor is already there.


The 3.2 distributor is purely mechanical and the dme (ecu) has no idea what is happening inside it.


.

The rotor is merely a current path for secondary current. The voltage discharge is created by the computer and does so independently of the rotor. Obviously, a 15º difference in angle between an advanced rotor and its contact doesn't make a significant difference in engine operation and as demonstrated by later distributors with fixed rotors.

I may still be missing something, so please let me know.

Sherwood

Later distributors with fixed rotors are much larger to seperate the terminals as mentioned earlier. The 3.2 dist terminals were deemed too close together to have a fixed rotor, apparently.

Whether fixed or not, the rotor has nothing to do with when the spark is discharged through the secondary system. When points were used to signal the spark, the rotor assembly includes the 6 cam lobes that also controls when the contact points open. Thus, when distributor speed increases, the contacts open sooner with the rotor in a more "advanced" position, and so on.

In a DME system, there are no points, thus control of the spark is performed entirely by the computer independently of the rotor position. With a known reference point, the TDC sensor and other inputs (engine speed, coolant temperature, throttle position, etc.), a predetermined computer map controls the occurrence of the spark for each revolution of the distributor (2 crank rotations) regardless of the rotor position, either in direct alignment with the dist. contact or a few degrees further away. The rotor does not have to "catch up" with anything. If the maximum spark advance by the computer is 30ª BTDC, the maximum distance from the dist. contact is 15º.

I assume a larger distributor cap provides more physical separation between the 6 spark plug contacts to reduce crossfiring as the rotor conducts the "spark" to the appropriate contact.

Sherwood

Well you got one part right-- in a DME triggered ignition there is no longer a cam in the distributor to tell the ignition when to fire. Nor is there a reluctor or hall sensor in there either. Porsche moved the crank position sensing to the flywheel in '84 with two sensors and a magnet. Later iterations use missing teeth. But in the Motronic cars that still use distributors, the rotor phasing mechanism is still present.

How does the rotor tip move to a place other than where it started? That's what the advance/phasing mechanism does. As RPM increases, the rotor tip moves ahead of where it started, so that when the spark fires, it's under the contact. If it didn't move, it would still be in its original position relative to the distributor shaft (and by extension, the crank) but it would be passing under the contact some amount of time after the coil had fired.

I don't think you can dismiss the inclusion of rotor phasing as "immaterial" when it was used for 13 years across three different iterations of the design.

First, we know that from a theoretical standpoint, with the small caps required to fit behind the heater duct, there's a danger of crossfire between the rotor and the adjacent contact if the distributor is locked. As you are aware, Paschen's law tells us that the breakdown voltage between parallel plates is a function of the pressure of the gas they sit in and the distance between the plates. Leave the pressure discussion for when we talk about spark plugs: it's the distance that matters.

At a given spark voltage, putting the distributor contacts closer to the axis of the distributor shaft will place them closer together, 2 * Pi * R. So if you make the cap bigger, you can put the points further apart. What you are counting on in this situation is the absolute distance between the rotor tip and the adjacent contact being sufficiently great to inhibit a spark to that contact instead of the one it's supposed to hit.

If we're down to speculation, as against the idea that the advance mechanism was included by the same engineers who gave us the five blade fan and the cup-type tensioner, I speculate that concerns over reduction in the number of moving parts first and cost of goods second would have sent rotor phasing the way of the carburetor at the earliest opportunity were it not justified by necessity.

A couple Pelicans have tried locking their distributors to run the MSD-6AL-digital.

http://forums.pelicanparts.com/porsche-911-technical-forum/485288-msd-6al-2-programmable-ignition.html

http://forums.pelicanparts.com/porsche-911-technical-forum/500866-msd6al-2-locked-distributor.html

from the Bosch manual on Motronic

subsection on rotating voltage distribution
"Reliable high voltage distribution can only be guaranteed within a limited dwell angle range which is inversely proportional to the number of cylinders. Centrifugal rotor adjustment provides adequate range extension on 6 cylinder engines, but 8 cylinder systems usually need two 4 cylinder systems"

"A couple Pelicans have tried locking their distributors to run the MSD-6AL-digital....."

It appears these owners were using distributor-based points or reluctor to trigger the spark. These distributor-based signals are all dependent on the advance mechanism of the distributor, thus "phasing" must exist between signal and rotor. If the signal is performed externally, rotor "phasing" is no longer an issue.

Sherwood

"At static timing the rotor is lined up with the terminal."

Actually not! Some 3.2s I've worked on had that setup which resulted in mis-fires
at higher RPMs, e.g. ~ 4K, and my shop (customer) tried replacing everything,
DME ECM, wires, AFM, etc. When I checked the rotor's initial position, it was set
directly on the terminal pole.

For proper ignition advance over the full RPM range, the rotor needs to position
itself a little before the terminal initially and then a little after the terminal at
the higher RPM. If it's set (the rotor) too far 'back' (retarded) initially, then spark
jumps to the prior cylinder. If it's set right on the terminal initially, then the
spark will eventually jump to the next terminal at the higher RPM.

"but 8 cylinder systems usually need two 4 cylinder systems"

Or a huge distributor & cap as used by Mercedes on the 8 cylinder engines.
As mentioned in this thread, the limitation is the size one can make a distributor
& cap when using a fixed rotor.

In a perfect world perhaps , or at least is less of an issue.

I don't think anyone here thinks the engine won't work w/o the advancing mechanism. Nor does it appear that anyone thinks the spark distributor has anything to do with triggering or timing in any way.

The question was why the mechanical advance was there at all.

My guess is that porsche wanted the system to be as bulletproof as possible even if the car was being operated at max load in the pouring rain and had 150k on all the ignition components.

I guess it happens with some set ups. How do you set static timing so the rotor position is a little shy of the contact? By marking the rotor tip visually on the housing at TDC, then rotating the housing accordingly (1/2 timing advance spec) with the cap installed?

Sherwood

An article on rotor phasing:

rotorphasing

Two rotors. Notice the width of the contact tip on both:
http://forums.pelicanparts.com/uploa...1320705663.jpg

vs.
http://forums.pelicanparts.com/uploa...1320705734.jpg

The former can "reach" closer to the dist. contact as it advances past the static timing position.

As the article mentions, one can split the advance difference and position the rotor accordingly so the rotor tip doesn't venture too far from the contact.

Please note: The above only applies to ignition control outside the distributor (i.e. computer or equivalent).

Sherwood

"My guess is that porsche wanted the system to be as bulletproof as possible even if the car was being operated at max load in the pouring rain and had 150k on all the ignition components."

As it applies to the KISS principle, an unneeded mechanical advance mechanism introduces more complexity where it's not essential.

Sherwood

Hi All,
reading through the replies and discusions here has been very interesting and enlightening in more ways than one, however, i cant help wondering if some folk are just defending porsche's decision to still incorporate these items [for what ever reason] without looking at the arguement from both sides, and realy thinking about which scenario would be most consistently affective,accurate, long term.
Anthony.

Maybe Porsche did not want to spend money creating a new convention. As far as the advance components go, It looks as though Bosch supplied them generic components from their most commonly used parts lines reducing the overall cost. Perhaps there was quite a bit of inventory left @ Bosch that mandated the utilization of the remaining inventory(housings,shafts,etc.

HI Aaron,
Perish the thought ;)
But seriously, it does make you think...
Anthony.