From: "Paul Lipps" <elippse@sbcglobal.net>

Date: January 8, 2008 12:23:25 PM PST

To: lml@lancaironline.net

Subject: Ignition timing

As you say, anyone who wants to take the time could probably figure out the timing if they wanted to by parametric testing, but why?

Why?  I think mostly in reference to how the mixture is controlled, which is still manual.  For instance, if the "load advance" starts at 30" there will be a reduced detonation margin at 25" than if the load advance started there.  If the timing were fixed from 30 to 25 I would not hesitate to lean to best power at 25 inches.  If the timing were already advancing at that condition I would be reluctant to do that.  The flame speed goes up as the mixture is leaned from the full-rich position to best power.  If the timing is limited to a modest value and there were a failure that advanced the spark to the limit the pilot would be able to use more of the power - otherwise the manifold pressure would have to be limited to a lower value.  Without knowing the curve and not receiving any guidance the pilot is incapable of knowing how to react.

But what was listed as his timing curve was not correct

I've seen many posts like this before - the information from an initial post is labeled as "wrong", but no correct numbers are offered.  Should the initial poster just keep guessing until the information is no longer wrong?  I suggest if someone knows the right answer he post it.  I would just like to know the timing on my engine.

Do they want to build one of their own; if so, Klaus should keep it secret.

Many years ago GM and Ford would trade cars at the beginning of each model year.  It was easier to just trade than to make the competitor go buy one from a dealer.  Yes, one can make it inconvenient for the competitor to copy the design or the for the customer to know its characteristics, but that's all it does -make it inconvenient, not prevent it.

Do they possess the smarts to make an evaluation of the curve and assess whether it works properly?

That would be very difficult to do, but if I were building one I would keep the timing fixed until 25 inches and then advance it linearly up to 10 degrees at 20 inches.  The "speed advance" would be fixed except for starting.  If I had a more sophisticated system (microprocessor-based) I would keep the timing fixed until the rpm was over about 1700 so that run-ups could be done with a known timing.  Come to think of it, I have done that and it worked quite well.

Well, if so, publish what a correct timing curve should be,

See above, except that "correct" is a big word - any timing algorithm selected will necessarily be a collection of compromises.  There certainly is no one algorithm that is "correct."

and then measure Klaus' and show him what's wrong with his approach!.

I'm not sure there is anything wrong with his approach, but detail timing values might not satisfy everybody.

Why shouldn't a business hold secret those intellectual properties that they feel were the result of lots of work.

Yes, one can, but at the same time the value of holding information secret has to be balanced against the benefit of making them available.  For instance the approach of "I'm not going to tell you the timing, but I will sell you a device that will measure it."  Not a customer-friendly practice.

    Keep in mind, Klaus has an I/O on the unit which allows the user to change the timing as much as +/-7 deg. if desired; you can use it to come up with your own optimized timing vs MAP and rpm. And on a dual installation, there is a cross-connect which advances one unit when the other is off to regain most of the dual power; you can experience it when you do the pre-takeoff ignition check. The rpm will drop when one ignition is switched off, then climb back up after 1 or 2 seconds! See how your engine performs on just one mag. I've heard several reports of a mag failure causing the engine to overheat on just the remaining mag.

I think I would rather not have the interconnect, as a failure mode might affect both.  One reason to have independent systems is to keep them independent.  Yes, one could "over-advance" (or "correctly advance") one side of the other died, but then the engine would be operating with less detonation margin - is that important?  Don't know.  What happens if one side incorrectly went to the higher advance condition?

Gary Casey

ps: Slight correction to another post:  With the ignition retarded (like running on one mag compared to one EI) essentially the same portion of the fuel is burned in the cylinder.  With less advance more of the charge is burned later in the cycle so the expansion of the burned charge is less.  Less expansion means less cooling (and less energy being converted to useful work) so the exhaust temperature goes up.  Almost nothing is actually burning in the exhaust manifold - but enough to occasionally be visible.