m42racer

For sure the Timing belt introduces some error in the Ignition timing, but within the typical 944 application, the Triggering at the Cam is still a good choice. Absolute accuracy would demand the Triggering from the Crank, and any sync on the Cam. A sync pulse does not have to be super accurate, as it usually can happen within a certain window of Crank degrees. The Cam trigger was done for ease of application, where the absolute accuracy is not required. Even then, accuarcy is help by teeth that have a radial dimension greater than any Cam movement caused by the belt flex. Any belt flex that caused any error would first cause an engine misfire from a triggering error, not Ignition timing error. The trigger wheel I saw has 23 teeth on it, with the 24th missing as the Cam sync. The EFI system reads these 23 teeth and any misread would result in a trigger failure which would result in a misfire. The system would simply miscount. Once the system has counted the teeth correctly, the system does the math and fires the Ignition based upon the number of cylinders programmed at the right amount of tooth count. If there is no miscounting of trigger teeth, resulting in no misfires due to a triggering error, then the Ignition will be as accurate as the EFI systems software. There is a lot of mystery surrounding this topic. The term spark scatter has been suggested here caused by the Timing belt flex. Distributors which are gear driven of the Crank suffer from this also. Most common cause of this is cross firing within the cap due to incorrect rotor phasing. Many times where very powerful Ignition systems are used, the Distributor Cap size is increased to help eliminate this problem.
Any 944T application will not suffer from the trigger driven off the Cam. The bigger the teeth, the less likely the problem and any problem will show up first as a trigger problem. Hope this answers some of your questions.

NZ951

Ok did anyone read the part about the hand puppets?

Chris White

Ok, here is some objective info on the crank trigger vs cam trigger.
Cam triggers have 3 inherent problems –
1) belt stretch / flex – this can result in a loss of accuracy in the timing signal since the trigger wheel is not indicating the crank position accurately. Not a huge error but it can be noticeable.
2) Rotation speed – the Cam rotates at half the speed of the crank, effectively reducing the accuracy by half.
3) Smaller diameter trigger – to fit within the confines of the cam area the diameter of the trigger has to be smaller, this also results in a higher possibility of error.

So add those up and we have cut the timing accuracy by at least 50%. I am interested in the previous posts that expound the more power has to be better but then seems to have the position that accuracy greater than 1 degree is not really needed. I believe the exact opposite. If the plug will fire correctly in all conditions then more power to the spark will not yield a significant difference.

(insert the sound of the sock hand puppet coming off)

Now to the interesting stuff – the issue of accuracy. Looking at the system as a static device (no motion) you can easily convince your self that the trigger wheel can be pretty darn accurate. The interesting part is how the timing get calculated. For the sake of round numbers lets figure an engine can accelerate form 3000 to 6000 (unloaded) in ½ a second. At 3k rpm the time for one revolution is .002 seconds or 50rps (revs per second).
At 6k rpm the time for one revolution is .001 seconds or 100rps. In the space of that one second the average rpm was 4500 which means that during that 1/2 second the engine made 37.5 revolutions. A 3000 rpm change (3k to 6k rpm) in 37.5 revolutions equals a change of 80rpm per revolution.

So what does that mean? If your system (trigger wheel plus software algorithms) can’t accurately cope with a dynamically changing environment the result is an inaccurate spark. When you do that math you see that the in one revolution of the engine the rpm can change by 80rpm. That’s a change of 2.67% in one revolution of the crank at 3k rpm. Some systems rely on the TDC signal to perform the calculation – those systems will have accuracy problems in very dynamic operating conditions. BTW – using a cam trigger doubles the delta – the cam rotates once per two crank rotations. So that 2.67% error is now compounded to 5.33%.

Before we go there – I am not saying that this is how a Link systems works, I don’t know how the spark timing algorithm is set up for those units (according to the teeth counting description in a previous post the Link does calculate the spark timing that way). I do know that there are quite a few set ups that do rely strictly on the TDC signal for timing calculation. The Electromotive system uses all the pluses (60 tooth wheel) for ‘on the fly’ timing calculations and in fact they have the paten on that design. Many folks, including Porsche, are now using the 60-2 concept (the S2 and 968 engines are all factory 60 tooth triggers on the flywheel). So the Electromotive design ‘sees’ the position of the crank every 6 degrees and will calculate the change in rpm at each tooth – resulting in a max error of .044% in the scenario we just calculated.

Back to spark power – once you get past the point of igniting the mixture properly I don’t see where the more is better argument applies – especially at the cost involved. 150mj of spark is a lot of spark power (tec3 spark output). In the real world I have never had a spark related problem associated with a Tec3 (other than a faulty spark plug). In fact I have a particular customer with a 930 that has just the spark side of the Tec3 hooked up and his CIS system has been malfunctioning – he has been running track events at 1 bar of boost and a air/ fuel ration of under 9.5:1. He did melt his rear bumper but the engine never missed a beat for the last 10 track days. (I have tried to get him to fix it but he likes driving it too much!). Its so rich that you eyes water any where near it, but it keep running!

I will stand on my original point – if an ignition system has enough spark power to perform flawlessly why do you need more? If it was a no cost option with no down side I would agree, more is better. But if you are saying that more spark power will always result in better performance then I have to disagree on that point, especially at the costs involved.

Conclusion – look for accuracy in your ignition!

Chris White

PS – If anybody absolutely insists on it I can set up a Tec3 to run a CDI set up. All it takes is a couple of extra handfuls of money. Oh, yeah – I will also give you a written guarantee that there will be no additional rwhp due to the addition of the CDI set up.

macnewma

This is awesome guys. I am really learning something from all of this. I wish Rennlist had a special area for the more advanced topics.

At the risk of sounding ignorant (there is a first for everything ), don't MSD and Crane claim to have CDI setups? Those aren't particularly expensive. Am comparing apples to apples here?

Max

Chris White

CDI vs Inductive is just about how the energy for the spark is created. CDI is not that exotic – it is standard in many vehicles, boats and such.

The key thing is that you need to control what the CDI does – many of the aftermarket folks (crane, MSD and such) can supply a CDI box that takes the place of the conventional coil. Other systems end up being a ‘stand alone’ ignition system – these will typically have their own triggering system.

The down side of CDI is that it usually has a very short duration spark. Some manufactures work around this by having multiple sparks. A well engineered inductive system can produce a long duration single spark. At high RPM the Electromotive system’s spark can be as long as 60 degrees of crank rotation….that should make sure that the mixture is ignited!

In reality I believe that any of the decent stuff out there (including the stock Porsche system) are up to the task of creating enough spark energy to fire a well tuned street engine. If you are trying to boost spark output to fire a non well tuned engine then spend your time and money to fix the root cause instead of just dealing with the symptoms.

macnewma

I see. That makes sense.

Thanks for all the good info guys. Many of the advanced issues surrounding engine management are unknown to some of the guys here (myself included). A little education helps with determining solutions for my car.

Max

m42racer

Mr White,

Great points. I am also glad we are adding some interest and possibly information to some. But I do disagree with some of your points and the generalization about the Spark timing.

Please explain how the Cam speed effects accuracy. And what data do you have that this triigering and the points you mention contribiute to 50% inaccuracy. There are many engines that have only cam triggering. Some have gear drives, some have chains, but alot have belts too. No question, there is always some belt influence. Take a look at alot of Japanese engines. Do you think the Lexus which has 2 Cam Triggers on each bank suffers from Spark timing. I expect not. But I cannot be sure either. AS for the implication that some EFI systems are not accurate within 1 degree, maybe that true. I have never tested any of them. But I am absolutely sure that the one you sell is neither more or less accuarate than most. Neither of us can connect those dots.
I did not have the time to check your math, but I assume its correct. But so what. What is your point. If your implying , which is the way I read it, that your system is more accurarte because its reads 58 teeth over a system that reads only 23, I disagree totally with that statement. Again, you are generalizing again, with no knowledge. To that point, would that make the Link more accurate than your system if we on our typical 944 system use the factory 132 tooth count from the flywheel? Absolutely nonsense. The number of teeth does play a part, but its the software and how the input information is processed thats important. % of error mean absolutely nothing unless we were comparing the exact same software and dissimilar triggering. Filtering plays a huge part here also. The tooth count at high speed, along with the sync pulse is always an issue. We all know that the system you use uses a standard 58 tooth wheel. I asked today, and the Link uses a multitude of triggering systems. That is why the stock factory 132 +1 can be used, the 968 58 tooth trigger can, along with many other types. The 24-1 cam trigger is nothing new either. Both Ford and Toyota use this is many of their engines. And on the Cam, in some cases.
Your post is full of %'s which read as total generalizations, assuming many things, none backed up with facts or data. The use of the 58 tooth wheel and the processing is typical in most EFI systems, not just the electromotive. You would not count the 58 teeth and not use any of them. Then you would miscount, and then the system would never know where it is. Please explain your point. I missed that too. What is different from your system to, say the Motec or EFI Tech, or Pactel, or Link which also use the 58 count, and count 60-2 or 58 teeth every revolution.
The real point here sir, is not that we disagree, but that there are many choices out there. Are some better than the others. Maybe, but neither you and I are experts to make that an absolute. The real decision comes down to the users budget, the cost of the system, the cost of installation etc. Where factroy triggering can be used, factory sensors etc, the the end result not suffer from any use of these must play a part in the final decision. Would ypou agree?

As for your point about accuracy over power. I do agree. But my point was when all things are equal, is the use of CDI over Inductive a better choice. I thought I put forward fairly good data, backed it up with equations to prove the differences. Many high end EFI companies, use CDI over Inductive. Why do you think they do this?
Mr White, what is the difference in the coils supplied with your system, that can generate the spark energy you state as an Inductive system, when some of the very best CDI systems along with Ferrite type coils can only match these numbers. From my research, all of the spark energy that can be generated in an Inductive system is within the coils magnetic field.
As for your challenge sir, its not about challenges, its about facts, use of common sense in applications and honesty, what systems really will do. From my own experience, and your example, a 930 engine with CIS will not run any better with CDI than Inductive. I proved that not to be the case, and so have many others. Your example is probably one of the best test cases. Firstly, if you happen to foul a plug, common in CIS engines, the inductive won't even start the engine. Next the idle is way better, and the overal power is up, due to better combustion in a Hemi Chamber with long spark travel. Twin plugs help some, but the use of CDI is a real difference maker.
As for your retort about more Ignition power is no better if you have enough, I have never seen enough sir. To run greater spark gaps, where more energy is required to jump those gaps under boost, to get a hotter flame, a bigger controlled explosion and have complete combustion is something many EFI companies and OEM's still work towards.
Mr White, we agree to disagree, and you have your reasons. I welcome different opinons and facts. I am no expert by far, so anytime I can be educated I welcome it.

Chris White

OK, I’ll try again…
Back to the accuracy thing – the tooth count does not matter as much as how it is used.
Here is some more basic info – the typical way some systems calculate when to fire the spark is to see when TDC passes and then fire the plug at a certain time after that. “Other” Management systems do not ‘think’ in angles, they work on a time based structure. So the system has to calculate at what time (in milliseconds) after a particular event (usually the Synch trigger) that the plug should fire. This works nice if you are running at a constant rpm. If the rpm is very dynamic then the ignition timing will be off by the amounts I calculated in the previous post.
Electromoitve’s patent is based on the concept of angular calculation of spark and dwell timing. This functions quite a bit differently and without a large black board I don’t think I can get fully explain the difference. In essence the Electromotive system calculates the angle between events and uses the 60-2 tooth wheel to trigger the correct firing angle. This is key because the time element changes with changes in engine speed but the angle does not.
Other stuff - The basic fact is that halving the rotational speed (cam trigger vs crank) will decrease the resolution by 50% is true for all systems – including the Electromotive system.
As for questions about spark energy – the other issue covered by patents is the method that is used to fire the coils. A feedback system is utilized to monitor the ½ charge point of each coil firing. This info is then used to calculate the correct dwell timing enabling a full charge for each coil firing up to 12,000 rpm on a 4 cylinder engine. The coil charging info is also driven via the angular calculation to make sure that the full amount of energy that the coils can generate is available at the correct point in the cycle.

OK, then back to the last Link infomercial…I mean post.

Please explain how the Cam speed effects accuracy. And what data do you have that this triigering and the points you mention contribiute to 50% inaccuracy. There are many engines that have only cam triggering. Some have gear drives, some have chains, but alot have belts too. No question, there is always some belt influence. Take a look at alot of Japanese engines. Do you think the Lexus which has 2 Cam Triggers on each bank suffers from Spark timing. I expect not. But I cannot be sure either.


For most applications the cam timing method is ‘close enough’, if you are content with close enough then so be it. Since you seemed to be concerned about gettign every last possible joule of spark energy I must have mistakenly though you were concerned about accuracy. Hmmmm….maybe this need for more spark energy and your concern about fowled plugs has some corlelation to the ‘close enough’ spark timing?

I did not have the time to check your math, but I assume its correct. But so what. What is your point. If your implying , which is the way I read it, that your system is more accurarte because its reads 58 teeth over a system that reads only 23, I disagree totally with that statement. Again, you are generalizing again, with no knowledge. To that point, would that make the Link more accurate than your system if we on our typical 944 system use the factory 132 tooth count from the flywheel? Absolutely nonsense.

By definition the higher the tooth count the higher the ‘possible’ accuracy. If you wish to “totally disagree” with that then I can’t help you, its just a reality. What your software does with this increased resolution is another issue.

number of teeth does play a part, but its the software and how the input information is processed thats important. % of error mean absolutely nothing unless we were comparing the exact same software and dissimilar triggering. Filtering plays a huge part here also. The tooth count at high speed, along with the sync pulse is always an issue. We all know that the system you use uses a standard 58 tooth wheel. I asked today, and the Link uses a multitude of triggering systems. That is why the stock factory 132 +1 can be used, the 968 58 tooth trigger can, along with many other types. The 24-1 cam trigger is nothing new either. Both Ford and Toyota use this is many of their engines. And on the Cam, in some cases.

Read the opening paragraphs and then ask your friends about angular calculation vs time based. I totally agree with you on the software issue.

You are confusing me now – earlier you were professing aboiut the virtues of CDI and how it is good for high end racing application and now you are quite content with Ford engine management? I don’t get it.

As for your point about accuracy over power. I do agree. But my point was when all things are equal, is the use of CDI over Inductive a better choice. I thought I put forward fairly good data, backed it up with equations to prove the differences. Many high end EFI companies, use CDI over Inductive.

“All other things being equal”…I guess I was looking at the bigger picture and felt that I should mention that accuracy is more important than absolute spark power. I don’t doubt that you can get more spark power out put with a high end CDI and your math can prove that just fine. But out there in the real world a well tuned and managed engine just doesn’t benefit from that level of overkill.

Mr White, what is the difference in the coils supplied with your system, that can generate the spark energy you state as an Inductive system, when some of the very best CDI systems along with Ferrite type coils can only match these numbers. From my research, all of the spark energy that can be generated in an Inductive system is within the coils magnetic field.

The difference is how you manage them. To get peak performance form a coil you need to optimize the dwell for any given situation and that is one of the keys to the electromotive system

Mr White, we agree to disagree, and you have your reasons. I welcome different opinons and facts. I am no expert by far, so anytime I can be educated I welcome it.

So can I count on you to pick up a copy of “The 944 Engine Handbook” next year? (Bentley Publishing). I will be glad to sign it for you!!

Summary – I will take accuracy over power and that’s why I like Porsches over Muscle cars!!

Chris White

Oops! The quotes don't show up the right way!!! Sorry for the confusion, I hope you guys can figure out which are quotes and which is me!!

Duke

This is very intersting stuff.... but I'm more keen on seeing pics of those 16v turbo engine's of yours Chris

m42racer

Mr White,

This thread started out as who is running wasted spark etc. I added the posts about wasted spark on sequential fueled engines and why not do both, then I got into the differences between CDI and Inductive, and you have turned this into some sort of pissing match between the two of us and the electromotive system and some of the systems I have quoted. In no post have I endorsed any system, but rather offered what some do , and what one of the currently popular systems offers for 944 cars etc.
At the very least offer up valid points that make sense. I have read your latest post about the timing angles etc, your opinion of CDI V Inductive eyc etc. I emailed it to some people I hold in high regard about EFI systems. I copied copies to 2 EFI companies for their engineers to look at and give me some idea of what you are talking about. After they all expressed an opinion of you, your theories, and your apparent intimiate knowledge of both their software, they both suggested they should give up and just buy from you. Well not exactly, but thats the nice version.

You make statements which show off your intentions here. They are not to discuss, nor to infrom, but to **** off. Your quote about infomercial for example. Yes sir, I purchased a Link system for my new 944 engine. I have been involved in an engine project for sometime now, and the system has been used on the dyno to test many of the new engine parts. I really like it. Many others appear to also, if sales are any indication. I also have a EFI Tech system on my latest 930. I have not endorsed either system over any other one. I have simply offered facts, what is offered by some and tried to offer some good explaination why most CDI and EFI aftermarket suppliers go the CDI route.

I am sure you have a lot of passion for the system you sell. It shows. Thats great. Doesn't make you right. Me neither.
As for your explaination of how the electromotive works, I have absolutely no idea what you are saying. I have no reason other than good theory, good understanding of how the typical system works, and a lot of practical engineering, to question anything you say. I just cannot understand a thing you are saying. It makes no sense to me or others who have read what you say. This is like a game of ping pong.
This was not my intention when I posted what I posted. However I always welcome critics and opinions. Many times an open mind will teach you something. I welcome the opportunity to learn something. The only thing I have learnt from you Sir is, well lets not go there. I just want people to come with sensible knowledge and not say things that make no sense.
For those who have read our posts, good luck. I have tried to offer good basic engineering data, equations to calculate the differences one will get when using CDIvInductive, but this thread has turned into a pissing contest, or will if I return Mr Whites favor. I wish not to, but wish you luck Mr White in all of your endeavors.

You Sir are quite something. BTW you won. I lost.

Funny thing though, there is a lot of really smart people on my side of this issue. They must be as dumb as I am. Who knew.