Capacitive or Inductive Ignition?
 

While thinking about Wasted Spark setup for the 84-89 Carrera Cars in this thread:
http://forums.pelicanparts.com/porsche-911-technical-forum/433152-diy-wasted-spark-84-89-carrera.html

It quickly developed into a discussion about Capacitive Discharge Ignition (CDI) or Inductive Ignition.

I figured this topic is best for a new thread, so here it is.

Anyone using CDI? Would love to hear from you.

Thanks

CDI has had the reputation of giving CIS a bad name and was unable to compete with todays high energy ignition systems.

I'm running an M&W 4 channel CDI system. I'm splitting 3 channels to make 6 signals to 6 waste spark coils. So far it's working well, but I'll probably switch to a 6 channel CDI in the future.

So will you go to coil-on-plug and then use 1 dedicated channel for each plug still in wasted spark configuration with 2 channels firing together (meaning 3 pairs)?

Thanks for the input.

I'll still run the same 6 waste spark coils with each coil feed by one CDI channel. There will only be 3 inputs from the ECU, so yes, 2 channels will fire together.

Got it thanks for the input.

One con on CDI is that some say the spark time is to short and this can cause lean mis-fires. Have you seen this at all in the low RPM range?

Thanks.

why is CDI giving CIS a bad name? i have not heard this.

...me either:confused:

EDIT: in addition, the original poster's car wouldn't be CIS if it's an '84 Carrera, right?

Correct, 84-89 Carrera is inductive ignition, my car is a 84.

I have heard this CIS cars with CDI have issues but not sure why folks say this? But keep in mind that those older CDI boxes are primitive in function compared to the more Modern CDI boxes. Many of the newer CDI boxes do multiple sparks below 3000RPMs.

But let's keep this going, would love to here from CIS owners on why the CDI system failed them.

That's because you haven't talked to any technicians who have worked on CIS 911s for a living. The general consensus is that 99% of their problems are ignition related.

Unfortunately, by the time the techs get them, their owners have screwed up the CIS so badly chasing down ( CDI related) ignition problems that it becomes impossible to know where to begin.


Cheers,

Joe

FWIW, I went with Electromotive XDi crank-fired ignition (inductive) during my twin-plug 930 engine rebuild (CIS). It seemed to be a "turn-key" solution with a decent track record. And at the time I was way over budget, and couldn't swing the cash for the 12-point dizzy and related parts. It worked very well for me. However, I always lusted after the "big headed" ignition dizzy.

Fast forward, I have since removed Electromotive and gone with the JB Racing 12-point dizzy with two MSD6AL CDI boxes, and two MSD Blaster SS coils. I'm very pleased. I wish I had factual dyno numbers to back it up, but my CIS based turbo engine does indeed perform better in the under 4,500 RPM registers. It pulls harder, and feels torquier. At the upper end, I notice no difference.

http://forums.pelicanparts.com/911-930-turbo-super-charging-forum/413780-cdi-coilpack-ignition-options.html

These kinds of threads make me nervous because they so often degenerate into something nasty. Both systems worked well for me. I'm happier with the more mechanical aspects of running an ignition distributor than I am with having Electromotive coil packs. My car is a daily driver and I personally feel that I'd actually have a shot at a road side repair if faced with ignition trouble.

Obviously your engine is EFI and thus, YMMV...

Hi Joe,

OK but I'd like to hear what the technical issues are? For starters the CIS CDI used points to fire it, I suspect this could be the cause of some of the head ache, but I'm just speculating.

Here are a few thoughts:
1) So are the CIS CDI boxes simply failing because of age?
2) Back in the day when these boxes where new how where they?
3) Many of these boxes are 30 years old or more, I'm a EE and simply amazed they even still work. These are high voltage boxes and this type of circuitry simply takes a beating.

Then modern day CDI (digital) systems have also come a long way and most likely I'd use a modern day system and not the old CIS CDI setup. With that said it's still very usefull to learn what was bad or good about the older systems.

What I'd like to get out of this thread is concrete reasons why CDI is good or bad and the same goes for Inductive.

I have my own opionions but wish to hear from others as well.

I have. Being that my experiences are with CIS based turbo engines, I'll have to keep it there. But during my rebuild, several folks suggested I keep CDI. Even if it meant running the stock ignition system, and just plugging the extra spark plug holes (twin plug), until I was ready for other options. However, I was anxious to fire the extra plugs and I liked the functionality of being able to dial in my ignition curve, offered by Electromotive. Again, I have no regrets. I'm just happier with my JB Racing set-up.

First, you are correct this type of thread can get nasty and I hope that does not happen. Let's keep an open mind and just think of it as learning.

It's kind of interesting you say you saw improvement in the 4500RPM range. Here's why that may be: the 4500RPM to 5500RPM range is max torque range and under these cylinder conditions are the most extreme pressures. Spark has a very difficult time jumping a gap under extreme pressurized vapor as compared to lower pressure vapor. Then I also bet that if you use CDI you opened your plug gap to .05 to .06 and the reason you can do this is because CDI delivers higher voltage. Opening the gap puts more fuel into the gap and thus a better chance of starting the flame front. My bet would be that if you keep the same stock gap with CDI you would notice no diffrence.

The bottom line is that both inductive and capacitive systems function well and so long as either system fully ignites the fuel and starts the flame front then no improvement can be made. However, one theory is that at max torque our Carrera Ignition may not be optimal, actualy it's not the ignition it's the fact that the plug gap is to small and the ignition can't drive a wider gap. So one could fix this with a CDI system or a better Inductive system.

Just my thoughts on this, but I'd love to know if you also widend your plug gap when you went to CDI?

Thanks for the input.

From the Innovate Forum, posted by Klauss, the inventor:

http://www.innovatemotorsports.com/forums/showthread.php?t=3201&highlight=911s+ignition

More spark energy depending on grounding would really depend on how the CD system is designed. If the discharge current through the coil has to run through the engine's ground strap, you can have some losses.
CD systems use the coil as straight transformer. For example with a turn ratio of the coil of 80:1, and a CD cap voltage of 450V, you'll have theoretically 36000 Volts to initiate the spark. But, once established, the burn voltage on the spark gap is only about 100-200V. This in turn gets transformed back by the coil (as transformer) as an almost short and discharges the cap very fast. That's why CD ignitions have such a short spark duration. Gets worse with solid copper core spark wires.
Many years ago I built myself a CD ignition system that initiated the spark with a conventional CD, but that system switched the coil to 12V when the cap was almost discharged (but the arc still burning). This still used the coil as half-wave transformer, but then running from 12V (yielding 960V no-load spark voltage). The spark duration was adjustable with a pot to up to 2.5 msec. Worked great, but halved the life of the plugs, and required a heat-sink on the coil for high RPM (6-cyl at 7500RPM).

I did eventually open the plug gaps up. I'm at the office right now and the exact numbers escape me, but I think I'm running a .045 gap, twin-plug. Bosch W4CS spark plugs

Interesting, it's just a educated theory on my part but I think the idea of wider gap helps under hi-torque conditions. So whatever system (CDI or Inductive) you can achieve this with really doesn't matter.

Then I've also heard that inductive systems don't have enough time at hi-RPMs to fully charge the coil. This is true, many Inductive coils need as much as 5milliseconds to fully charge, and at 7000RPMs you have only 2.5milliseconds at best. But we must be missing something because we all know these cars don't mis-fire at hi-RPMs. I think the explanation may once again come down to cyl pressures, above 5700RPMs our torque really starts to fall off and thus cyl pressures drop and therefore a half charged coil has plenty of energy to start the flame front. Once again just my thoughts.

Low RPM is great. I've noticed a high RPM problem but it's intermitent and I didn't get a chance to figure it out before I fried my clutch. Hopefully I'll get a new clutch in soon and get a chance to find out if there's really a problem or not.

Another question I have on CDI is why do many of these units fire multiple sparks below 3000RPMs? Just wondering what others think about this.

My impression is that certain manufacturers use COP CDI when it comes to highly boosted turbo engines and inductive discharge for ordinary N/A engines.

I don't quite see the value of multiple sparks. When flame front is created, it's created. CDI also allows another tricks due to it's quick discharge. It allwos ECU to measure ionization current for example.

SAAB used this for ages, BMW just about started. The fine thing with CDI and ion-sensing is that you can detect the phase and thus get rid of cam position sensor AND use fully sequential ignition and injection.

You start cranking and fire twice. Once the engine catches up, you use ion sensing to detect combustion and kill the ignition/injection frequency in half and off you go.

So CDI is really the state of the art, but is probably more costly to manufacture.

My thought as well, once flame front starts what's the point of more spark. So why do CDI folks do this? Could it be that maybe at low-RPM (or idle) the first spark does not get the lean (14.7AFR) mixture ignited and the extra sparks help with this? They must have added this multi-spark feature for a reason.

CDI's have a shorter spark duration compared to their inductive brethren so there is the possibility of a flame-out. At low RPM, you can cram in multiple spark sequences to insure this doesn't happen. As RPM's increase, the spark window decreases and multiple sparks are not possible.

I mean no dis-respect, but I'v never heard of Flame Out? Once the fuel is ignited I don't think it can/will go out. What I think happens is that it simply does not light the first time if the mixture is lean as is often the case at idle or low RPM lo-load. In this condition the second spark fires it, but the timing is off (slightly more retarded). For this reason I think it's better not to have the multiple sparks, would be nice to be able to turn this feature off. This way you know what's going on.

Combustion needs three components; fuel, oxygen and a ignitor. If it's overly rich it won't fire, if it's overly lean it won't fire. Of course, you can get by with just fuel and oxygen, but then you have detonation. :(
A flame-out can occur under conditions in the combustion chamber due to the turbulent conditions during the compression cycle where the mixture is not equally distributed throughout the chamber. Ignition conditions for combustion is not static in an internal combustion engine, unlike a stationary chamber that has had time to equalize fuel and oxygen components.

Dave,

Thanks for the education. So how often do you think the 3.2L engine may be prone to flame out? Hopefully not very often.

I'm just trying to figure out if the multiple sparks are really needed in a properly running 3.2L?

Also, from what I'v learned so far I think that a wide spark plug gap (0.05" - 0.06") is really the trick to getting fuel ignited. The wider the gap the more fuel fits into the gap. But you need plenty of spark to jump a wider gap and I think CDI can help with this. What's your thoughts here?

Sal,
The CDI in the Porsche's is actually pretty good and I don't believe they use a multi spark ignitor. The CDI systems inherently provide a higher spark voltage than Inductive, hence the wider plug gaps. It's really hard to beat the MSD systems the a cost per performance gain. I know a ton of people on this board that use the MSD's and they swear by them. You have to spend a LOT of money on an inductive system to get that kind of performance.
Keep in mind the EDIS and other Automotive general purpose designed systems are only going to get you about 45KV. Capacitive will easily get you nearly 50% more KV. That being said, if I were twin-plugging an engine, I would design a wasted spark. Twin plugging will get you an additional 2% HP and a cleaner pipe (i.e. emissions) to boot.

[QUOTE=WERK-I;4236294]Sal,
The CDI in the Porsche's is actually pretty good and I don't believe they use a multi spark igniter. QUOTE]


The Porsche CDI design was completed and in production long before the benefits of multi-spark came to be known, if they exist at all. I've only read the literature from the manufactures and have not seen the results of any independent dyno runs as to it's benefits. Sounds good in theory at low RPM to insure complete ignition of the flame front. The reason they go to single spark at higher RPMs is due to the recharge time for the CDI capacitor.

Higher spark voltage and a wider gap has been proven to be a benefit for turbo applications but I'd like to hear an tested results of multi-spark with a turbo. Why? Because I've been thinking about trying it but hate to waste time and money if it has been proven ineffective.

Jim

One fundamental difference is that the spark energy in a CDI system needs to be stored and retrieved from a capacitor of some sort compared to the inductive system where it is stored in the ignition coil.

In my opinion a capacitor will age and change properties over time while a coil will not. The Bosch CDI boxes use what I believe is some sort of unipolar paper capacitor with electrolyte filling. With age these capacitors loose some of the capacitance and start to develop DC leakage and polarization effects. This will eventually lead to failure of the box. Even modern capacitors will show dielectric fatigue effects over time.

I don't think that component aging is an issue with transistorized inductive systems found on the 3.2 and 3.6 engines. With these failures are mainly due to failing solder joints.

Another observation from running both CDI boxes and Motronic boxes on the bench with a single spark plug hooked up is this:

The Bosch CDI boxes seem to have no problem driving a single spark plug with 300 sparks per second. That corresponds to 6,000 RPM for the CDI box and 36,000 RPM for the spark plug !!

When I do a similar test with a 3.2 Motronic the spark plug start to behave like an electron beam welder way before I reach 4000 RPM. It does not produce individual sparks anymore but what looks like a constant arc. So there must be significant differences in spark duration between the two systems for what it is worth.

Ingo

Really? Then why are Porsche OEM CDI plug gaps so small when compared to MSD?

Maybe because the factory recommends the plugs changed every 30,000 miles? How many of us follow that recommendation? :)

sorry sal, i cant let this one go.
are you for real? 99% of CIS is CD related? what about the earlier carbed/MFI, why arent their problems 99% CD related and why dont CD's give them a bad name? there are a lot more adjustments on carbs tahn CIS. sounds like you dont know much about CIS or you would know most of CIS problems are due to old cracked/broken hoses or other rubber on the injection system. CIS is very susceptible to air leaks. as far as screwing up CIS, there is only ONE adjustment, except fot the idle, to screw up. where people make their mistake is in thinking that as long as the CD is making noise, the hi pitched whine, that it is working. when the cap goes bad, the oscillator in the CD still works, but the cap does not charge up. and how do you know what techs i talk to? is the tech that work on the racecar "spot" back in the 90's acceptable?
and i still have not heard that CD's give CIS a bad name, not even once on this board.


sal,
i have repaired several of the bosch CD's. most of the time the big cap in there goes bad. they are drying up from age/heat.
i had a permatune when i bought my car. changed it over to MSD. could not really tell any difference. beru wires were put on just before i got. i understand that these do not work well with the MSD. i have also tried different plug gaps. even up to .065, no difference noticed. i plan on changing over to the magnacore.

Ingo, when you say 4000RPMs do you mean crank RPMs? so that the plug is firing about 33 times per second? (4000/60=66 per second, plug fires every other crank turn 66/2=33)

In theory, as RPM increases in an inductive ignition you tend to run out of Dwell time to charge the coil. But as RPMs increase the coil also starts to act more like a transformer, meaning (in Theory) if you drive the coil at say 1000 cycles per second (1000hz, on and off very fast) then even though the input is a DC square wave the coil still tends to act like an AC Transformer. It simply induces the primary signal into the secondary coil. So if the coil is wrapped 100 secondary turns for each 1 primary turn (1/100 ratio) then the 12v square wave on the primary tends to produce 1200v on the secondary. I wanted to mention this theory because it may be the reason that the spark looks like a constant beam on the bench in free-air at hi-RPMs. Just a theory of course.

OK, what you say really helps make my theory even more valid. If the ignition system (CDI, Inductive or whatever system you like) is properly starting the flame front at all RPMs and Load conditions then NOTHING can be improved (except for maybe dual plug). The point is that moving to a new system will only improve things if something was wrong with old system.

So I'm not surprised that changing to MSD made no noticable diffrence, this simply means your old stock system was working and not producing any mis-fires.

Then, once again the larger plug gap helps get flame front started more reliably under max-torgue conditions (extreme cyl pressure). But if the smaller gap of say 0.035" is already functioning under these conditions then moving to 0.065" would not make any diffrence.

What I'm trying to figure out is if the stock system has some limitations? maybe it simply does not in which case it's a waste of money to change it. Sounds like in your car the stock system was doing it's Job just fine.

on a stock porsche, i dont think you can do much to make the ignition much better, unless you have points.
remember though, on my car i have the beru wires which are not suppose to work well the the MSD and my car is stock with CIS. i dont know if you will notice any difference until you start changing compression unless you have a dyno. i would be interested in what a longer spark would do.
here is something you will enjoy reading. i tried to contact him, i was interested in what he thought about plug gaps and heat ranges of plugs.

http://www.clubwrx.net/forums/tuning-electronic-engine-management/14426-engine-basics-detonation-pre-ignition.html

Nothing wrong with points for driving CDI ignition. The points current is only 420mA so the replacement interval is literally tens of thousands of miles and they are about ten bucks a set.

Everyone says "use a Pertronix" hall effect sensor or optocoupler but when those things fail you can't tell without a scope or logic probe. At the roadside it's very easy to replace a set of points.

Anyway, back to this excellent thread-- I do think it all comes down to spark duration.

still have my points and no plans to change. i put that in to cover the ones that dont like points.
very interesting. sal has good thoughts and ideas. this stuff helps you learn. i dont think i would spend the time or money to implement any of this since my motor is stock, but still interesting.
i work on radar systems for the FAA. a radar system has whats called a pulse forming network (PFN) the is esentially the same as a CD and coil, except the PFN has much much higher current. so i have always wondered about the longer spark versus multiple or higher voltage sparks.

The basic question still seems to be unaswered. What's better longer cooler spark with smaller plug gap? or shorter hotter spark with wider plug gap?

I think they both have pros and cons and what may work in one car might not in another.

But maybe for our P-Cars either works fine, I say this because folks have reported good results with both CDI or Inductive stock and non-stock ignitions.

For those that have tried something non-stock did you notice any other improvements? Like idle or cold start?

Now I'll tell you why I went down this path: My 3.2L is a Euro engine with 20/21 Web Cams, ported polished intake, SSIs and 2in2out M&K pipe, fine tuned chip for this engine. The trouble I'm having is lean misfire at idle, to fix this I bumped idle up to 920RPMs and advanced Idle Ignition to 5deg BTDC. This really cleaned up the idle, but I still notice some misfire if I let the mixture lean out to say 14.7AFR (O2 Sensor Closed Loop) but if I richen the mixture to 14.2AFR (no O2 sensor) the mis-fire goes away. Keep in mind I'm only talking about idle here, everything else is fine. Now this is a Euro motor that was not speced for 14.7AFR as the Euro engines do not have O2 Sensors. The Euro cars run at about 14.2AFR (open loop with no O2 sensor). I suspect the higher compression in the Euro Engine does not like 14.7AFR at idle. So the idea I may try is a simple MSD 6 CDI box just to see what happens.

Sal,
I don't think the problem you're experiencing is ignition related. After reading your mods, it may all boil down to induction. Have you spoken to Web Cams on this matter?
Ignition choice? I would pick an ignition that has the hottest spark with the longest duration through the widest gap. Unfortunately, that beast does not exist. That being said, I would go for the most reliable system out there...probably the MSD. Geez, even the roundy-round boys of NASCAR are using these things and they're turning close to 9K RPM with them.

Dave,
I have spoke to WebCams they claim the 3.2L should idle fine with these cams. Could be the entire combination of the setup. For now the idle is just fine since I don't run the O2 sensor, I simply have the AFR set for Euro spec at 14.2AFR and it idles nice. But I do think I'm going to try the MSD 6 CDI with a matched MSD coil.

I don't think there's such a phenomenon as a cool or cooler spark. As it relates to jumping an air gap, current systems either produce a single spark or a longer duration spark. Wider gaps require higher voltage to jump the gap and requires an ignition box and components with this capability.

There are ignition boxes that do produce high voltage and long duration spark. Again the "hotness" factor is not part of the equation. Do the pro teams use the 6Al box? I'd be surprised. I think they use a higher level product, IMHO.

Sherwood