How far can our ignition systems go?

[cribbj]

CDI works on the principle of charging a capacitor to a much higher voltage, say 400v, then this capacitor dumps its charge into the primary side of the coil which produces a much hotter, albeit much shorter, spark on the secondary.

Some purported downsides to CDI are:

1) Because the spark is a much shorter duration, there is not a complete burn of the mixture.

2) Possibly because of 1) above, some CDI systems are not known for their smooth idle. Some designs get around this by producing a "multi-spark" discharge at idle.

3) Not all coils are suitable for CDI. Any coil with an integral ignitor cannot be used with CDI.

4) Because a CDI system discharges all of its energy in typically 1/3 the time of an inductive system, it tends to find the weak spots in an ignition system faster.

5) Cost. To do a CDI system right requires at least another box or two in the system. These boxes cost anywhere between $1K and $2K.

Downsides of an inductive ignition:

1) For a single coil, distributor based system, there is not sufficient "dwell" time at high RPM to charge the coil(s) to produce a hot spark. (This disadvantage is reduced or eliminated by using COP's or CNP's)

2) Multi-coil waste spark setups can have the same problem as 1) but this can be reduced by boosting the charging voltage on the primary side of the coils. For example at 12v the Mototron coil requires about 1.6 msec to charge sufficiently to produce a 50 millijoule spark, however if the charging voltage is boosted to 16v, it only requires 1.1 msec for the same spark energy.

It would seem the ideal ignition would use inductive technology at low RPM and CD at high RPM. Interestingly HKS sell a module called the Twin Power DLI which does just this, and it is a commonly used solution on Supra's which have had their ignitions reconfigured to waste spark setups.

 

[JustenGT8]

Thanks for the quick rundown John.

So seems CoP should be enough if not run as waste spark?

 

[Lextreme II]

I am sure John will jump on this one. I found this COP is direct fit to the 1uzfe and bolt right onto the stock valve cover holes.

 

[cribbj]

Justen, most COP's can be run with either an inductive or a CDI setup, so long as they don't have integral ignitors. CDI systems don't require ignitors.

Some coils either in a COP, CNP or other configuration, are better suited for use with a CDI setup than others because they have very low impedances. This helps keep the voltage rise times very short, which is a benefit to having fast, predictable spark discharges. The Honda CBR motorcycle coils are a good example, and they actually are part of an OEM CDI system. Most (all?) Japanese bike ignitions are CDI now.

David, that looks like a Honda S2000 or Accord COP. Is it Denso P/N 099700-070? I've got 4 of those here, and at 169mm overall, it's longer than any Toyota COP I've seen, and looks like it would fit down in the valve cover of the 1UZ just fine and have plenty of plug engagement.

It's still not long enough to do what I want, however, which is something like the photo below. Even these longer Honda coils are going to require a short spacer/extender to get them up above the valve covers.

(BTW, those are GM and/or Chrysler 300M COP's in that photo and they're 168mm long also)

 

[JustenGT8]

So between the CBR or the VVTi UZ CoP which would you suggest?

Next point of clarification. I'm pretty sure the SM2 doesn't have enough outputs to run 8 channel ignition, but with CDI not requiring an ignitor does than mean an 8 channel CDI avoids having to run waste spark?

 

[cribbj]

So between the CBR or the VVTi UZ CoP which would you suggest?

Next point of clarification. I'm pretty sure the SM2 doesn't have enough outputs to run 8 channel ignition, but with CDI not requiring an ignitor does than mean an 8 channel CDI avoids having to run waste spark?

First, the VVTi COP has an integral ignitor, so that automatically eliminates using it in a CDI system. Look at the list of coils I provided a few messages earlier and find those that do not have integral ignitors if you're interested in running CDI. There are certainly more than these out in the market - most of the older Bosch and Beru coils that are for the BMW/VW/Mercedes engines do not have integral ignitors.

In fact, apparently Bosch & Beru only switched over to pencil coils with integral ignitors in the last few years for BMW (I have access to the BMW TIS), while Denso has had them in Japanese motors since the mid/late 90's.

Second, I think unless you can configure the SM2 to address the coils in a binary fashion, and/or use a multiplexor, you're still going to be limited to waste spark, because it's the number of ignition outputs that the ECU has that determines whether you can run a true 8 output sequential ignition, or have to run two sets of 4 coils in waste spark.

BTW, Kdog, who is far more knowledgeable in this area than I am, feels there is no downside to running waste spark.

 

[JustenGT8]

Issue is i don't want to spend $1K plus just to run a couple more psi Totally stock setup with regapped plugs is doing fine @ 17psi 400+rwkw.

I'd like to go to CoP if only to tidy up the engine bay but if this is going to cascade into a bunch of issues i'll just stay with what is currently working?

 

[cribbj]

I thought I had some specs on your SM2, but couldn't find them, and apparently the Autronic website is still down after their fall out with Ray Hall?

Honestly, if I were you, I'd look hard at dumping your dizzies and going full crank triggered on the ignition. The sensors are already on the engine to do this, and I think the Autronic should read VR sensors as well as Hall's?

Whether you go COP or CNP, you'll not have to be concerned about the ignition holding you back if you want to crank up the boost, or wind that motor past 8000 RPM.

Benefits of crank trigger are a much more accurate reference for the spark, and reduction of losses in the system by eliminating the electromechanical distributor devices. Plus, if you go COP, you can get rid of all or most of the ignition secondary system, which is where most of the problems occur.

I would guess that if you wanted to go crank trigger and COP, and you did your own wiring, you could probably do it for the price of the coils, and whatever the reconfiguration of your Autronic would cost.

Distributor technology is 100 years old. Shoot, even Detroit has finally seen the light and is getting rid of them!

 

[JustenGT8]

No probs with going CoP with the SM2. At the time it was wired up i hadn't found a CoP that would go close to fitting was all. I think i need another reluctor for the crank angle sensor and it will mean stripping the loom for a significant rewire.

Just some discussion on this thread suggesting some CoP aren't very powerful so while you gain from losing the dissy you lose from a weaker coil?

 

[cribbj]

Surprised your crank angle sensor isn't already installed and wired up? If you're using it for the fuel injection, it should be able to do double duty for the ignition too?

I think we're speaking in relative terms wrt the COP's and their outputs. They may be smaller than one of the existing coils, but they're still only having to spark a single cylinder (albeit twice an engine cycle), where the existing coils are sparking 4.

Plus there are no secondary losses as there are with a coil, a distributor, and 4 long high tension leads.

I don't think I've seen or read an objective comparison between the power output of a coil & distributor setup, and a COP setup. That would be an interesting comparison.

 

[JustenGT8]

Sorry, i meant cam angle sensor. As you say, crank angle is set up. Well i have all the bits, when the built engine goes in i'll give the VVTi ones a wurl 1st and if no good fork out for the extra ignitors and try the CBR ones.

 

[cribbj]

Justen, aren't you running a cam angle sensor? I thought you were doing full sequential injection with your Autronic, but you must be batch/bank firing?

When you get ready for a set of coils, email me and I'll send you the latest data I've compiled on what fits best with the least issues. I'm hoping to eventually put the data up here on Lextreme when I've finished, but I keep finding promising new coils that ought to be looked at, so it remains a work in progress.

Lately I've been curious about the Mitsubishi and Isuzu coils as they look to be longer than the Toyota coils.....

 

[KDog]

Justen, (other information included for people who may not be familiar with CDI or Autronic)

For your engine I would be going wasted spark CDI.

1) You already have the ECU to do it (the SM2 or were you buying another one)
2) For your application there is no disadvantage (John will be pleased that I do know of some) in running wasted spark (the autronic ECU whilst doing sequential injection for 8 cylinders only have 4 ignition outputs).
3) It does away with the expense of the dizzy's, coil leads, rotor button etc so you can remove these from the engine.
4) I would be inclined to ditch the stock reluctor crank/cam sensors and replace them with hall effect sensor rather than using a reluctor interface (the autronic SMC/SM2 ECUs only have hall effect inputs so for reluctor sensors you need an interface)
5) The stronger spark (Autronic CDI produces one of the biggest sparks you will ever see) from the CDI will easily ignite the mixutre at whatever cylinder pressure (boost) you run.
6) If you can't find some COPs that fit nicely then you can use the good old fashioned bosch MEC717 coil for CDI use.

Just to play DA on Cribbs post about the downsides to CDI.

It is a common misconception about CDI that the short spark can cause a problem. The role of the spark is to ignite the mixture it is not there to sustain the burn, the flame kernel will do this. Because the spark is very short but very powerful it can ignite almost anything.

I have heard some critics state that the short spark time can result in the mixture not being ignited (hence multispark CDIs) becasue it isn't near the spark plug when the spark occurs. My response to this is that any head which has such poor mixture distribution (don't get me wrong the A/F ratio in the chamber, but it shouln't be this uneven) will be a problem regardless. The fuel mixture is drawn into the flame kernel as it burns, if you have the corect ignition timing and A/F ratio then there will be no problems. It takes a very small amount of energy to start the kernel (~6mJ) which is far less than most ignition systems output.

There is of course the costs involved in the CDI unit and the extra coils.
If you need help wiring then let me know, most installers get it wrong (either material or install) which means the system doesn't run 100% or it interfere's with the ECU.

 

[JustenGT8]

Thanks kdog, between your and John's advice i have a good feel for where i should head (finances allowing).

I'll have to go and check what sensors i have for what as i'm sure i have the stock sensor on the crank with a reluctor interface but from memory the cam sensor doesn't have a reluctor....hmmm it must do so i'm a bit confused...wiring was done over 4 years ago now so recollection is a bit fuzzy. I do know the timing signal betwen what you see on the ECU and what advance is in realilty is 5 deg out because of the teeth....just something i have to remember for tuning session

 

[Zuffen]

John,

Have you looked at the Mitsubishi Lancer Evo COP?

There is a company here selling them on a billet plate (a good way to mount them and you could put a piece of folded allow or ventilated alloy over them) their Website is http://hioctaneracing.com.au/eshop/product_info.php?cPath=35_86&products_id=1040

The photo isn't too good, but I've seen another and the coils look like they could be in the right size range.

 

[cribbj]

Cheers Rod, I'll check them out. I haven't acquired any of the Mitsubishi or Isuzu coils, but those in that pic look a bit like the Honda S2000 coil I have.

 

[cribbj]

Here's an interesting bit of information concerning the ignition systems used in Formula 1. In 2000 and before, I was fairly sure the F1 teams were fairly evenly split between using CD and inductive ignitions. To satisfy my own curiousity I recently asked someone close to the sport what the F1 teams were now using for ignition systems. This fellow's company provides meteorological services to a number of the F1 teams, so he's a regular at the races, and is well known by the teams.

Here's the response he received from a representative of the Renault F1 team:

"They are not capacitive discharge systems and everyone uses coils built into the spark plug caps now. The ECU controls the low voltage power output to the coils and then normal inductance produces the high voltage."

Interesting that they're all using COP technology too, although it's probably not the garden variety we see from Denso. Here's an interesting page from Bosch Motorsport, showing their COP technology for F1. These are probably more similar to what the F1 boys are using: http://www.bosch-motorsport.de/content/language2/html/2847.htm

 

[KDog]

A few years ago I was contracted to Denso for F1 (We had a flame measurement system they wished to use) they ran this system, whilst its not CDI, its also not what you would call inductive.
Due to IP I am not allowed to disclose any specifics about the exact coils/discharging invloved but it is a modification of COP inductive with secret sauces.

Due to control ECUs now in place ie not just the Mac/Dens/Fer ECUs on the cars pretty much everyone has been running the same system for a long time.

 

[cribbj]

I've been reading/browsing the Bosch Gasoline-Engine Management (3rd edition) lately, and what a wealth of interesting information is there. One particular page in the ignition section caught my eye last night, so I've scanned it and have attached it below.

The page gives a summary of the two types of high voltage distribution systems - the old electromechanical style distributors, which they summarise as being "....no longer of any significance to new, modern-day engine management systems." and the distributorless (stationary) distribution system, which we know as COP's or CNP's.

What's interesting about their summary of COP systems, is the last paragraph on the page, which is not very supportive of dual-spark ignition coils, ie waste spark systems. Albeit, they are referring to coil sets which ignite two cylinders at a time, ie 4 coils for a V8, not COP's per se, but the principle is still the same.

Bosch summarise by saying "Because of these limitations, dual-spark ignition coils cannot be recommended."

The limitations they spell out, are exactly the problem I had with my first waste spark ignition system back in the early 80's offshore, where we had a combustible mixture in the manifold that was ignited by a plug firing on its exhaust stroke, and the resulting backfire destroyed most of the top of the engine.

So, my interpretation is that Bosch do not recommend the use of waste spark technology in an engine. Interesting......

Full credit to the authors of this book for this scan; anyone interested in the history of engine management systems really should pick up a copy. It's chockablock full of useful information:

 

[Zuffen]

I agree wasted spark does concern me.

When the best ECU only supports 4 igniton drivers it does make it hard.

I want COP/CNP but need to figure a way around the problem.