Ion sensing ignition

[MattKab]

Has anybody here considered Ion sensing ignition for a huge aircooled engine?

MegaSquirt team appear to be getting close, I have a paper given to me by a friend that almost has me sold..

Matt

[yetibone]

I stumbled upon one of those ignitions on a Saab 900 Turbo, mabye like a '96 model or something.

As I understand it, the ECM commands each coil pack to apply 80 volts to their corresponding spark plugs during a compression stroke. If that current finds continuity to ground, then ionized gasses are present within the combustible mixture, and detonation will be eminent. The firing event is delayed until closer to TDC so the flame kernel can expand while dynamic compression drops as the piston travels downward, instead of expanding as dynamic compression rises as the piston reaches TDC.

If that were applied to a VW, full engine management would be nessesary. It would have to use a crank triggered ignition, and sequential DIS ignition (not a waste-spark two coil type) with a cam sensor. The ECM would have to be able to sense a voltage drop in the secondary ignition circuits individually (4 in all) and quickly delay firing if the voltage did drop at that plug.

Imagine the benifits in a turbocharged engine. That could make a turbo Type IV feasible in essentially a bolt together kit!

[Wally]

So, if I unerstand correctly, its an ignition system with a build-in knock sensing and automatic retarding device?
Clever! You then don't have to 'listen' for knock with all of its difficulties on an aircooled engine (see other thread :-) )

Greetings,
Walter

[Richie]

Originally Posted by Wally

So, if I unerstand correctly, its an ignition system with a build-in knock sensing and automatic retarding device?
Clever! You then don't have to 'listen' for knock with all of its difficulties on an aircooled engine (see other thread :-) )

Greetings,
Walter

Hello,

Yes, in a way. But it's far from simple. A very, very high current is sent through the sparkplug, just prior of ignition. The sparkplug then acts as a sensor as it measures the "density" or the way the mixture ionises (sp?). I really don't know how to say it otherwise. The rate of ionization determines the moment of the ignition. It doesn't really retard but it's more some sort of time-travelling because it shifts the ignition forward or backward according to the ionization. Goal is to have the PPP (peak pressure point) around 30 degrees ATDC IIRC.
Well, that is what I know about it, maybe someone with more indepth knowledge can chime in?

Cheers,
Richard

[yetibone]

Actually Richard, that's pretty accurate. It is a very complex system, built around a simple idea. Ionized gasses are conductive, or can carry electrons, much like lightning jumping from cloud to cloud during a thunderstorm. When a low voltage load, roughly 80 to 100 volts is applied to the end of the spark plug, it will strike an arc, so to speak, in a heavily ionized environment.

Instead of that current jumping to the spark plug's ground electrode, it finds an easy path to ground thru ionized combustible gasses in the chamber. The ECM can then detect a drop in voltage at the tip of the plug by monitoring how much current it takes to maintain that 80 to 100 volts.

The higher the current needed to maintain 80 to 100 volts, the greater the drop in voltage. The greater the drop in voltage, the greater the number of ions present in the combustion chamber. By monitoring the current (amps, not volts) the ECM can calculate how long to delay the firing event.

Ionization of a combustible charge is due to heat, which is obviously a byproduct of combustion itself, but as that charge is pulled in and compressed
into 1/10th it's original mass at 14.7psi (pressure @ sea level) it begins to superheat. That's when ions start gathering together.

An ion sensing knock retard system is superior to a piezo-electric knock sensor in that it can detect the knock before it happens. A piezo-electric knock sensor senses the knock itself, and retards the ign timing after spark knock has already occurred.

[Wally]

Cool stuff!
Does anybody know if there is a free programmable versions of such a system available?
Thanks for the excellent explanation guys!
Walter

[Richie]

Hello,

You're right it's very cool stuff.
Free programmable? No, not yet but maybe you'll find this usefull.
http://www.hut.fi/~vvartiov/ion/DIY-Ion-Sensing-2.pdf

Cheers,
Richard

[MattKab]

Thanks Richie!

It was the 'old' version I saw..

Here's more great reading:http://www.hut.fi/~vvartiov/ion/ion.html

Does this look promising or what?

Matt

[cj011]

what year cars does this come off of for Saab and Saturn? Are there any specific turbo models that utilize this technology? thanks,

[yetibone]

Originally Posted by cj011

what year cars does this come off of for Saab and Saturn? Are there any specific turbo models that utilize this technology? thanks,

I know for fact that a '96 Saab 900s 2.3 turbo had this system. I'm sure later models do to, but I doubt that any of the old longitudinal engine'd 900's ever had even a DIS.

I don't know about Saturn. The latest model Saturn I've worked on is '97, and it had a Delco waste-spark DIS.

[zen]

does it come with an idiot light??

sorry, i had to do it. sounds like cool technology. biggest question will be cost/benefit.

[yetibone]

Originally Posted by zen

does it come with an idiot light??

sorry, i had to do it. sounds like cool technology. biggest question will be cost/benefit.

Lights, we dont need no stinkin' lights!

...well if Megasquirt makes one available, then it'll be pretty affordable. The benifits would snowball... :righton:

[MattKab]

The current SAAB DI pack fits very neatly between the cam box covers on the 95 and 93. It is a single component that drops onto all 4 sparkplugs and has 1 mulitpin plug on one end.

I'll see if I can get a faulty one to cut up and have a look

Matt