Ion sensing circuitry for 4 wire coils

[dnb]

Just thinking about Delphi's solution to getting an ion current without a dedicated power supply...

I plan to implement the circuit from Delphi's coil design for one of my coils as a test.


I'm not convinced that 80v is sufficient to make a detectable ion current, so I plan to use 200v zeners.  (39p each from Farnell)

I am wondering what the best value of capacitor is?   
Assuming the spark is at TDC:

At low engine speed (1000RPM) we need there to be enough charge to power the ion circuit for just under 6ms, and the circuit will have 55ms recovery time (allow for wasted spark)
Conversely at high engine speed (7000RPM) we need charge for 0.85ms and have a recovery time of 7.7ms

So taking the 2 extreme requirements, we have 6ms use time and 7.7ms recovery time.  Not exactly ideal, but I suspect it doesn't matter too much about recovery time.

Charging time is not an issue really - there is the energy there is.  The only snag is that if the spark is a long way advanced then the ion circuit needs to be powered for more crank degrees, so more time.

To some extent, the value for C1 depends on the resistance and inductance of the secondary coil, so there will be tinkering required to get it right...   I'm planning to start with 2 microFarads and see what happens.

As rough "guesses", R1 should be 2k and R4 should be 38k to give us a reasonable chance of good scope-measurable signals.  This is based on ion currents being of the order of 1mA and the zeners being 200v.

Does anyone have any thoughts before I attempt to destroy my coils?

Thanks.

[Sprocket]

Is this any help, or have you already seen it

http://vvnet.fi/ville/ion/DIY-Ion-Sensing-2.pdf

http://www.freepatentsonline.com/6741080.html

[MWfire]

In my case ion curent is about 0.1mA.
So r1 should be 10-20kohm(i put 50kohm timmer, so i can adjust desirable output). This will give 2-3V signal. It's desirable to protect dac input(or scope)with zener diode.
Capatinace should be capability give 0.2mA for 4ms(0.8uC charge). 1-10uF should be ok.

[Jorgen]

We looked at this circuit back when we started with the ION work and IIRC we found that it was primarily meant for missfire detection. I think that it can be used for PPP and knock detection as well but that it would be harder to detect PPP without a steady supply. There was also some problems with the internal diode that is fitted in the secondary circuit of most 4 wire coils.

I think that 80v will work well for ION measurement, I would try that first. I would try this myself but even if all my cars have 4 wire coils all also have the internal diode.

As a sidenote I scoped the ignition of my BMW this weekend to try to detect an intermittent coil problem. As I run the coils in dual-out mode I saw alternating combustion and no combustion. Given the very different patterns when the coils was ringing after the ignition event think that it would be possible to detect missfires by only measuring the primary voltage of the coil. I will need to hook up the home signal to the scope to know which trace is the combustion. Warning, a pretty rugged scope or probe is needed for this measurement as I saw >400v for the first 4-5us of the spark event.

We could charge the ION capacitor from the primary side, this could be done with or without CDI and should make it possible to use commonly available 4 wire coils.

About ADC protection, I would try to use a BAV99 double diode instead of the zener, they are far quicker and more likely to save the ADC from any spikes.

Jörgen

[moahten]

Have you made any progress with the ION-Sensing system?

Thanks,
Mo-