Homebuilt Ion Sensing - Fuel Economy, Hypermiling, EcoModding News and Forum

Harlan · 11-11-2016, 11:01 AM

For those of you who don’t know Ion sensing is a method of measuring combustion chamber pressure using the spark plug. It involves high voltages measuring the current flow across the spark gap, and the current flow is roughly proportional to chamber pressure. This can be used both for detonation detection and for closed loop ignition timing for maximum torque. This can be used for both best power under all conditions or for best efficiency under all conditions.
After a few failed attempts I now have Ion sensing working on my 2004 Mazda Rx8. I am currently moving forward for closed loop ignition timing. One of my end goals is to push the lean limit of the engine while advancing timing for high fuel economy.

Here is what I have so far.
Testing Ion sensing on the bench:

https://www.youtube.com/watch?v=AAuMCs_jrrE

First test of the piggyback ignition timing connected to the car without coils:

https://www.youtube.com/watch?v=uw29szxoRJ8

Second test of piggyback on the car with ignition coils, and with the car running:

https://www.youtube.com/watch?v=q2DUA-rTDEc

First test of Ion sensing on the car, shows Peak Pressure Point relative to TDC:

https://www.youtube.com/watch?v=hP6tso6cnbk

Next test is going to be measuring PPP while driving, and after that I’ll be trying closed loop timing at idle.

Harlan · 11-12-2016, 05:50 PM

Part of the reason I'm doing this is to work out lean burn on a rotary. I previously tried lean burn by adjusting the MAF sensor voltage until the throttle was wide open and power started dropping off. The biggest problem was that I had no way of controlling timing, and a lot of energy was wasted out the exhaust with higher EGTs.

With this I hope to have enough control to run >20:1 AFR with proper timing and get some solid data. My end goal is 40MPG+ cruise, and then to implement a control strategy to bring me into and out of lean burn smoothly.

Harlan · 11-14-2016, 07:18 PM

Just tried a simple closed loop ignition timing in idle. It worked, but I am having problems with the ion sensing signal being too weak so I still need to tweak hardware. I will post an update with video when I get around to shooting some.

Once I get this straightened out, my next step will be to go into Open loop fueling with closed loop timing. This should support lean burn, because the ECU won't let me go into lean burn on closed loop. I plan on starting another thread specifically dealing with lean burn and why it works well on a rotary. That is, assuming I'm right and it does work well on a rotary...

Ecky · 11-14-2016, 08:24 PM

Forgive my ignorance of rotaries, but I was under the impression it was pretty easy to destroy them.

Anyway, subscribed.

Harlan · 11-14-2016, 09:38 PM

It's easy to blow a coolant seal or warp an end plate/housing if you overheat. Easier than blowing a head gasket on most engines.

Also must turbocharged engines don't survive long, but I'm pretty sure that's from running too much timing/bad tune. Tuning rotaries is a real black art, hence the ion sensing.

Other than that they are pretty similar to other sports cars on reliability. If you premix oil with your gas and don't baby them too much they can last without issue for years.

redpoint5 · 11-15-2016, 02:25 AM

Why is it called ion sensing instead of pressure sensing, or current sensing?

Harlan · 11-15-2016, 02:49 AM

It's called ion sensing because it's measuring ion density to approximate pressure. I didn't develop the technology or the name, I'm just developing my own system that uses the technology.