in-situ cylinder pressure tuning

[Holmz]

Has anyone used the pressure sensors that Kinser and other make that have a sensor that side in the side of the spark plug?
Looking for how one maximises the power/torque using just the pressure measurements, as well as how to get the MBT values for spark advance.

[RaceMechaniX]

I believe you mean Kistler sensors. AVL also produces one.

Yes, I have experimented with both direct measurement and the spark plug versions. Challenge is the resolution and filtering unless you are running through another piece of equipment. The frequency response of these sensors are extremely high and even the best ECU's typically do not have high enough frequency inputs to make use of the data. Typically you run the sensor to its amplifier and the output goes to your combustion measurement equipment which links the crank position and cylinder pressure. This hardware samples in the range up to 1GHz.

It might be possible to average and filter the data to smooth it out into a useable trace, but setting up an ECU to use it is a challenge. Perhaps with the latest M1 generation and Simulink this is possible.

TG

[Holmz]

I am thinking of using a 192 kHz.24 bit sound "card" to sample it. Really an channel audio device which is fire wired or USB'ed to the lap top.
Then I can get, crank, cam, throttle, and MAP in there as well, so I can determine RPM, and crank phase, etc.

192 kHz is very fine resolution, or 5uS between samples.
I would think that I probably only need some where in the 20-100uS/sample ?

[SprinterTRD]

A while ago we used Optrand spark plug cylinder pressure sensors

[Holmz]

A while ago we used Optrand spark plug cylinder pressure sensors

And???
Did you get any insights into tuning the engine with it?
And how was the data used in the process?

[RaceMechaniX]

Holmz,

The logging rate is OK, but filtering will be your biggest challenge in addition to syncing with TDC. Most in-cylinder combustion systems sync with the crank at 0.1Deg crank angle resolution and need to be physically calibrated with a piston TDC via LVDT. A thermodynamic drift occurs in the measurement from true TDC when measured.

If you are looking for knock limit and some basic other data, your set-up should be fine. Otherwise you are looking at a more sophisticated system.

A good scope will also do the trick.

TG