Results from individual cylinder tuning

[Stocky]

Thought some of you might like to see this.

Below is a screen shot of some data logged on my AFR Tuner.

I've got individual Lambda sensors in 3 of the header pipes (waiting for a 4th sensor to show up in the post)

I've added some labels to the shot to help it make sense.

What you are looking at is a roll on (some variation in throttle position) in 4th gear.

RPM is at the top in orange.

Throttle position is in the middle in blue

Then the individual A/F ratios of each cylinder are shown in Red at the bottom.

For arguments sake, lets say the target A/F is 13:1 (good power and fast acceleration)

The Red scale goes from 12:1 to 14:1

http://www.asfphotos.com/upload/1248504461.jpg



So.....

Over all it's a bit lean (more than a bit actually)

You can see it in this shot but it's more evident in other parts of the data that the variation (and the actual A/F) vary more at part throttle than full throttle.

This engine made 182HP at the back wheel at Casino. The tune at 100% is pretty good.

However, at anything but full throttle it's all over the place.

The funny thing is, the fueling at part throttle feels fantastic.

The data above is also with some tuning of individual cylinders already done (not much though)

In a min I'll post up an example of the cylinder offsets so far....

[Stocky]

The base tune was done over 18 months of running in the mid pipe with just a single Lambda sensor.

This created the Map for cylinder 1 ( of course it's actually a combination of all 4 cylinders)

This is an example of the base map (that Cylinder 1 now uses)

http://www.asfphotos.com/upload/1248533500.jpg



The main AFR Tuner constantly attempts to tune cylinder 1 to what ever A/F Ratio you target (eg 13:1)

When you run the expansion unit and map individual cylinders the base map on Cyl 1 becomes the reference for the other cylinders.

The other 3 cylinders all tune by "offset" to cylinder 1.

So for example, if Cylinder 1 is right on target but cylinder 2 is a bit rich then cylinder 2 will be offset leaner for that throttle pos/RPM

Below is a section of the cylinder 2 offset so far and this is only about 20 mins of riding.

http://www.asfphotos.com/upload/1248601772.jpg



The data says that the tune at part throttle is shocking but the bike feels wicked smooth and powerful etc.

After only 20mins of riding it's already significantly better than before and there is obviously a long way to go.

Can't wait to get a 4th sensor and get some time on it.....:D

[Tony OW31]

That's awesome, once you have mapped it, will it self adjust on the fly using the O2 sensors?

[DaveK5]

Could be wrong stocky..but isn't the variation in A/F ratio due to the great rate of change in when rolling on,it does all seem to come together when at or near WOT.
Do you expect to get the ratios of all 4 cylinders following similar patterns right through out the RPM range?

Bloody interesting stuff...cheers.

[Stocky]

Tony OW31 - It slowly maps itself and the cylinder offsets while you ride.

Dave K5 - Looking through the data there is a bunch of time where the throttle is constant (eg 26% or something) and the RPM is rising, the A/F is still all over the place.

I don't expect it'll be perfect when there is a rapid change in throttle position but when it's constant or not changing fast I hope to get all the cylinders tuned instead of a good average.

After a long time with the single sensor in the mid pipe the A/F ratio was pretty good at most throttle positions and rpm commonly used.... however... this is now shown to be just a good average.

The mid pipe might say the A/F is good over all but it can be made up of any combination of rich and lean individual cylinders that average out ok in the common mid pipe.

Looking at the data and talking to some guys who've been running the expansion unit for a while there is an awesome amount of part throttle power and response to be had.

I'm told my data is pretty typical and that V Twins are typically much worse.

Even a Desmosedici is all over the place when you check individual cylinders.

I guess that's why Moto GP bikes etc all have a Lambda sensor in each cylinder......

[BANDITROD]

mate i dont know what your talking bout but it sounds fucking awsome ..

[RUSS M8]

That's very interesting Stocky... Will be interesting to see the difference in seat of the pants and dyno when all 4 are the same (or at least closer than that!). Are you still running standard throttle bodies and injectors?

[MONO]

Gerard are you running a pc 111

[Stocky]

Running an AFR Tuner and then an ATX4 expansion unit.

The AFR Tuner works a bit like a PC3 but it runs it's own Lambda sensor and tunes itself while you ride (among other things)

Then the ATX expansion runs another 3 sensors that enables it to tune individual cylinders while you ride.