MicroSquirt.com read only forum (Jan.15,2018)

I have code 3.770 on a Sequencer for a 2.8 liter turbo charged inline six.
The car is my daily driver and I have the base tune fairly well nailed down
so I decided to tackle X-Tau.

I read everything I could on this feature and did a couple dry runs enabling
it in TunerStudio 2.06 to check the defaults. I turned on X-Tau with WE
and let the car idle to warm up. Once up to temp I went for a quick drive
and noted my idle and cruise AFR had richened up by 2.5 to 3 points.

I started by scaling the X from the default of 30% at all RPMS. Note that
the car ran fine, otherwise, no issues at all with drivability, so these changes
were all about getting the EGO numbers back to where I had them.
The RPM range is 750,1500,3000,4500,6000 and my Accel/Decel percentages
are at 15,12,10,7,5. The car is still smooth, and EGO is closer to my
desired numbers. I don't want to reduce the puddling too much, so I
started looking at the timing, and these defaults also seemed way too
high by default. I started triming them as well, with the Accel at
(300, 250, 200, 150, 100) and Decel at (350, 300, 250, 200, 150).

I realize that I have to the best/right settings for my car, but I am hoping
I can get some confirmation I am going in the right direction. Another
config item that may help is that I have 42# injectors, which leads to
relatively low pulsewidths and duty cycles.

The X-Tau should only affect your PW and hence AFR if you are accelerating or decelerating. It will persist a bit afterward, but only on the order of a second or two. If it is affecting your idle and cruise then it is being triggered by too much jitter in the PW and you need to tune that down until it doesn't affect idle and cruise. You should be able to see this on a datalog - the X-Tau correction is shown there and it should stay at 100% +/-1or 2% at idle and cruise. It sounds like this is exactly what you did by reducing the sensitivity to changes in PW (the X component). And yes the large injectors, small PW at low rpm are likely a factor in the jitter.

You can see the same sort of thing with tpsdot or mapdot when you have a lot of jitter in tps and map (because of noise on the signal or a tpsdot/mapdot AE trigger threshhold that is too low.) It will cause AE to be on all the time also.

BTW, if you are using the same tps/ map AE and adding X-Tau, then if it is helping your Accel/ Decel, then great. If it's not helping then you should either drop the X-Tau or, if you want to experiment, start reducing the tps/map AE by say 10-20% and make up for it in X-Tau if you need to. Again evaluate - am I better off or not. If better you can continue on this path.

Excellent, those pointers are exactly what I was looking for. I was and
am very happy with standard AE, and thought to make it better. Knowing
how to interpret the correction and more importantly when is key getting
this tune right.

My X values are now in the single digit range while the time values have
only been reduced by about 20% from the defaults. I think I have what
I need to get this right, and if diminishing returns kick in and the result
is not an improvement I can always go back to my working standard AE
settings.

I captured a datalog on the way to work this morning and found that at idle and
cruise that correction was ranging from 98% to 117%. The I noticed something
odd. During actual acceleration events the correction would tighten up, from 99%
to 102%, giving the impression that the activation of this feature was inverted.

Subjectively the car is running very well, and it is only when I look at the logs
and TS guages can I see it is not quite right.