Gernby

As I learn more about tuning, I'm realizing that I've done just about everything backwards. When I started tuning my car months ago, I started with full throttle fuel tuning, then ignition timing, then part throttle fuel. However, I've notcied that my fuel tuning seems to be great one day, then crappy the next, and my LTRIM and STRIM is unpredictible depending on climate and heat soak. I realized that this is probably due to the compensation tables, but figuring out how the adjust them is tough.

Here is what I've done in an attempt to get them straightened out.

1) Create a throttle map that will allow me to hit the same spot on the fuel table precisely over and over again. I did this using the same approach that I did for part throttle fuel tuning.
2) Wait for the car to be totally cold
4) Start the car and hit the datalog button
5) Immediately pull straight out of the garage, and accellerate from about 2K RPMs to 5K RPMs in 2nd gear. Note: the car heats up very quickly, so don't dilly dally around. Every second counts!
6) Slow down, then accellerate from 2K to 5K again
7) Repeat step 6 for about 5 minutes
8) Park the car, and let it idle for 10 minutes to build up some heat soak.
9) Start repeating step 6 again for another 5 minutes.
10) Stop the datalog, and go home.

The goal above is to see how the STRIM and LTRIM change as ECT rises from ambient temp to about 195 degrees, then see how STRIM and LTRIM change as the car idles (and heat soaks), then see how the STRIM and LTRIM change as the IATs drop from ~130 F to 100F as the heat soak diminishes.

When I create my throttle map, I wanted to limit MAP to about 400 mBar across the whole RPM range, so that step 6 would give me a very repeatable pass through 3500 RPMs at 400 mBar each time. When I looked at the datalog, I decided to ignore all RPM and load levels except for 3500 RPMs at 400 mBar. I created an Excel table with IAT, ECT, STRIM, and LTRIM for each time that I passed through 3500 RPMs at 400 mBar so that I could identify the trend as ECT and IAT changed. There was an obvious trend, which I did my best to apply to the tables. This took some serious through, even though there really aren't that many values to edit.

The end result is that I wound up with much more consistent TRIM values, but then I had to go back and "fix" fuel tables for part and full throttle. Basically, I believe the compensation tables should have been fixed FIRST, before tuning ANYTHING else.

Any thoughts or suggestions?

Gernby

On a side note, I'm really beginning to believe that it would be impossible to do a great tune without street tuning. I have no idea how anyone could create a good tune on a dyno alone.

s57_s2k

S2000 does not have MAf (Mass Air Flow) sensor so it does not see how much air is entering . it calculate air based on MAP ,RPM ,IAT so you just can not have perfect tune in speed density system (weather changes , humidity is not constant ) .

this is why they made closed loop operation to be close to perfect .

I think you can have a good tune on a dyno alone (in our condition "flashing the OEM ECU" ) if our ECU is well calibrated (I mean IAT , ECT table ) from honda .

scareyourpassenger

How has this been working out for you?

It is hard to believe that Honda didn't get this close to start with.

Gernby

I think it's been working better, but it was already pretty good.

I understand your point about Honda getting the IAT / ECT compensations right, but you have to consider what I've done to change what "right" is. I believe the changes I have made could easily change the "thermal dynamics" of the engine. I'm not sure where the ECT sensor is, but I know that the IAT sensor is pretty far from the combustion chamber, and I've made changes that will impact how the intake charge will be heated along the way (carbon fiber intake chamber, hondata intake manifold gasket with throttle body bypass, 80% water / 20% antifreeze coolant mix with Redline Water Wetter).

scareyourpassenger

That makes more sense then. I am keeping the stock airbox and intake gasket so I might be ok. Only way to tell is to finally get around to some partial throttle fuel adjustments. I decided to wait till the new header arrived

scareyourpassenger

I have noticed quite a range in the af values from day to day and from when the car is first warmed up to further into a drive.

It looks as though the af is rich starting out and leans out as it gets warmer. From day to day it is also inconsistent some days it reads 12.8:1 and others it will dip into 12.3:1. I am unsure wether this is due to the stock wideband reading everything different when hot vs cold or outside temperature.

Gernby

Hard to say if the OEM wideband is the culprit, but humidity can also impact AFR, since there is no sensor for it. If it is the compenwsation tables, then I suggest altering just one table at a time in small incriments. You really just need to change the slope of the compensation to make it more consistent for your build. Basically, don't think too much about the raw numbers in the tables. Just increase or decrese the numbers a couple % at a time. If I were in your shoes, I would start by decreasing the IAT fuel compensation in the low temp range by 1% and increase the IAT fuel compensation in the high temp range by 1%

If you are using the values from the canned calibration, then my suggestion above would be to change the trim% in the high flow table for 71* from 3.0 to 2.0. Then I would change the trim for 104* from 0.0 to 1.0. I would probably also change the value for 158* from -3.5 to -2.5. Then I would do some more datalogs to see if it got any better.

scareyourpassenger

I really need to get a wideband installed. Humidity has been fluctuating quite a bit lately so that might be the case.

I will adjust the values to see if that will help but I really wish I knew how much I can trust the oem wideband. If it turns out tto be very temperature sensitive it might make sense to create a heat sink for it.

Gernby

IIRC, the OEM O2 sensors already have heat sinks.