What the hell am I doing wrong!!! Chrome Tuning

[RealMcCoy]

I am going to take issue with the theory that closed loop only targets stoic... I understand that the fuel table doesn't control the target AFR, but somewhere in the code there is a target average 02 voltage that the ECU is trying to achieve via STFT. That target varies with load and temperature. The ECU is fully capable of targeting and maintaining closed loop operation between mid 15s and low 13s before it reaches the limit of it's range and drops into open loop.

My best example that can easily be observed would be the warm up phase... Watch when it goes into closed loop, and what the AFR is doing as it warms up.

You can also see it driving with a light load that's not enough to force open loop. The AFR will start to go rich, and it's still in closed loop.

[Greg E]

4)The fuel map is not actually an "afr map". It is a somewhat happy co-incidence it works out roughly this way.

Correct it's a correction factor map essentially. I should have populated it with percentages instead of AFR numbers but again that idea is something carried in from the EVO guys before I fully understood it....

[Greg E]

somewhere in the code there is a target average 02 voltage that the ECU is trying to achieve via STFT. That target varies with load and temperature.

Nope. I wish this was the case as that would have been simple to adapt to say, wideband feedback...

Closed loop is nothing more than a PID. What that means is it's just responding to cycling. Not a target voltage but a response to a reaction. I'm sure they had their reasons to this approach but I was never able to successfully create my own target voltage feedback operation back when I was experimenting.

[DrGonzo]

So as for condition of the car. Zero Boost, Vacuum leaks on the intake side and the exhaust side was pressurized and all leaks fixed there also. So I am sure it is not caused by random air entering/exiting the system. I only run 93 Octane in the engine as that is the premium grade we get around here. Never ran anything less than 93 in it.

I'll get the laptop plugged into the innovate controller and check the settings in there. Maybe something is off in those settings which could explain some anomalies I have run into with tuning. I did get the innovate controller used and bought a new WBO2 for it. I should be able to confirm the scaling from within the controller itself.

As far as the scaling using in Chrome for the Innovate WB and the formula in Evoscan, do those look correct? I wondered if they may be wrong as they were something I found on the forum and looked correct. Pretty sure the WB scaling I got from Innovates site themselves so I think that is correct.

And yeah I have to agree with GregE that using the AFR instead of percentages in that table is a bit confusing as that is how I read it. I have yet to even touch that table so I'll make a mental note of it for down the road.

Once I get some free time to actually work on the car I'll look into the WB to see if there is an issue with it. I know it goes through the calibration cycle as it should so it may end up just being faulty.

[DocWalt]

Honestly, I'd replace the Innovate with AEM's new UEGO with the Bosch 4.9 sensor. Nearly everybody that I know with an Innovate hates them due to random bizarre issues. Yes, it's great that it can be calibrated but you shouldn't need to do it more than once (or again if you change sensors)

[anyonebutme]

FWIW, my wideband reads super lean at idle and low loads but matches the dyno wideband at WOT. I think it's the parts store replacement sensor.

[edit]: AEM Uego.

[DrGonzo]

Ran an AEM for a while, but didn't want another gauge in the car. At the time I think they only made the gauge/sensor combos but ti looks like they have a standalone now like the innovate. Not really wanting to change it out unless I can prove it is indeed faulty. I don't DD my car so spending a min or two every so often to run it through a free air calibration is fine with me.

[Jimvr4]

AEM UEGO 30-2310 has no guage in the car. They've had this one for a while.

Back to your Innovate. The ideal scaling values would be:

0 7.35
1 10.358
2 13.366
3 16.374
4 19.382
5 22.39

I believe Chrome limits these entries to tenths so what you had there was reasonable. However if you read 16 AFR when the actual AFR is 14.7 then there would have to be a 0.4V error in the voltage reading. The reading for 14.7 AFR should be 2.443V

I'll suggest the most likely issue is the sensor ground is not connected to the proper signal. This is very important for accurate readings. The correct sensor ground is on the ECU not chassis ground.

[DrGonzo]

Yeah I'm pretty sure Chrome rounded them all up when I entered them in.

I'm positive I have the sensor ground to chassis. I never ground anything to the ECU.

I have to do some rewiring anyways for the BSLCD as I need to tee off the WBo2 output to the BSLCD so that it is able to read the WB directly and not through the ECU. The BSLCD to too laggy when reading data from the ECU to be useful for the WB data I need it to display. I'm hoping that Tee'ing off the WBO2 signal to 2 sources does not cause any issues with signal degradation....

[DocWalt]

Yes, sensor ground must be on the ECU's reference. Otherwise you see these whacky problems you're having.