Getting Lambda to 0.91 in Closed loop

[meatbag]

I have been working on a customer car for awhile now that is running E85/E98 on a 13.5:1 compression LS7 motor with a 251/264 .666lift 112LSA cam. The car is driven quite a bit and in different climates so I dont want to give up some of the closed loop attributes but the car runs smoothest while cruising(no bucking) with a lambda just richer then stoich. I have played with the narrowband voltages and stoich settings in the ECU but other then keeping the car in open loop I can never get the ECU to maintain a 0.91 lambda in open loop. The car is very sensitive to msec changes in injection time since it has 1000cc injectors so I have reeled in the proportional trims to keep the oscillation around stoich more in check but other then that I havent had much luck. Timing has been adjusted in these areas to minimize the surging as much as possible but if i could just get this one last part worked out it would be perfect.

Any ideas or do I just need to run the car in open loop the whole time?

[chevy lover]

try to maintain the VE table for closed loop first because it affects on open loop later

start again with closed loop tune

[Mez]

Why do you want .91 Lambda in closed loop?

[DSteck]

What kind of injectors are you using? Without proper data, you won't get consistency.

You won't get anything besides 1.0 in closed loop.

[meatbag]

What kind of injectors are you using? Without proper data, you won't get consistency.

You won't get anything besides 1.0 in closed loop.

ID1000's so I have all the proper data and everything is entered. I am running the car MAF only so I havent paid much attention to the VE tables. Are they still referenced even if you are not switching to SD at any point?

[DSteck]

SD calculations are not used if you are in pure MAF unless the MAF fails. You cannot run 0.91 lambda in closed loop, that's all there is to it.

[meatbag]

SD calculations are not used if you are in pure MAF unless the MAF fails. You cannot run 0.91 lambda in closed loop, that's all there is to it.

Just wanted to make sure I wasnt missing something. I might try rebuilding the tune in using SD then.

[DSteck]

SD still won't let you run closed loop at .91 lambda. You can run open loop and control your commanded EQ ratio based on a ton of different tables though. It is pretty nice. I use that option to help huge cam setups drive better by manipulating fueling at certain operating conditions. You can run pure MAF to do that, too.

[Higgs Boson]

The goal of closed loop is 1.0 by definition.

You need open loop.

[pontisteve]

Not sure why you would want to run closed loop, since an open loop tune on large cam cars is more likely to be smoother from lack of oscillation. Plus, you can command a slight rich condition, which the motor will like, as you are leading towards.

I think you might want to experiment with injector timing delay to improve your drivability some first, before chasing other options. That cam is big.

But if you really want to run .91 @ closed loop, there is a way. Innovate LC-1 widebands have 2 programmable analog outputs. You can set one up to simulate a narrowband O2 sensor. And then you can skew it. The only time I've heard of this being used is when someone wanted to lean out the car beyond stoich slightly for mileage, while still keeping the O2s.

[hymey]

Keep it in OL and run .91 lambda in OL, CL waste of time , run more timing at cruise and advance injection timing to suit camshaft it will clean up, all my e38 setups drive like stock,

[pontisteve]

Advance injection timing? Or increase injection delay?

Even with a pretty hot cam, I think .95 lambda would be sufficient.

[LSX378]

Keep it in OL and run .91 lambda in OL, CL waste of time , run more timing at cruise and advance injection timing to suit camshaft it will clean up, all my e38 setups drive like stock,

Do you mean boundary angle?

[Cobra Dave]

I have one that runs 0.91 in closed loop. (Which is not what I want).
(GMPP LS3 complete package)

The car is with its customer but when I get my hands on it again I will be playing with the switching points for the O2 sensors.(Thanks to advice given on here by "Pontisteve" and others.
Currently they are set at 650mv (From factory). I'll change them to 450.
So, if yours are set at 450 I would give 650 a try.

Would be real nice to know the outcome.

[meatbag]

I have one that runs 0.91 in closed loop. (Which is not what I want).
(GMPP LS3 complete package)

The car is with its customer but when I get my hands on it again I will be playing with the switching points for the O2 sensors.(Thanks to advice given on here by "Pontisteve" and others.
Currently they are set at 650mv (From factory). I'll change them to 450.
So, if yours are set at 450 I would give 650 a try.

Would be real nice to know the outcome.

Ive tried this but to no avail. Not sure if its because of the ethanol content or not, i would think that would not affect it. Probably more so because of the inherent design of a narrowband sensor and its voltage curve. I think I might try the LC1 idea and use it to simulate a narrowband to trick the computer.

[pontisteve]

The O2 just finds stoich, no matter what the fuel. The commanded air/fuel ratio = the stoich ratio / the EQ ratio. So stoich basically affects open loop and wot modes. If you're running 10% ethanol, you should punch in a 14.08 stoich before beginning to tune your VE and MAF tables.

I don't know exactly why some O2 voltages are set in tunes to 650. Most switch at .450. I would think biasing that number up or down would slightly richen/lean the engine. That number shouldn't be something we solve for on the car. It should come from the O2 sensor manufacturer.