AFRs too lean - what's up?

[caramall2]

Any ideas why my AFR is lean and won't stick to the map values? My map says I should be around 11.0 AFR for load and RPM, but I keep getting 11.8, 12.1, etc. I'm only boosting at about 14 psi at the moment so certainly not too high on the boost. MAF isn't that high, I calibrated the intake last summer, I have upgraded injectors (though I haven't updated the pump), injectors are calibrated (latency and scalar), injector duty cycles aren't too high, etc.

 

Could my LC-1 O2 sensor be off?

 

Have been checking knock issues (FLKC, FBKC, IAM, knock events, learning view, etc.).

 

I changed the LC-1 sample rate to one step slower than "instantaneous" (can't remember what it is in the LC-1 software) thinking it might have been occassional spikes.

 

Been kind of swamped so have had lots of time to analyze, but still not making sense to me.

[Legacy_GT_Pilot]

A few thoughts ...

 

You could have a pre-turbo leak allowing un-metered air to enter the system leaning out your AFRs.

 

Your sensor could be off - you could try re-calibrating it and see if it reads any closer to your target? Or you could borrow a sensor from someone else and see if it still reads the same.

 

Your MAF calibration might be off. When you calibrated it last year, were you hitting your AFR targets?

 

You could try cleaning your MAF sensor as well to see if that improves things.

[Spec B]

+1

 

Assuming it is not a leak, you may need to rescale the maf. The cooler weather may have an effect on the scaling.

[boostsr20]

What injectors? Stock pump?

Offtopic, but your total timing values seem really low. Are you in the process of getting this problem figured out before tuning for power?

[Ridgeracer]

Run a Learn Viewer. Really interested in seeing your fuel trims...

[caramall2]

Thanks all for the inputs. Engine is brand new and trying to be careful with the tune. Also put on an AVO 380 (slightly used).

 

I've been a bit swamped which has tripped up me systematically going through things...I think I really need to do that and you guys have given some good input.

 

- Injectors are DW 740 cc. Started with settings many here have used, and then did partial check (didn't let it run long enough though) using RR injector tab.

 

- Stock pump. I have an AVO waiting on the shelf and would like to get that in (stock should be able to handle things right now though).

 

- Learning View is attached. (It was a few weeks ago, but car was doing the same thing.)

 

- Wideband: I am going to order another O2 sensor as a back up, and to check if it gives me the same readings.

 

- Timing:

-- I was trying to stop some FBKC and events that were happening just as I rolled on the throttle so I started to pull back timing a lot (was probably a bit hyper cautious and am a newbie at tuning).

-- Right now I think I'm way underpowered...timing pulled too far back probably and boost only getting up to 14 psi (another issue...might be the temperature correction table thing-a-ma-bob (at cold temps restricts boost (or I think it was WGDC)), or that my actuator arm isn't tightened up enough yet). Always a joy when there are like 3 different things not dialed in right ;-)

 

- MAF & MAF sensor:

-- I was hitting proper AFRs last fall before the engine problems. I did put in TGV deletes as part of the engine rebuild...that might throw things off.

-- It's mounted in an AEM CAI...can double check it to see if it's clean. Another member who's pretty smart on tuning (lives nearby) got the same MAF scaling results as me (and we worked somewhat separately on it)...I think he has TGV deletes too. I even changed them about 5% or so for the heck of it to see if that helped AFRs.

-- Guess I can try a rescale, but if something is off on injectors/pump/etc, then my rescale would be "off" right? I guess it doesn't matter if either of them is off a little, as long as the net result (MAF and injector settings) is correct AFR per my map. I can do the whole "loading a map with mostly 11.0 AFRs" and rescale the MAF, seeing if I can bring it in line.

 

- Leaks. That's something I guess I should really check. I think everything is ok unless the intake to turbo connection has a problem (that's hard to see/get at it seems).

 

Thanks again guys for helping.

[boostsr20]

I also just got my car back from an engine build and I also did TGV deletes at the same time. I haven't done anything yet to my map other than disable the TGV CEL's and the car is running flawlessly and much more powerful than before the build. My fuel trims are back to almost identical from before at 2%,0,0,.8%.....

 

As far as timing goes, I'm running about 4 degrees more than you in the 5100 range and so is BlackGT. He's running a ton more power than both of us are too and he's getting away with it. We are both in Michigan and I don't know about his local stations but I know that I'm still getting winter blend fuel for another week or so. All in all, I think you'd probably be safe to slowly add a couple degrees back into your base map.

 

Get a can of carb cleaner and start spray around all connections and places that would have been touched. My car after being put back together had a massive leak at the TGV causing a crazy rich condition at idle.

[qikslvr]

When was the last time you pulled your WBO2 sensor out of the DP and did an "open air calibration"? I know I have to on my LC-1 every 3-4 months to keep it accurate.

When my first sensor went; it started reading out of range and then finally would read 4.70 at idle and go negative under boost. Then it just went totally crazy and died all in one day:lol:. The new sensor has been rock solid for a couple weeks now while tuning my 20g.

What boost control solonoid do you have?

[LittleBlueGT]

As far as timing goes, I'm running about 4 degrees more than you in the 5100 range and so is BlackGT. He's running a ton more power than both of us are too and he's getting away with it. We are both in Michigan and I don't know about his local stations but I know that I'm still getting winter blend fuel for another week or so.

 

His timing is a little low, but in all honesty so is mine on my pump tune. (11:1 AFR)

 

I cannot run anymore timing on our pump gas and not get knock, even my alky tune has timing that is comparable to many 93 tunes I see here on the boards.

 

FWIW I can easily run a couple more degrees (or more) timing on gas just south of the border.

[boostsr20]

I wish we all got the gas they do in Japan

[caramall2]

O2: I did do a free air about 3 weeks ago, but will try again. Do you know if the LC-1 keeps all its settings (sampling rate, etc.) after it's disconnected from power? I am going to get a backup sensor too just to double check.

 

Boost: I have the stock boost solenoid.

 

Forgot how crappy our gas is. Sounds like that's a big driver on the timing issue.

[caramall2]

Well, I want to do some long logging on the RR injector tab, but, in the mean time, I set the majority of AFRs to 11.0, and then did some logging. I then sorted by MAF, calculated % increase or decrease in MAF scaling required, then averaged for each 10 g/s block (e.g. 101 to 110, 111 to 120, etc.).

 

Attached is the graph. Interesting. If there's not some issue with the fuel pump, etc., and MAF scaling is the issue, this certainly shows problem areas. Just wouldn't think it would change this much since my engine replacement, TGV deletes, etc.

[caramall2]

Did a second logging and got fairly similar results (the two are overlayed...second one in red had a little more psi on boost (about 1.5 more max). That gives me a little more faith that the sensor is reading correctly. I definitely want to:

 

- Install my AVO fuel pump...just to make sure there's no supply issues (which I think could be a factor up high).

- Clean the MAF sensor.

- Install another O2 sensor just to see if it reads the same.

 

Also, I did an injector tab on Rom Raider and it suggested decreasing the injector scalar from 715 cc (what it suggested a month ago) to 709 cc.

 

I also want to see what happens if I change the MAF scale according to my results....to see how close to AFR map I get.

 

For the rest of you logging, do your AFRs bounce around a bit at WOT...e.g. 11, 11, then maybe jump to 12, then back down to 11? I know LtBlue and I have discussed this. I moved my sensor down about 12 inches (was real close to the turbo) but that didn't seem to change it much. In the LC-1 software I also slowed the the sample rate to every 1/6 of a second.

[RIPjeep2002]

Whats a normal AFR on a stage 1 tune? Mine is 11.04 consistently.

[Spec B]

you need a WB02 to know what you are hitting. The stock sensor readings are useless (if in the stock position) while under boost.

[caramall2]

Mine jump around...e.g. if you go down the log at WOT readings will go something like this over one or two seconds of reads:

11.2

11.2

11.0

12.1

11.3

11.2

etc....

 

Meanwhile the map for that instant might be calling for 11.1 or something (depends on your tune/map...your 11.04 under WOT and middle to upper rpms is typical). They might average out to around a target, but they jump around more than I think they should. I know LtBlue used his same sensor on another car with the bung further away from the turbo (about where my new one is) and it read really consistent.

[caramall2]

Oops, I didn't catch that. RipJeep2002, Spec B is right. If you're just relying on your stock sensor, it doesn't work under wide open throttle conditions...it will just read aprx 11 and stay there no matter what's happening. If you don't have it, you need an aftermarket O2 wideband sensor. I always recommend this and to check your tune, no matter who gave you the map....better safe than sorry.

[Spec B]

FWIW, here is a good example of what mine looks like vs what the map calls for..pretty consistant.

 

Spec B AFR Error.csv

[caramall2]

Thanks, good to compare with. There's certainly some movement, but only +/- 0.1 or so...not much. I'd have no problems with that. It's when I get my "12.1" popping in there I get nervous (but don't have much experience to go off of).

 

I definitely want to get my AVO pump in to see if that steadies things at all. Also, my MAF error plot is showing a consistent trend...if that still shows up after the AVO pump install, I'll probably change the MAF scaling.

[lepich]

Sorry for off-topick, but it is related..

Can AFR become leaner when intake temp gets lower? (ddenser=more air). Lets say it was tuned on 70F outside, and now 40F? I've seen leaner AFR this morning, but a couple days ago it was OK, and was much warmer here, in Atlanta:-)

I do nor see any Air temp compensation for a fuel in our ECU.. Have I overlooked it?

[caramall2]

There is a MAF compensation table for intake temp (at least in my LGT hex file for ROM Raider).

 

Also, from what I understand, some sensors (I think ours and most modern ones) use a heated wire anemomenter. The temp of the wire decreases as airflow increases and likewise, the conductivity changes too. As such, it has somewhat of a built-in compensation for temperature. I'm no expert, just what I read.

[Spec B]

+1, That is how I uderstand it.

[qikslvr]

You should try using Logworks to log your AFR on a WOT run. I know it won't give you RPM data, but at least you would be able to see if it's actually the AFR that's off or if RR is lying to you. I know my RR will post incorrect values and flip cells at times. I'm thinking that you are seein an 11.1 instead of a 12.1 at that one point, but RR is misposting it. Again, Logworks would at least give you and accurate data curve to go off of.

 

FWIW I put my LC-1 into a different DP(Invidia cattless) with the bung a couple inches further away from the bottom curve of the DP and it now reads rock solid. It was all kinds of flakey and skewed in my Autospeed catted DP.

[lepich]

There is a MAF compensation table for intake temp

 

Thank, I got it now.

Now it is warmer outside, without any change in tune my AFR went back to norm, (12-12.5). I am running 50/50 meth, so may be it is related to water temp or something like this.

[caramall2]

Interesting idea about log works....I'll have to try that. Also interesting how a different pipe changed things...maybe there is flow and "turbulence" issues for different points in different pipes.