EA says add 4% more fuel....already too rich

[tremec77]

I have been running some logs through EA and it keeps telling me to add fuel. My Lambse are around 14 and my AFR is around 13.4 but it keeps telling me to add full to the MAF curve. Over all, it seems to be wanting me to add about 4% more fuel. Why would it tell me to add more fuel when I am already too rich?

[Chucko]

Please tell us more. Under what conditions? Are you using a wideband to confirm the AFR? Did you start with an accurate curve for the MAF? How about the injectors? Injector characteristics for the Ford Racing injectors can be found at their website, do a search here and you'll find the links.

[tremec77]

The MAF specs are in my sig. The MAF transfer seems to be spot on. I messed up when I said it is already rich. What I meant was I am commanding around 13.5 ARF at cruise and my LAMBSE and wideband match pretty close. Even with them matching it is still telling me to add more fuel.

I am using the FMS 30's and I have my slopes set to equal each other. I have tried using the High and Low slope values that are on FMS site but it made no difference in what EA suggested for the MAF curve. When I run a log through EA it tells me to add about 4% more fuel to the entire MAF curve.

I'm starting to think I may need to change my slope settings but I'm not sure how I should change the values to fix the problem.

[Chucko]

At cruise... that tells me you and I are having the same problem.

I am also seeing this behavior at light throttle moderate speed street cruise conditions. LAMBSE at 2000 RPM and ~15% load is around 13.4 in my latest logs. The EEC also seems to want to command richer mixtures at closed throttle with loads in the single digits. At other combinations of load and RPM, commanded AFR is much closer to stoich.

My current theories are:

• Cam overlap drawing lots of intake air through the cylinder and straight out the tailpipe. Injector timing after the exhaust valve closes isn't helping much.
• Vacuum leak, or other unmetered intake air source (e.g. PCV leakage). Lots of potential mechanical causes for this...

But it's a little encouraging to know that I'm not the only one!

[cgrey8]

If EA is calculating using the WB as a reference, then it shouldn't be asking you to enrich the MAF if both the LAMBSE and WB are relatively close to each other most of the time.

However if you are using a WB as your reference, but not datalogging the WB OR are datalogging the WB but have EA setup such that it is not using the WB, then EA is "assuming" you've filtered the datalog of all Open Loop datalog entries (and preferably transient Closed Loop entries) and thus is assuming the EEC was aiming for 14.64 AFR. When it see's the EEC with LAMBSEs of 13.5 when it is assuming the EEC modulated to 13.5 because the EEC was trying to maintain 14.64, then this would be an indication that things were running lean. And in this case, EA would ask you to enrich.

Without a datalog and a tune, it's difficult to say much more than that.

[tremec77]

Okay here is my latest tune and my latest datalog. Maybe you guys can see something I'm not. I will add that I am selecting "use WB" in the EA options and I am also filtering out transient and enrichment fuel.

http://www.mediafire.com/?yzcyg3mnomj DATALOG

http://www.mediafire.com/?nhomnz1joe3 TUNE

[cgrey8]

The 1st part of the log was started while the engine was in Closed Loop. The LAMBSEs are in the 16s indicating the engine is a bit too rich and the EEC is having to pull fuel by leaning out the LAMBSEs. The interesting thing though is the AFR is saying things are way lean. But when you look at what the HEGOs are doing, all is working and the LAMBSEs are just responding to what the HEGOs are telling the EEC.

At position 7.47731161, the EEC suddenly goes OL. At first, it appeared this happened for no reason. But what it appears triggered OL was the Throttle moving. Once the TPS returned Closed, the EEC returned to CL. Something in the tune is switching you to Open Loop when the throttle is blipped a tad. The only thing I can imagine that is doing this is MFA or some other setting that is forcing OL. What I don't know is if the actual OL trigger is TPS or Load. Just looking at the Loads doesn't appear that things are so high that the EEC should be forcing OL.

As the log goes on, I see short intermittent points where the throttle returns to closed and the EEC returns to Closed Loop. During these periods, the LAMBSEs are again responding as expected based on the HEGO values.

The next point of interest in the log begins at 49.15573. Here you see that HEGO1 is finally going lean showing the EEC has found "near stoic" for the Bank 1. However Bank 2 continues to indicate rich and the EEC continues to leanout Bank 2. A normal engine doesn't normally require a big difference in LAMBSE between the two banks. So I keep looking down the long and what do you know? The LAMBSEs return to a more equilibrium the longer the engine is allowed to stay in Closed Loop which means one side is just wetting up more than the other while in Open Loop...and just takes longer to dry off. During that time, the EEC leans out the wet-up bank. Looking further into the log at other periods of long Closed Loop times, Bank 2 shows a pattern of being the bank that takes longer to lean out after being in OL. So I'm not terribly concerned about this.

But at this point, I have more faith in the HEGOs than the WB. You have or more of the following:

• The wrong transfer programmed in for the WB
• Voltage error in the pickup
• The WB controller itself needs calibrating

And you need to find out what's jacked up in the tune causing you to go Open Loop when you step on the throttle.

[tremec77]

And you need to find out what's jacked up in the tune causing you to go Open Loop when you step on the throttle.

I have my TPS voltage set to trigger OL at tip in. I did this because if I allowed CL to operate all the time, it would cause all kinds of drivability problems. I believe it is because of my cam which is a pretty aggressive custom grind.

[cgrey8]

That would be a good plan if your WB was working. But since it isn't, EA is recommending the wrong course of action...to enrich when things are already way too rich, but being reported as lean by the AFR numbers.

[tremec77]

Do you think it could be because I have my WB set for 14.6 to be Stoich and in BE I have it set for E10 fuel which I think is stoich at 14.0?

[cgrey8]

No, your WB is indicating you are in the 18s...if I remember your datalog correctly.

[tremec77]

Well, I did the self test on the Dynojet WB and it came back good. I also double checked my transfer function for the WB that I have in BE and it too is good. Any other ideas?

Also I noticed that in the logs, when I got into OL the WB matches my LAMBSE pretty close. IF my WB was bad I wouldn't think that they would track so close in OL.

[EDS50]

Well, I did the self test on the Dynojet WB and it came back good. I also double checked my transfer function for the WB that I have in BE and it too is good. Any other ideas?

Also I noticed that in the logs, when I got into OL the WB matches my LAMBSE pretty close. IF my WB was bad I wouldn't think that they would track so close in OL.

Do you have EA and your WB configured for e-10 like it is in BE? I just ran into this problem. I didnt have EA or my LMProgrammer for my WB configured for e-10, they were still set to 14.64. CGREY8 enlightened me to check these settings. Here is a link tothe discussion:http://eectuning.org/forums/viewtopic.php?f=2&t=15829. I figured it would be worth a shot for you to check out.

[tremec77]

I will check this out and see what the settings are. I don't think I have EA set correctly now that I think about it. Were do I make the setting changes in EA??

[EDS50]

I will check this out and see what the settings are. I don't think I have EA set correctly now that I think about it. Were do I make the setting changes in EA??

This is what I was getting at. There's more than just setting the Z_AFR scalar. You must set the Z_AFR scalar AND the Fuel Setting parameter in BE. In BE2010, go to Settings>Fuel Settings. There you should see a place to indicate your stoic AFR. Otherwise if you have Adaptive Learning ON, it'll reset LAMBSEs back to 14.64 despite what Z_AFR is set to. With Adaptive Learning disabled, the problems are probably minimal but still important to be aware of. When this was implemented, I wanted BE to reference Z_AFR, if present, and only make this BE setting editable for strategies without a Z_AFR scalar, but Clint thought that was a bit much. So you must change it in both locations.

Next, you have to setup EA to the same Fuel Setting that is set in BE and in your tune. Since the datalogs don't pass along what LAMBSE values represent stoic, EA and BE must be manually kept in sync. BTW, this setting in EA also affects the Injector and Fuel Pump sizing calculators.

And if you have a WB, you'll also need to retask the WB to know that stoic (lambda = 1.0) is 14.08 as well. Using the EGR input or DataQ, this is just a matter of updating your transfer function. If you are connected to the WB serially, this requires you use the WB's proprietary configuration software to change.

Give the thread a read I provided a link to.

[tremec77]

Setting EA correctly fixed it. THANKS!!

[EDS50]

Setting EA correctly fixed it. THANKS!!

Good to hear you got it fixed.