fuel tuning for 3v mustang

[stangslayer]

I've got a 2005 mustang gt that I just installed the on3 70mm turbo kit onto and am have a few question about tunging the fuel curve to address my lean condition under load. I have installed 60lb seimens deka injectors, mafia and 340lph aem stealth fuel pump as supporting mods

I have set the injectors up using specs from ford racing 60's as no one seems to agree or even have all the specs for the SD's so that may be a portion of my problem but I had to start somewhere.

I have not done anything to account for the new fuel pump. I haven't noticed any tables that seemed appropriate to update due to the high flowing pump. It flows 340@43psi and I believe the mustang is suppose to have 39psi stock. If anything this should be giving me a richer fuel mixture but I would like this to be correct if it needs to be addressed.

I suspect my issue lies in the MAF. I am using the stock MAF with the help of the mafia range extender set on 3. The correction multiplier I've listed below comes straight from diablo but multiplying mag maf value by .59 seemed like it would lean out the mixture even further. The stock setting at 5volts is something like 303g/. what I did to scale the table is multiplied it by 1.41 to bring the max value (5 volts) to 427g/ so it would register more airflow at each voltage level. Did I do that correctly?

Setting Correction Factor Range Notes
0 No tuning required Stock Stock/No Change
1 No tuning required Stock minus 5% Leans Signal
2 24% or multiply by .76 5.5v – 5v or .90 Tuning Required
3 41% or multiply by .59 6v-5v or .83 Tuning Required
4 53% or multiply by .47 6.5v-5v or .77 Tuning Required
5 62% or multiply by .38 7v -5v or .71 Tuning Required
6 70% or multiply by .30 7.5v – 5v or .67 Tuning Required
7 No tuning required Stock plus 5% Richens Signal

To make matters worse, the charge pipe that I have the maf installed in is 3" diameter and I suspect the stock intake piping was larger which is further throwing my maf calculations off. Once I find someone who will let me measure the stock diameter I plan to use an "area of a circle" calculation to find the difference between the 2 and multiply my maf files accordingly.

I have only done a couple miles of street driving to figure out what areas need tuning it seemed that as the engine warmed up the lean condition worsened. At idle I had 11.8-12.2 which although rich I though was at least a safe place to start. I was only able to accelerate under very light load (not even enough to get to drive speed limit) where my wideband was showing 14-15 afr. Any harder and afr would jump to 18.

I have read else where it is a good idea to tune in open loop before enabling closed loop. Is this a good strategy? From what I can see (newbie tuner), the main table for influencing the fuel curve are the "maf airflow vs voltage", and the "injector offset" (I'm at work and going off memory). I know there are multipliers and tip in's but are these the 2 main areas to get it to at least stay under 18 afr?

Can someone also please explain "Load with failed MAF"? I don't like to change things unless I have at least a rough understanding of what I'm doing. Throttle % and rpm are the 2 axis but I cant figure out what the values are that make up that table.

thanks,

[fullthrottle]

if you dont have correct injector data, its a certainty that it will be a problem. all fueling requirements rely on the correct injector data to begin with.

[stangslayer]

thank you for replying.

Currently I have the published Seimens Deka 60 values entered to get me started. I am expeirimenting with various changes but some of the scanner feed back is throughing me for a loop.

When my car starts up in open loop for the first 15 seconds my wideband shows it running very lean. once it goes into closed loop and the afr comes back around 14.5 it is showing a commanded lamba of .6 to .8 but my narrowband O2's are usually over 900mv (which in my understanding means they are seeing a rich mixture) but my short term fuel trims are maxed. And also the "O2 sensor voltage'' and ''O2 sensor voltage SAE" rarely agree and many times are opposit each other. How can this be? Which one do I use?

Thats all for now. I plan on doing some test drives tommorrow as long as i can get on the same page as my scanner.

[fullthrottle]

the o2 sensors see anything to the left of stoich (14.7) as rich. and anything to the right of stoich (14.7) as lean hence the term "narrowband" 02 sensor. comparing the narrowband o2 sensors to your wideband does no real justice. I cant open your log file. Are you using the beta version of HP Tuners or 2.24?

[PST]

First, the Ford racing injectors absolutely are the seiemens deka injectors. If your data is correct as in straight from ford, for the correct injector, and no copying errors, then your good.

After startup fuel.. look at base fuel cold table, and for the same scenario warm, look at base fuel table. These are your target EQ (basically lambda) values. The ECM is trying to hit these values using injector and MAF data. If it's not hitting properly, then one of the 2 is most likely wrong.
I always use the non SAE pids whenever possible. The SAE pids have always seemed slow or delayed to me.

Now to your first post...
You can log fuel pressure and most of the data needed to check your fuel pump. Your car uses PWM to control the fuel pump and obtain desired pressures. Just because a pump is rated at 43psi doesn't mean that's the pressure your getting.

Turn the MAFia off for a little bit.. using fuel trims dial in your low airflow regions of the maf table. Not all points will need the same correction, but from them you can pick an average correction percentage that can be applied to the rest of the table. This will be nowhere near correct for the high airflow boost regions, but should get you closer than a stock table. After that re'enable the MAFia and apply your correction factor to the MAF table, hi and low inj slopes, cylinder displacement, etc.

At this point setup a histogram to log wideband LAMBDA vs commanded lambda against your MAF volt scale. Force open loop, then start logging. Apply your histogram lambda error factor to the maf table, rinse repeat, etc. Once you have your actual lambda hitting within 2-3% of commanded, you have your maf calibrated pretty good. Try to make it error on the rich side with all your error numbers being negative. Then allow closed loop fueling and see where your fuel trims go. If they are off, it will need to be addressed with a diagnostic approach.

Load with failed MAF gets filled out AFTER your fuel and timing are all finalized. Setup a histogram to log the pid AIR LOAD against the same rpm and tps scales used in the load with failed MAF table. Copy and paste.

[PST]

Almost forgot. When tuning your MAF, you'll want to turn off the MAF and cyl air anticipation logic stuff. Don't re-enable until after you've put it back into closed loop and verified correct fuel trims.

[stangslayer]

.

Turn the MAFia off for a little bit.. using fuel trims dial in your low airflow regions of the maf table. Not all points will need the same correction, but from them you can pick an average correction percentage that can be applied to the rest of the table. This will be nowhere near correct for the high airflow boost regions, but should get you closer than a stock table. After that re'enable the MAFia and apply your correction factor to the MAF table, hi and low inj slopes, cylinder displacement, etc. .

I can't find a table for fuel trims so i am assuming you are referring to the "high and low flow rate multiplier"? Your reference to Hi and low injector slopes also sounds like the same table so I am sure I am wrong for at least 1 of my assumtions lol.

PST, I am using 2.24

[PST]

Fuel trims are logged. Short term and long term. You need to do lots of searching and reading.

The fuel trims are used by the ecm to correct fueling. You will use the logged correction factor to modify the maf table

[stangslayer]

I have verified all injector settings match fords specs and my maf function is multiplied to match the MAFia setting with not other adjustments done from the stock profile. My fuel trims are pretty close wiith 2-5% leaning of the STFT at various no load rpm positions. I would think im at a good starting point if it wasnt for a lean spike in afr every 3 seconds if i bring the rpms up to a steady rpm around 2000-2400. Everything looks good and stable for 3 seconds and then SPIKE, dip and recover back at 14.6 and repeat 3 seconds later and so on in that pattern. When the AFR spikes I feel and hear the engine miss and fall on its face for a split second almost like the injectors forgot to fire for a moment. Does this sound like a tuning issue or a mechanical malfunction of the injector or mafia?

I am not able to completely remove the mafia as the stock harness doesnt reach the maf location so I retested with the mafia set at 0 and maf file corrected back to stock and the problem persisted.

There are 3 obvious spikes towards the end of the log file that are the best example of this situation.

[PST]

You don't have to remove it, just turn the nob to zero

However that miss is most likely an ignition issue. My bet is a coil. Seen that exact scenario many times.

[stangslayer]

I didn't have the lean spikes before the turbo went on the car so I am going to continue getting the tune to a better start and see if it goes away as the tune improves.

After a carefull test drive to get more data on the state of my Maf curve I am a bit confused. My computer won't screen shot so I had to use my phone for the image below. This is near the top end of a 3rd gear pull that I shut down at 3300 when the afr still wasn't falling to the comanded 12.05 that it should have been at for 2500rpm and 63% throttle. Looking at the data now I see that my StFT was pulling fuel and the "commanded lamba" was being corrected to 1.15 instead of the .82 value it should be reading. I read this as meaning my Maf curve is asking for too much fuel and so it is being trimmed, execpt I actually DO want that much fuel! I must be wrong in my rationality but why is fuel being trimmed?

[stangslayer]

The event i described above is at 5:26 in the attached log

[PST]

It's not going into open loop soon enough. There are adjustments to be made for this, I'll dig through your tune tomorrow morning for sure.

[stangslayer]

After that re'enable the MAFia and apply your correction factor to the MAF table, hi and low inj slopes, cylinder displacement, etc.

Can you please explain the reasoning for applying the same correction factor to the hi and low slopes, cylinder displacement and other areas i am unaware of? I have only applied the correction factor to the maf table and the car runs whell when warm, butI am fighting a lean tip in condition as I posted in another thread so I obviously missed doing something. Maybe this is it, but i dont see the reasoning behind it.

If before the addition of the mafia, 1.52v was = to 25g/sec, and after the addition of the mafia 1.08v was = to 25g/sec all that matters to the ecm is that it needs to supply fuel for 25g/sec of air. right? Please correct me if my theory is flawed.

thanks