Here a couple of things I’d like to understand:

During PE (and assuming COT is disabled) the PCM will inject a certain amount of fuel ON TOP of what it would need if it was in closed loop.
Example: 1.2 PE means 20% more fuel than stoichiometric = 14.7/1.2 = 12.25
This should be the AFR I see on my wideband.
Correct?

IFR: I have installed a new set of flow matched Bosch injectors. They are built same as the stockers: they are rated at the same pressure and just flow a little more.
I also installed an additional fuel pump in line with original one.
Old injectors: 28.4 lbs/h
New injectors: 36 lbs/h
Old pressure: 55 PSI
New pressure: 64 PSI
The factor to calculate the new IFR is SQR(64/55)x36/28.4 = 1.3674
So I should multiply the old IFR values by this factor.
Correct?

VE: I drove around for a while in SD (MAF fail freq = 0, low octane table = high octane table). Result: all cells between 0 and 15 were +/-2% LTFT.
Can I assume that the stock VE table is correct? In this case I think the new IFR table is fine.
Don’t look at my signature: right now the engine is NA; it only has a K&N FIPK filter and has stock internals. That’s why I assume the stock VE is correct.

MAF table: after putting back the program to work with the MAF I logged Raw Frequency and LTFT. This allows me to make a pivot table in Excel and I can see for what frequency I need to change the G/sec value of the MAF table.
The result is fine, the method works: in closed loop all LTFT are pretty close to zero.
I had to change the stock MAF curve just a little (2-3%) and smooth it down so that it still looks like a curve.

Now the problem: In PE and commanding an AFR of 12.8:1 (see above how) I can measure values in the region of 13.5!
12.8 is about 14.8% more fuel, but 13.5 are only 8.9%!

Where is the mistake?

Is the stock VE table wrong and makes me believe the IFR table is correct?

Should I change the MAF curve in the high frequency range?

This of the commanded AFR is a BS? Should I just add/reduce fuel until is fine and don’t bother about the theoretical numbers?


OK, this was a long post, but how tell those things in a simpler way?

Thanks - Stefano

Your post is quite confusing since you state the engine is stock but are tuning MAF and VE tables, why I do not know.

There is zero reason with stock engine to be doing so.

I'd suggest go back to stock setting for those functions.
Set IFR to vales that would be close which is about 25% higher values then stock values are and set PE/RPM values across the table at 1.145

Do the flashing. clear any error codes, do idle relearn, drive about 30 miles for PCM to relearn and then do a testrun with a scanner recording.
From that data tweak the IFR to get LTFTs +/- 5% and only then tweak PE/RPM values.

From what I see in your post is your jumping all over the place and making one area better but then hosing that with changes that are not needed for the PCM will readjust and once LTFTs are close PCM will adjust properly and tune will be fine.
I suspect what you read on other forums has you chasing your tail.

MAF sensors measure airMASS, even if so called porting is done that does NOT change the sensors measuring ablity one bit.

VE as I have shown are for startup and decel under 4,000 RPMs and as a backup if MAF was to fail, THAT IS IT, so screwing with those tables only forces PCM to calculate functions like torque, engine load, airmass, etc incorrectly and thus adjust incorrectly.

Thank you for the answer,
Actually I haven't changed the VE table: I made a run in SD only to see if the IFR table was correct. Maybe not the best idea

The MAF is also stock and I was thinking it was an option to tune fuel by changing the MAF table (like before with the MAF-translator...)

But I understand what you say, It's probably so that I'm making too many changes in places that I shouldn't touch


Another question: is the IFR table supposed to be a straight line? Or can it be curved so to have more/less fues at certain MAP values?

Thank you - Stefano

The PCM could care less what the "line" looks like so yes the values needed for your IFR can be different

The pretty looking straight line comes from the cyberspace tuners who care more aobut what a line in a graph looks like then what the PCM really does, looks at a value and simply uses it.

Is it a good idea to log for MAP and LTFT and change the IFR table accordingly?

This only for cells 0-19.

I see LTFT values around 0% for cells #10 and 15 and about -8% at idle and cells like #0 and 1, that's why I was thinkin the IFR table could be something different than a straight line.

The other possibility is to change the MAF curve logging for raw frequency and LTFT.

I know both systems work(IFR and MAF), I just should choose a strategy and not change both as you wrote before:

OK, I left the MAF and VE table stock and played only with the IFR (logged LTFT's vs. MAP).

The IFR table is now a curve with a little higher values at low MAP.
The LTFT average is around zero +/- 2%).
The car itself drives as usual but the LTFT are more consistent. Tuning the MAF curve they changed depending on the driving time and on the wheater. Now they remain pretty constant.

The STFT's used to bounce up and down a lot (+/- 7-8%). Now they are about +/- 3%.

The big surprise is the fuel consumption: On the highway at constant speed it used to be much better than on country roads. Now it's about the same. I mean: on countrie roads it needs as "low" as on the highway! This in winter and during short trips.

Keep in mind that as long as the STFTs are within its operating window then LTFTs values are ignored.
Being +/- 2% is just fine so you did a good job on tweaking IFR and as you mentioned its more constant which is good because that puts less work onto the PCM's slow 17 Mhz CPU and gives it more clock time to do other functions.

Its hard to look at gas mileage on cold winter weather but I'd suspect the MPG will increase a bit in summmer for keep in mind winter gas makeup does not give good MPG.

You might try a test of setting IFR a bit leaner across the board, as long as you do not have ping then engine will perform better and MPG will also increase.

Being you have water injection I would not expect a linear IFR table since when WI is on makes AFR about 8% richer thus IFR needs to be commanded with shorter injector pulse width when WI is on.


Lastly keep in mind if you tuned it correctly for better performance then MPG would take a hit but thats the purpose of performance tune right ?

But you can tune it leaner and still get performance by tuning the timing corretly to prevent the ping yet using a bit more fuel to prevent ping and using higher timing makes the power but in any case your on a better tuning path then you were before and the STFT is saying just that.

What is the difference in the USA? Over here I think it's the same

Does it mean that with positive LTFT's the engine is running light lean to stoichoimetric? And with negetive LTFT's it's light rich to stoichoiometric?

WI turns on at 2 PSI and off at 0.5 PSI. For my particular (low boost) setup this additional enrichment happens in open loop only.
I was able to tune the IFR only in the closed loop range (20 - 90 kPa), over 90 kPa I just set it as a linear continuation of the curve.

I now have to check what happens in PE...
How does the PCM calculate the IFR under boost?
My MAP only sees 105 kPa max, the IFR goes only up to 100 kPa. Does it keep the last value @ 100 kPa? Does it extrapolate up to 105 kPa then it's linear?

I ask because there's something I don't understand:
After 105 kPa I think the PCM keeps a constant value. In reality there's less fuel flowing because of the backpressure in the intake manifold. But the PCM doesn’t notice that.
At the same time the MAF is checking for the amount of air and commands the injectors accordingly.
With constant PE vs. RPM values I should see the AFR rising proportional to the RPM (and to the boost).
In reality the AFR remains constant (that was with water alone. So the AFR wasn’t affected by the alcohol).
Why?

Something similar happens when the supercharger is disconnected: with constant PE vs. RPM values the AFR at WOT decreases at high RPM. Linearly.
Can I assume that my MAF is seeing more air than what really flows?
In NA and at WOT I see a constant MAP of about 90 kPa, so this is not something I can adjust with the IFR.
Is there some other factor that influences the AFR at WOT?
COT is disabled and the temperature is 190 – 200F. What else could it be?


Stefano

In USA due to smog rules during the winter the gas makeup is different then in the summer since engines run richer in colder weather so more O2 is in winter gas and make have more Ethanol ( corn)which produces less MPG then summer grade which has more gas then corn.

With Boost and if over 1.2 grms/cyl as example which is the max for timing table then PCM just uses what the highest value its table supports.

Being engine is at 100 or 105 KPA still means engine is in WOT and O2s are ignored and so are the fuel trims so its simply functioning on what PE tables values are

With boost that means both PE tables of RPM and temp have to be adjusted.
Not sure what you mean but IFR functions up to 100 KPA ( as per LS1edit) but other tuning tools only show table to 80 KPA which I think really for them is vaccum values

Again as to fue trims, STFTs are what PCM adjusts IFR and not LTFTs and if STFTs are close to zero then its a mute point as to effects to overall AFR.

I think you may be looking too much into this for as long as NON WOT has fuel trims close and during WOT you have good timing and little or no ping then fine tweaking tune buys you nothing.

Consider EPA (smog) they feel no problem as long as fuel trims do not exceed -13 and +22 so in your case with +/-2 of fuel trims your pretty well tuned in.

This is what happens at WOT (non supercharged): the blue line is the commanded AFR (1.145), the red line is the result (smoothed).
Why is it a curve? I was expecting a parallel to the commanded AFR, but not a curve.

As you stated COMMANDED, it commands but that is not what the AFR will be due to other conditions

In this case it is most likely timing scaled to go lower as RPMs increase which would cause AFR to be richer

If your running boost I suspect upper RPMs are set for low timing

As said right now the blower is disconnected and I only have a FIPK filter (similar to the blackwing filter for the corvette).
VE table is stock, Timing table is stock (28* at WOT), MAF table is stock.
Only the IFR table has been changed, but at WOT I saw a constant MAP so the PCM is always using the same IFR value.

That the measured AFR is lower than the "commanded" is not so strange (IFR is maybe still not right), I just don't understand why it makes a slope. Like if the MAF was seeing more air than what it was.

Unless there are other factors that influences fueling that I don't see in Edit.
There is a PE vs coolant temperature, but at 200F the correction is zero.
The measured timing was also 28* without KR.

I know: the AFR is not too much out and I could correct it in the PE vs RPM table and call it a day. I just was wondering why it's a curve and not a line...

Thanks - Stefano

Short report

It's a while now that I changed the IFR table and the tune is still OK.
LTFT change a little depending on wheater conditions (go a little positive when cold) but that's OK.
What I did was log for LTFT, fuel trim cell and MAP.
LTFT are averaged, MAP is rounded to the next MAP value in the IFT table.
I used only cells 0-19.
I then put LTFT ans MAP in a pivot table (Excel).
The result is a correction in % of the specific IFR value.

So far so good!
I used to tweak the MAF table but never with good results. At least not as good as only changing the IFT table.
A side effect of changing the MAF table is a wrong G/cyl value which affect timing also!!!
I used to reduce the MAF values to adjust fueling: less G/sec means less G/cyl and in general higher timing than necessary (PING!)

Here the actual IFR curve: an untypical shape, but LTFT are now +/- 2% with an average of almost zero

Stefano

Good, glad to see your getting purged out of what you see on forums saying to screw with MAF and VE tables when as you saw it just blurs the results in scanner data and gets you no where to having clear fuel trims.

You cannot blindly use cells 0 to 19 as a bulk value since some cells are used for decel and fuel learn is off thus your trim average is bloated since its adding into the trim average when PCM is commanding fuel off thus would be leaner on decel but being added to when fuel is commanded on

I was thinking only in cell 21 the PCM will turn off fuel learn (in fact I see an AFR of around 13 on the WB but the LTFT remain constant at about -2%). During fuel cut off it's another story.
I also see cell 0,1, 4 and 8 in decel (mostly 0 and 1) but the AFR is stoichiometric (so I assume the PCM is learning).

What cells would you consider? I tried 0-19 and 0-15 and the result is exactly the same.

For MAP over 85 - 90 kPa it's quite difficult because it goes in PE, so I just continued the line by feeling (values 90 - 100 kpa). Is there another way to determinate this last piece of the IFR curve?

Stefano

Its not black & white on decel since engine changes have moved vaccum around, DEFCO or cell boundaries were changed via a tune, etc

There are really two conditions, certain cells like 21 are large lift of gas pedal but other cells are smaller lift but in both cases if not set to a conditional math in Excel would be added as if they were part of true LTFT averages.

Only setting conditions in the math in Excel can you find true values but in the end, again of trims are within -/+ 5% of STFT then PCM is ignoring LTFTs so you can chase a ghost that is not there.
As to WOT there is no performance difference if AFR is 12.4 or 12.6 so as long as there is no ping and timing is close to around 28 degrees for a NA engine then tweaking tune any more does not buy you anything.

Lastly cars are not driven like being on some chassis dyno so it must be considered that simple movement of gas pedal can skew trim averages but mean nothing to overall performace.

OK, I managed to separate the decel values in Excel.
It's true: it doesn't follow the cell numbers.
Mostly cell are 21, 0, 1 and 4. But not only.
Result: the same IFR line as I posted before.
I'm happy with the result, I just don't understand the shape of the curve at low MAP. It just doesn't make sense (to me).
Do you have an explaination?

Stefano

Not sure what your asking since as here is an example of a testrun and as engine load increases so does airmass which increases grams/cly to injector ON time

I just wanted to say that this IFR curve is a little strange in the low MAP range.

As you said Excel would average everything (also during decel)so I changed the spreadsheet to consider only the scan data related to acceleration and not decel.
Considering acceleration only doesn't change the IFR curve, so I guess this is what my engine wants.
(This curve is the result of a 2 hours scan, using only cells 0-19, cell 15 only when not in PE).
I was just surprised to see lower G/sec at 30 kPa than at 20 kPa.

Stefano

Stock engine at idle around sealevel is around 30-35 KPA Map. Compare 10-30 KPA when fuel learn is ON and you also see less then 30 KPA, torque less then 30 lbs to negative where 30 plus KPA engine is pulling more engine load and timing to maintain idle

Your values almost look like your scanner is backwards and not KPA but in inh

Being 100 KPA MAP would be WOT ( low vaccum) and 20 KPA ( high vaccum )

We are at about 500 m here (1640 ft), at hot idle I see about 30 kPa.
With AutoTap I can't see if the fuel learn is on, or at least I don't know how (do you know if Autotap can do that?)

I adjusted the values to fit the LS1Edit IFR table. It's easier to compare / change the values. According to the scanner the pressure is in kPa (??) and not in inHg

Ah, that says a lot. If your going to trust your tune changes to what your scanner tells you can be a mistake.
I quit using Atap 5 years ago when I saw too many things wrong. Fuel learn is a EPA PID since they dictate when trim values are averaged in or are ignored and you need to know what PCM has either commanded injectors off or went to min pulse width.

It is an important PID to know when engine is going into decel or is in decel or when PCM is truely into power enrichment mode

Ease Enhanced OBD-II Powertrain Scanner ?
Are there other producs you would suggest?

I see what you mean by "you need to know what PCM has either commanded injectors off or went to min pulse width". Right now I see LTFT / STFT values that don't make too much sense.
After a while I'm at idle the fuel trims go negative without an apparent reason (?)
I have #36 injectors and at minimum pulse width they could still put too much fuel in the engine.

Also the averaged cells change depending on the wheater. It was 0% at 60F and +2/+3% at 40F.
I wonder if this is the result of a wrong tune or of wrong data from the scanner (???)

What a mess!

I suggest the Ease scanner because they are a tier 2 suppler to GM so their scanner was designed with GM engineering input and information that is not known by the public for their enhanced PIDs plus how the VPW functions.

If interested you can purchase it via our frontstore and this is the model you would want :

Ease GM Enhanced OBD-II

Your autotap does less then about 10 frames a second where with Ease you get 40 F/ps.

In your case PCM cycles are then missed where with the Ease your getting far more data cycles.
Also the Ease supports all GM PIDs, has bi-directional, full EPA IM snog tests and O2 tests.

OK, let me think about that.

Here another question: is there a natural reason of having different LTFT's depending on the wheater? Temperature? Air umidity? Pressure?
Or can it be related to a tuning problem?

The difference between 30F or 70F is 4 - 5%
Is that normal? Should I just take something in between?

Sure weather will have a effect.
Try using the weather calculator on my website,
direct is Weather cost to performance

Make weather changes with it and see the effect it has. Hot air adds to the correction factor of HP/Tq which is a loss of performance where colder airmass would increase performance so it clearly has effect to AFR and fuel trim values

You see a hugh effect to weather to AIR effect on performance.

Clearly you know in summer heat engine does not perform as well as nice fall or spring weather so air quality effects AFR and reason why using the MAF is much better adjusting tool for the PCM then those people deleting the MAF and use VE speed density as to how PCM determines how much fuel to add to qualty of airmass.

As to tune corrections I slant on the lean side as long as there is no knock pulling timing so there is no rule that LTFTs must be a perfect zero being as long as STFTs are +/- 5% PCM ignores LTFTs and does no correction to them

Interesting calculation! As I expected Dewpoint has a big influence... This is the only real problem of a MAF: it measures mass. Air mass but also humidity mass.

So with umid air the MAF will "see" more air than what it really is, causing more fuel to be injected. At this point the O2's will notice the difference and cause negative fuel trims.
Right?

In my company we have a couple of gas flometers: they are similar to a MAF in their function.
There's no chances to have a correct reading if we don't have a drier installed before them!

So what to do? Install a big drier in front of the car?

I do not see that as a problem for in fact that being MAF can sense humid air and adjust for it assures the PCM is properly adjusting for the quality of air.

PCM is not a perfect realtime monitor and adjuster since it only has to make changes over time and as long as PCM keeps AFR to within 1.5 times the FTP ( Federal Test Procedure) rate then all is well.

Really having the fuel trims ramp up/down (rich/lean) is good for that assures the O2s are properly also ramping above and below .450 mVolts.
If fuel trims stayed flat then O2s also would and that would confuse the PCM thus as long as STFTs are within that +/- 5% then engine is running perfect and forcing flatline of fuel trims would not make any more performance.

Take a simple driving up and down a hill, if PCM was realtime the trims would be lean, then perfect 0 and then rich going down the hill. If PCM realtimed those changes the engine would not run smooth that driver could feel and why PCM only reacts to LTFTs once every 10 minutes and once a second looking at O2 values.

Is it better to consider LTFT and STFT together?

correction = (LTFT B1 + LTFT B2)/2 + (STFT B1 + STFT B2)/2

This assuming that x% correction for STFT has the same weight for LTFT also.

Example: +4% LTFT and -7% STFT means a total adjustment of -3% (for a particular FTC)

OK, I'll answer myself: see here the result of 2 very long scans (3 hours each, visiting my parents for Easter).

Notice the IFR curves calculated from the LTFT's: they are different.

The other 2 IFR curves are the result of considering LTFT's + STFT's and are the same!
This means (to me) that the formula above (averaged sum of LT and ST) is correct.
As stated by JR LTFT will change only if STFT are higher/lower than +5/-5%. Considering LTFT alone can lead to wrong values (as I personally experienced!).

The 20-30 kPa area is pretty strange

I also wonder if and how much the EGR influences the whole story

Stefano

No there is nothing gained from adding short and long together.
20-30 MAP KPA would be funney since it mostly decel and low RPM range and sometimes fuel learn is off or on or just having switched from one to the other.
You'll gain nothing by doing all this since PCM is not perfect and many different driving conditions such as going down a hill from full to partly decel so numbers will never be perfect but again as long as trims rotate a bit lean and rich from zero assures O2s can also ramp up and down crossing 0.450 mVolts

If always a perfect zero then O2s would stay constant and that would confuse PCM.

Oh, I wasn't looking for performance but I wantend to understand how this part of the PCM works.
Since I loaded the new IFR curve (considering LT and ST) and resetted the PCM most LTFT are = 0 and only few are +/- 2%. STFT are most of the time +/- 3/5%.
This for different driving conditions (accel/decel) and outside temperatures (40/70F).
When I only considered LTFT, driving style and temperatures had a bigger influence.
But it's clear: as long it runs in closed loop it doesn't make a difference: the O2's will manage that

Yep as long as the trims fit within the STFT operating window tune is fine and long term changes then are just clean up to the PCM over time