MAF Scaling - Open/Closed Loop Fueling info

[Kodename47]

Originally Posted by steve99

Some jokers play around with iat/maf compensations

Thanks :p


I actually tried something yesterday that @mad_sb suggested a while back; Running full time OL just to see how good my MAF scaling was, especially as I had encountered an area where the 2 were vastly different (as described in a previous post). The curves were actually very well matched, so for those who spend run after run doing lots of CL testing to see how accurate their MAF scale is, I suggest that once you think you're close you run in full time OL and then leave the MAF after that. Obviously you can then return the maps to run CL again.


Here's the original thread for full info:
http://www.ft86club.com/forums/showthread.php?t=57279


Taken from that thread....
How to Force Full Time Open Loop:

1. Open your Fuel Map
2. Select all cells that are 14.7 and set them to 14.6
3. Set your Minimum Active Primary Open Loop Enrichment to 14.7
4. Set your CL to OL Delay timers to Zero
5. Save and Flash rom
6. Make sure your logger is setup to log commanded AFR as well as measured AFR

* NOTE - For those on ECUtek and don't have access to Minimum Active Primary Open Loop Enrichment, the stock value for this is 14.0, thus setting anywhere in the fuel map to lower than that will have the same result.


Use the same technique as above for Non-WOT OL scaling to get your results.

[steve99]

Originally Posted by Kodename47

Thanks :p

Rephrase " Some smart jokers"

[s2d4]

Why would you want to run full time open loop?

Are there any benefits?

[AL008]

This is my log:

http://datazap.me/u/toy-sooby/maf-re...g=0&data=1-4-6

Weird... My LTFT is 0% throughout whole log!

[steve99]

Originally Posted by AL008

This is my log:

http://datazap.me/u/toy-sooby/maf-re...g=0&data=1-4-6

Weird... My LTFT is 0% throughout whole log!

Yep you have done a good job with the CL maf scaling, the ltft =0 looks correct as if you look at the stft its oscillating + and - aprox the same values either side so its average is zero over time hence your ltft will be close to zero.

The commanded AFR and measured AFR are also following each other under load (on overrun ie closed throttle the difference is normal) so looks good

you might see a couple of percent on the ltft as the IAT temperatures change 20 degrees or more but thats normal

[AL008]

Originally Posted by steve99

Yep you have done a good job with the CL maf scaling, the ltft =0 looks correct as if you look at the stft its oscillating + and - aprox the same values either side so its average is zero over time hence your ltft will be close to zero.

The commanded AFR and measured AFR are also following each other so looks good

Thanks @steve99 for your great help and thanks @vgi for such a great program!

I noticed afr of 12 is that considered dangerous?

[Kodename47]

Originally Posted by s2d4

Why would you want to run full time open loop?

Are there any benefits?

@mad_sb noticed he thought the car ran smoother and TBH, so did I. I tried a full time OL map on mode 4 and the same map on mode 1 but running CL. I suspect that my CL-OL transitions haven't been perfect and wanted to check my MAF scaling in the transfer areas. This can only really be done well in full time OL. I won't run it all the time however.

I guess if you know that it's always hitting targets then running full time OL is possible, but then STFT aren't active so you won't generate LTFT and this could be detrimental with seasonal and fuel changes. Reading RomRaider and other Subaru tuning sites, some people do run full time OL. I guess you could give it a go and keep an eye and log it and see what you think.

Originally Posted by AL008

Thanks @steve99 for your great help and thanks @vgi for such a great program!

I noticed afr of 12 is that considered dangerous?

AFR of 12 where? What makes you think 12 is dangerous? No AFR is dangerous until it results in knock. TBH around 12 isn't dangerous at WOT, it's right on the boundary of best power AFR and being rich, which is ideal. For reference, there is a "typical" value of 12.6-12.7 that is a rule of thumb for best power. However this is more of a vague guide rather than a set number, as you trade off AFR produced power for ignition timing produced power. So long as it's not knocking, it's safe.

[steve99]

Originally Posted by AL008

Thanks @steve99 for your great help and thanks @vgi for such a great program!

I noticed afr of 12 is that considered dangerous?

The measured AFR under load should follow the commanded AFR, the commanded AFR is the AFR value from the tune table that the ECU is targeting at a given engine load/rpm. At higher RPM under higher load the tune will target AFR's in the mid to low 12 very high rpm may even drop to low 12 high 11's. (have a look at the OL fueling tables in RomRaider)

14.7 is stoich and is usually targeted during cruise and very low loads for max economy

Max power is usually around 12.5 -13 afr on petrol NA motors

low to mid 12's are considered on the rich side but these motors need to run quite rich as they are knock prone on our fuel anyway , its always a trade off more timing usually means more power (up to a point) and running richer than around 12.5 afr tends to loose power but provides some extra knock resistance, so experienced tuners will trade advance and AFR to find the best power at higher rpm under high load.

[AL008]

I am getting some LTFT around -5.47/-4.69/-3.91% around the following parameters:

Engine Load: Approx 0.3
MAF Voltage: 2.0-2.15
Engine Speed: Approx 2800-3000

Is it advisable to remove any LTFT that is in the good range in my data logs? i.e. around 0% and only process the problematic areas?

Also, how does temperature variance affect the LTFT?

i.e. If I data log and modify this at 60 degrees F compared to another day doing this at 80-90 degrees F?


Thanks.

[steve99]

Originally Posted by AL008

I am getting some LTFT around -5.47/-4.69/-3.91% around the following parameters:

Engine Load: Approx 0.3
MAF Voltage: 2.0-2.15
Engine Speed: Approx 2800-3000

Is it advisable to remove any LTFT that is in the good range in my data logs? i.e. around 0% and only process the problematic areas?

Also, how does temperature variance affect the LTFT?

i.e. If I data log and modify this at 60 degrees F compared to another day doing this at 80-90 degrees F?


Thanks.

Temperature effect things , there is another table IAT vs MAF V which is temp compensation for maf also air density changes with temp so ltft and maf scaling always varies with temp , the best you can generally do is getting ltft below 3% at the temperature you did your logging, it always best to do logging around 20 to 25 C as this is where the least temp compensation is applied to maf. at vastly different temps the ltft is going to vary a bit especially at low rpm in traffic on hot day where iat climbs. that sort of thing is why they have fuel trims to make minor compensation for differeing conditions and fuels such at e10 fuels, it also important to run same fuel.

If thats the only point thats a problem you can just manualy take say 3% off the values in that area of the maf scale to compensate for the negitive ltft at that point try to keep scale smooth no chunky bits :-)

[AL008]

Originally Posted by steve99

Temperature effect things , there is another table IAT vs MAF V which is temp compensation for maf also air density changes with temp so ltft and maf scaling always varies with temp , the best you can generally do is getting ltft below 3% at the temperature you did your logging, it always best to do logging around 20 to 25 C as this is where the least temp compensation is applied to maf. at vastly different temps the ltft is going to vary a bit especially at low rpm in traffic on hot day where iat climbs. that sort of thing is why they have fuel trims to make minor compensation for differeing conditions and fuels such at e10 fuels, it also important to run same fuel.

If thats the only point thats a problem you can just manualy take say 3% off the values in that area of the maf scale to compensate for the negitive ltft at that point try to keep scale smooth no chunky bits :-)

So just to get this right..

Decreasing values from the MAF scales is required for negative LTFT and
Increasing Values is required for Positive LTFT?

[Kodename47]

Originally Posted by AL008

So just to get this right..

Decreasing values from the MAF scales is required for negative LTFT and
Increasing Values is required for Positive LTFT?

Yes. Think of it like this:
Positive trims: MAF value is too low - ECU "see's" less air, adds less fuel - trims go positive to add in more fuel
Negative trims: MAF value is too high - ECU "see's" more air, adds more fuel - trims go negative to remove the excess fuel


Hence if the trims are -5%, you have 5% too much fuel so you can just reduce the area by 5% to re-align things.

@steve99 is correct, temperatures do make quite a difference. If you're logging in varying temperatures, it might be worth zeroing out the IAT compensations table while you do the MAF and then return it after.

[troek]

maybe this is in here somewhere and i missed it, but ill ask. So im getting virtually 0 stft and ltft while wide open throttle, but my afr is a bit low, i ran vgi's scaling app and its still suggesting a new maf curve for open loop. im assuming this is to bring the afr closer to the targeted value?

[steve99]

Originally Posted by troek

maybe this is in here somewhere and i missed it, but ill ask. So im getting virtually 0 stft and ltft while wide open throttle, but my afr is a bit low, i ran vgi's scaling app and its still suggesting a new maf curve for open loop. im assuming this is to bring the afr closer to the targeted value?

Hi long time no see,

Yes VGI utility will suggest MAF scale to bring your OL AFR in line with the OL Fuel table in RomRaider.

I think shiv alters the maf scale to run a bit richer than what in the OL fuel table for safety and knock resistance.

If you do lean out the top end just keep an eye out for any knock.

your fuel is very likely better than ours so you might be able to lean it out to values in OL fuel table with no knock.