Evo8Emperor

iTrader: (12)

Im guessing that you mean the smoothing table right ? I thought it changed also when you change the scaling because mine did.

I had to change some other things on the address's in the definitions. Now I have it reading to the tenth like jcsbanks for the sd patch.

I checked my non sd map to see if all the tables look fine and they do so there should be nothing wrong.

jcsbanks

Note that these tables do not affect load, so will not rescale your timing, open loop thresholds, knock thresholds etc. They are part of the airflow calculation that is multiplied by the AFR in your fuel map and your injector size to calculate an injector pulse width. [Other compensations are also in there such as coolant enrichment, acceleration enrichment, air temp comp, baro comp, knock comp, STFT, LTFT.]

03whitegsr

iTrader: (8)

I renamed the "MAF Scaling" and "MAF Smoothing" tables to "MAF Characteristic Scaling" and "MAF Compensation Scaling," respectively

I did this because to me, it looks like the MAF Scaling table defines the expected characteristics of the MAF and has the additional 140 adder. The MAF Smoothing table on the other hand looks to be a correction that was likely used to tune in the MAF from the factory to accomodate deviations from the predicted model they likely came up with of the MAF. Or maybe even fine tuning different MAFs for different models?

Scaled as percent (128) and it makes MAF scaling easy to do mathamtically as a 10% change in the the values gives you a 10% change in AFR.

I rescaled my MAF using MUT32 and actual AFR and got them to match based on averages with this method. The car drives better when warmed up now, but it runs leaner on cold start.

SWOLN

iTrader: (6)

I'm sorry, I feel so stupid sometimes, and I hate asking for help, when after it ends up being simple, but can someone look at my log. My STFTs are all over the place. I thought they'd be more steady, so that I can add the STFT and LTFT(LOW). Is my O2 bad or something? Thank you.

mplspilot

iTrader: (7)

Your STFT fluctuates between -5 and +5. Just use your LTFT Low value of 11.8 (12%) to start with.

Appauldd

iTrader: (22)

STFT is supppsed to be all over the place. It is the "now" adjustment where the LTFT is the long term effect of the now over time.

How that makes sense.

Based on the LTFT low being +11 you might consider changing your scaling in the 12V just a tiny bit.

SWOLN

iTrader: (6)

Thanks to both of you. Sending PMs so as not to clutter this thread too much.

XK120

iTrader: (18)

Sorry for the "noob" question, but at what point in the tuning process should I do this? Based on post # 24, it sounds as though one should rescale before actually tuning the car as it results in an improvement in overall accuracy. Also, I assume this should be done if I am using an aftermarket intake system...correct?
Thanks

shadow1

iTrader: (38)

yes and yes

sti.this

When you say you scaled as a percent, which table are you referring to? Scaling? Smoothing? And are you editing the definition to be scaled as a percent? Or doing some calcs on paper, xcel, whatever?

fostytou

iTrader: (15)

http://thefrost.net/randomfiles/tuni...g_Trending.JPG
(large teaser image)

Note, I'm putting together an example for people to look at to get an idea of how to scale the MAF at higher airflow frequencies. I need to log more before its complete, but its a really basic standard log, tune, repeat process. This should hopefully help get some of the MAFs in line and make cars more responsive/better gas milage/more accurate and easier to tune.

I'f I'm taking a while somebody kick me.

I'd like to add delta TPS in and a few other things, and it won't be the easiest to use... but someone can pick it up where I leave off if you want. Hope to have an update soon.

Appauldd

iTrader: (22)

The percent change is for the MAF Smoothing table. Ths table should be renamed too. Something like "MAF Adjustment".

After you use the MAF scaling to dial in your injectors. The MAF Adjustment (smoothing) table will draw you in to near perfect 0,0 LTFTs. It works amazingly well.

fostytou

iTrader: (15)

Ok guys, here are my first couple of sheets. So far with some mild driving today things seem to be running much smoother, but I'll report back if they stay that way.
(please note that one of the sheets is ~35mb, alot of cruising data)
http://thefrost.net/randomfiles/tuni...omb.cruise.xls
http://thefrost.net/randomfiles/tuni...p.FULL.WOT.xls

It may look silly, but here are the differences in my tables. I'm likely going to try to reset this table to "stock" and try to apply the same corrections to the MAF Compensation table as 03whitegsr suggested, just to see if results are any different or better.




A couple of quick notes:

-I divide actual AFR by AFRmap (because AFR and increasing the "weight" of the MAF value are inversely proportional) and then add LTFT current.
-To get a trend line just click any data point in your scatter plot and select "add trendline". I normally do polynomial to the 5th order for a smooth but accurate graph.
-To find what actual value to apply (see the larger cruise sheet's first tab) I simply pick points on the trend line rather than guessing what I should be putting in. Remember, we are going for accuracy here.
-I do ALOT of sorting and removing outlying data points. I also immediately remove any cells with actual AFR values richer than 10 and leaner than ~18 (would go even closer if this was a pump map).
-The more conditions and data you have, the better your trend lines will be.
-My car idled a little silly (even making strange lopy-er idle noises) after I flashed the new map, but after a couple of minutes of driving things were smooth sailing.
-note that if you have air leaks (open breather, boost leak, etc) your trend lines will be very good averages, but the still can't be "accurate"
-You should have your tune very well in order before getting to this... this should kind of be a final "touch up" step.

Beyond that I'm not going to add too much detail, but feel free to ask me anything you like.

shadow1

iTrader: (38)

Do you have the XML scaling equation for Percent128?

fostytou

iTrader: (15)

I would guess (I'm not sure) it is
Matt/03whitegsr... are you suggesting that we should not use an adder and should just be able to apply a 6% correction to the MAF Compensation table as (for instance)
95 - 6 = 89% (if scaled as Percent128)? Seems simple, I'm just trying to be very clear here. EDIT: ignore the next sentence----It seems like this would be near twice the weight since before we would calculate based on something closer to a %255 weight.

Is this because the adder is not applied to the weight of this table's calculation or something else I'm not thinking of?

Thanks for the tip!