crimson red

Golden,

>You can log the Load that comes from the MAF and MAP separately to see what's going on with your tune.
Oh! Golden!! Thanks you very much for the nice idea! I'm eager to log with your way. How can I do that? Are these items Load itself or some kind of calculation needed?

crimson red

evom mate,

I have to make sure that this issue occurs from either some kind of a hardware fault (corrupt MAP/MAF sensor, etc) or ECU settings.

As my opinion, this is not sensor fault. Looking at the pic, MAP and MAF sensor seems to work well. ( and There is pulsations of the intake air around 3000-5500rpm by FP RED compressor surge...I think. ) So I think this issue could be solved by correct tune.

Could you look at attached pic and answer me? Any different ideas? Thank you.

tephra

iTrader: (6)

You have access to the MAP_Calcs and MAF_Calcs..

Try logging them and posting a picture.

that should narrow down which sensor we need to look at

crimson red

Oh, I see... seems ECU does some sort of interpolation between MAPCalcs and MAFCalcs.　Very smart way, thanks so much tephra and Golden!

From the log and pic attached(with much increased 3xMAP tables as an experiment ), while pulling my car, the ECU was using value of MAPCalcs as Load until 3500 rpm. After that, Load was going back and forth twice between the MAPCalcs and MAFCalcs.

Ummm...probably I have to define both the MAF Scaling and the 3xMAP Scaling properly by myself. But so far, I have no knowledge about how one can create MAF Scaling and MAP Scaling.

I found some articles written about scaling a little. but couldn't find HOW-TO thread about creating both MAF scaling and 3xMAP scaling.

For example, I have a 3" AMS intake whereas the stock intake has 2.5".
Area ratio is calculated as 1.44. Can I get my version of the MAF scaling table by multiplication all values of MAF scaling table?

tephra

iTrader: (6)

actually I think the stock intake is 2.62".

so the AMS 3.0" is 1.31x bigger

and I gave Golden the MAP/MAF Calcs

richardjh

That graph of "deliberately induced load f**kwittery" is a brilliant example of how the internal calcs work.

I saw my own load zigzag on a long straight in 4th (track), and always wanted to get back and finesse it. As soon as I'm up for another track blast, I'll be all over this whole MAP_Calcs/MAF_Calcs thing. Thanks folks - more toys to play with.

Rich

crimson red

I took a mistake. Thank you for correcting the stock diameter! I got correct number.

tephra, can I have three questions to fix the scaling tables issue?

q1. Attached a pic comparrng two scaling curve. Light green indicates the MAF scaling curve for 3" intakes, dark brown for the stock curve. seems correct?

q2. When re-scaling MAF scaling tables, do we need a further fine tune? for example, for the high flow rate area, should multiplier be reduced gradually? ( probably for air flow loss. (my image: f(x)=a*x*x - b*x, x>=0, b is a small constant. ) )

q3. After re-scaling the MAF scaling tables, I may have to mod the 3xMAP tables too. My rough image is that :
1. initially import the stock values.
2. Take logs and referring to MAFCalcs/MAPCalcs (and MAF, Load), correct the tables.
Is this in the right way?

Thanks a lot for your help.

crimson red

Hello,

My head is being overloaded in order to solve this issue, but my trial and error increased your joy, I'm happy Rich.

crimson red

After several times stupid idea like previous post ( sorry tephra ), I started taking logs with x1.31 tables as a baseline.

As my first impression, while Idle x1.31 MAF scaling looks like a little too high, though fuel trims compensate it.

Thank you very much everyone who gave me comments. I'm waiting for becoming able to pull my car again in next Spring.

crimson red

I might have found a way to caluculate and plot actual MAF like data vs MAFVolts curve from a log.

My formula for caluculate MAF(g/s) from an EvoScan log is : ((((( 28.966 / 22.4 ) * ( 0.068046 * MAP ))/( 1 + ( 0.00367 * ( 5 / 9 * ( MAT - 32 ))))) * ( 1998 * 0.001 )) * ( 2byte-RPM / 2 )) / 60

For calculation, I use 28.996 (g / mol) as average molecular weight of dry air.

crimson red

I compared calculated maf and ecu maf by plotting in two cases.

Attached is the pic.

I didn't even think both were so similar like this. Very impressive...

tephra

iTrader: (6)

no offence, but I dont really know what I am looking at here

crimson red

Oh, please never mind, I'm just too much to know.
For example, how does ECU MAF act?, how does actual MAF draw curves?, why is the unit of the MAF scaling tables RAW x?(), ... etc.

In addition to it, as it's snow season in Hokkaido, I cannot pull my car until next Spring(it takes about 5 monthes...). I cannot check MAF_Calcs and MAP_Calcs from a log by pulling my car now, though I know it's the most important thing to solve this issue...Sorry... )

So, I'm doing the other things what I can examine now to solve this issue. Thank you.

crimson red

Think the issue was solved.

# It's so warm that some roads were not covered by snow. lucky day. I could pull my car.

It's my MAF Scaling tables that caused Load be plummeted. Attached is the graph after the tables was finely tuned.

Thanks a lot to everyone who posted comments!

sstevojr

iTrader: (7)

I've been working on something similar:
In EvoScan there is the logging parameter MAFgs; what 'engine sensor/signal' is used to determine these flows rate? I used about 10,000 lines if logs to create partial map of the actual flow. Then tried using known engine principle to create a new type of VE table. My thought is that if we can log the actual MAFgs (which I believe is the value in the 3x'sMAP tables), we can then set the Calibration Fuel Map to 100% and use actual AFR for the tables (open loop). But.....this all depends on the logged MAFgs being correct.
I'm no ROM disassembler, so I can't look into what the ID:238087f8 is in the 90107 (2010 RA). I've tried asking a couple times, but
Does anyone know how the MAFgs logging value is determined?
My experiments hit a road block in that I can't use any actual MAFgs values below 2k rpm and .25psi (low rpm, negative boost) w/o serious drive-ability issues. So I took my normal VE in this region combined it with the table based on logged values. A bit of smoothing, blessing from a Santeria priestess, and voila!


BUT.....this all means nothing of the logged values are.....inaccurate.