Notices
ECU Flash

Need help tuning Idle after MAF swap

Thread Tools
 
Search this Thread
 
Old Jun 1, 2011 | 07:23 AM
  #1  
HornstarBU's Avatar
Thread Starter
Evolved Member
iTrader: (7)
 
Joined: Sep 2003
Posts: 1,509
Likes: 1
From: Austin, Tx
Need help tuning Idle after MAF swap

I have a 4g94 engine, but am steadily putting turbo parts from the Evo onto it. It's got the stuff in my sig, minus the injectors, they never got put in yet.

The issue I'm having is that when using an evo 399 MAF, my idle goes to crap after the car warms up, with the A/C off. A/C on, it seems to do okay.

I will attach a log of my drive to work this morning, the notes in the lognotes column somewhat describe what is going on. Basically, the car dips to 10AFR rich when i let the clutch in, so I'm in neutral with the A/C off. It steadily tries to work it's way back to 14.7, but the rpm's dip every so often, and the AFR drops back down. I have done some adjusting to the ISCV, and it is getting closer, but not perfect. Another issue is that the MAF swap threw my fuel trims out of whack.

Basically, I'm not sure what to attack first....should I re-tune my injectors (stock 252cc) to get the trims right? Should I try running forced open loop and try to retun the MAF settings? Should I lower MAF comp in the lowest 2 cells where the car idles? Should I keep increasing ISCV steps with A/C off until it idles at the right AFR? Other thoughts?

I tried to swap the Evo MAF in once before, and this issue is what sort of kept me from staying with it....well, I'm more determined this time so any help is MUCH appreciated.

Log file - http://www.mediafire.com/?mu2pk70w4xkt4k9
94170015 ROM - http://www.mediafire.com/?5npx7h9vdtirxfl
Reply
Old Jun 1, 2011 | 08:28 AM
  #2  
stunt2's Avatar
Evolved Member
iTrader: (8)
 
Joined: Nov 2010
Posts: 753
Likes: 1
From: Georgia
lower MAF scaling values in 19-50 Hz cells
Reply
Old Jun 1, 2011 | 12:58 PM
  #3  
HornstarBU's Avatar
Thread Starter
Evolved Member
iTrader: (7)
 
Joined: Sep 2003
Posts: 1,509
Likes: 1
From: Austin, Tx
Originally Posted by stunt2
lower MAF scaling values in 19-50 Hz cells
I am going to try what you recommend this evening. I spoke to a friend whose on the forums from time to time, "RoadSpike" and he also recommended the same. He said it wouldn't hurt to do it in open loop (forced by setting load threshold to 10 for open loop) just so that it's a more accurate result. So, I will be doing that as well, probably in 2 steps though, to see what exactly happens.

My other question then....is why is MAF tuning required. It seems like it's a sensor, with specific values to be used for reference just like a MAP sensor...although you have to tune VE on SD so that's what RoadSpike compared it to.

If possible, can you or someone tell me what I am looking for as well? Lower those MAF scaling values, or actually, shouldn't I lower Compensation? Regardless, the end result should be closer to 14.7 AFR's and the fuel trims will eventually learn back to +/- 3 like they were once I turn closed loop back on, right?
Reply
Old Jun 1, 2011 | 01:42 PM
  #4  
l2r99gst's Avatar
Evolved Member
iTrader: (2)
 
Joined: Mar 2004
Posts: 3,499
Likes: 4
From: CA
Originally Posted by HornstarBU
My other question then....is why is MAF tuning required. It seems like it's a sensor, with specific values to be used for reference just like a MAP sensor...although you have to tune VE on SD so that's what RoadSpike compared it to.
The maf tables are what calibrates that particular sensor. If you change out the MAF sensor to one with a different size or to a different setup, then of course you need to recalibrate it. Your tables are set for your old sensor/setup. It's all in the details of how the Mitsu Karmen MAF sensor works.

What you need to do is just use your MAF tables to recalibrate the new maf. Either do it in open loop or stay in closed loop and use your fuel trims (short + long) as your guide. Do the adjustments for as many airflow levels as you can. Eventually you should have a new table scaled and calibrated for that new sensor, with the same exact trims and AFR (WOT) as before.

Last edited by l2r99gst; Jun 1, 2011 at 01:45 PM.
Reply
Old Jun 2, 2011 | 12:02 AM
  #5  
stunt2's Avatar
Evolved Member
iTrader: (8)
 
Joined: Nov 2010
Posts: 753
Likes: 1
From: Georgia
you shuold lower maf values itself, not compensation
if it's rich, lower maf values, if lean - raise them. tweak with it untill you set desired afr. don't be afraid, you can't harm anything

lower the values say by 20 percent first and start from there

Last edited by stunt2; Jun 2, 2011 at 12:06 AM.
Reply
Old Jun 2, 2011 | 03:49 PM
  #6  
HornstarBU's Avatar
Thread Starter
Evolved Member
iTrader: (7)
 
Joined: Sep 2003
Posts: 1,509
Likes: 1
From: Austin, Tx
In open loop its 1 Afr pt low pretty much across the board. At idle its 10-11 with the ac off, and 13 .5 or so with it on. Its making a lot of sense now, just have to fix ITT up like you said, with15% reduction or so across and more at the low low spot. Status update tonight.
Reply
Old Jun 2, 2011 | 08:34 PM
  #7  
HornstarBU's Avatar
Thread Starter
Evolved Member
iTrader: (7)
 
Joined: Sep 2003
Posts: 1,509
Likes: 1
From: Austin, Tx
This log, contains my last of about 5 reductions in MAF scaling. What I basically saw was this, throughout all of the changes, the values at the 31hz airflow seemed to get leaner, closer to 14.7. But, the one thing that's really bugging me, no matter what changes, I still saw this 1 thing. I stop, let the clutch in so it's in neutral, and the idle would stay a little too high, slowly thinking about reducing to the "target idle", eventually it would lower slowly, and then the AFR would go richer, the engine would "stumble", AFR drops to 10, and then the engine recovers, back to the "higher but still rich AFR, of like 13 or so by my last MAF increase.

I keep thinking my ISCV values may be wrong. ISCV steps in the base lancer rom are probably defined wrong, since the A/c off value is higher than the A/c on value (but low octane map is always used in base lancer, and MAF comp vs temp are reversed too) or at least named backward....regardless, could that be my issue? It seems fine with the stock MAF, so it's hard to say exactly....really at some what of a loss. Right now I'm set basically backwards of the base lancers values...which actually match up somewhat well with the 05 base lancer values, that do seem to be defined properly and correctly named.

I reduced from 150 units with scaling "uint8 down to 80, or if you use MAFScaling plus MaF adder scaling....then it's 360 down to 270.

Also, the other thing I notice, is that when i flip the A/C on, it still does pretty well, but with the MAF scaling changes so far, the engine starts to idle slightly high, and the AFR is fine, but as it slows lower toward target, the AFR goes fairly lean, like 16, where it used to sit nicely at 14.7 exactly with A/C on at idle, even in open loop only, at like 37hz or 41hz airflow.

So, let me know what you think. I do feel like it's a hell of a lot easier to test changes in open loop, so wish I'd have tried that earlier...still not sure about the idle drop or dip though.
Attached Files

Last edited by HornstarBU; Jun 2, 2011 at 08:37 PM.
Reply
Old Jun 7, 2011 | 08:09 PM
  #8  
HornstarBU's Avatar
Thread Starter
Evolved Member
iTrader: (7)
 
Joined: Sep 2003
Posts: 1,509
Likes: 1
From: Austin, Tx
Thank you guys for all of the advice so far. I did need to tune up the MAF Scaling tables to bring the AFR's more in line with what the Fuel Map had listed. My engine does seem to have a different volume/intake, so it doesn't really fit those same values used in an evo.

So, while the car is running quite well, and in open loop, the AFR's nearly match everywhere....I still have this rich idle with the A/C off. Let me try to explain it a little better, now that I have more exp with it.

When I stop, the AFR is decent, 14.7, and the idle is slightly high, like 1k. When it finally falls down some, to what is typical (desired idle), of 750.....the car goes rich there. When the A/C is on, it goes down to 1k rpm when i let off. A few seconds later the idle drops to 800 which is expected, and the AFR stays the same. In either case, MAF hz is basically the same. If I roll along in 1st, going as slow as possible, MAF hz is 30~40, same as idle, but AFR is fine with A/C on or off.

So, what could be causing the AFR to dip rich but only with the AC off...I also for testing sake set the AFR map much leaner in the lower regions and it idles leaner than 14.7....but STILL dips to 10~11 AFR when the idle comes down below 800 with A/C off.

Also, as a side note...how do you set ISCV values to their proper range. I'm guessing you lower them as low as possible, without the idle dipping too low too fast....Lets get that AFR fixed first though, lol. I really don't see anything that affects AFR with A/C on vs off....and there's no reason I would think that would be necessary either....still, thoughts?
Reply
Old Jun 8, 2011 | 10:57 AM
  #9  
HornstarBU's Avatar
Thread Starter
Evolved Member
iTrader: (7)
 
Joined: Sep 2003
Posts: 1,509
Likes: 1
From: Austin, Tx
I had a bit of an "ah HA!" moment on the way to a site this morning. While driving, here was my thought process....

The engine idles good at a/c on, higher up rpm's, 31hz maf, good afr....so why when the a/c is off does it idle worse, richer....same maf hz..............wait, same maf hz? but the idle is lower, and it's not working as hard, so why doesn't the maf register lower?

That's when I sort of put it all together, in that maybe it really should be reading like 25 or something when i'm idling no A/C. I hadn't thought much about it, but when i initially bought the MAF from an evo guy, he said "mint condition", so I bought it, like 50 bucks or whatever. When I got it, it had tons of "glitter" in it, just all over the inside, had this big circle on it, since he was running it no air filter and the honeycomb was ripped out.

So, long story short, I pm'd him, he said "it works fine without that dude...", so I just shrugged it off. Now, I'm thinking that since that isn't there to straighten the air out, that it's coming across the sensor sideways, making it think that more air is actually coming through, than really is. So, it runs rich.

I ordered a MAP sensor, 4 bar, and a IAT sensor w/ sd harness from spoolin up. I'm going to get that stuff pretty soon here, install it, flash the SD rom, remove the MAF from the picture, and see if the issue disapears. If it does, I bet I can write this whole thing off as a "bad maf".

thoughts? also...If i sell that maf, is that wrong? I would say upfront that the honeycomb is ripped out, but do you guys think it might be broken? It works aight, just not at the lowest airflow.
Reply
Old Jun 8, 2011 | 12:33 PM
  #10  
l2r99gst's Avatar
Evolved Member
iTrader: (2)
 
Joined: Mar 2004
Posts: 3,499
Likes: 4
From: CA
If the honeycomb is out of the MAF, that could very well be most, if not all, of your problem.

Another thought is that your idle with and without AC on are actually at different Hz values, but the logging step of 6.25 (IIRC) isn't enough to show it. For example, if you idle at 28 with AC off and 34 with AC on, then they will both register as 31 Hz. You can try 2 byte logging to be more precise.

But with your MAF and the honeycomb issue, etc, I think your choice of going SD will probably be worthwhile and may very well cure your issues.
Reply
Related Topics
Thread
Thread Starter
Forum
Replies
Last Post
mrfred
ECU Flash
441
Aug 1, 2022 06:05 AM
Dapper Dan
ECU Flash
59
Feb 5, 2021 04:57 PM
simple_lancer
ECU Flash
77
Oct 14, 2020 02:18 AM
Raceghost
ECU Flash
24
Aug 10, 2018 10:39 PM
Kracka
ECU Flash
8
Oct 3, 2009 04:13 PM




All times are GMT -7. The time now is 02:12 PM.