Lean-out issue around 3000rpm
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
...
I'm still slightly weary of the MAF scaling since I have never done it before... wish I had something to look at for a reference or close starting point. My X is my only car right now because my Evo 8 is down so I have to learn it and get it right the first time lol
...
I'm still slightly weary of the MAF scaling since I have never done it before... wish I had something to look at for a reference or close starting point. My X is my only car right now because my Evo 8 is down so I have to learn it and get it right the first time lol
...
Newbie
Joined: May 2010
Posts: 5
Likes: 0
From: Sydney
Im using a 3" Synaps intake with a 3a Racing dry pod filter. I dont have a heat shield (yet) but have installed the stock snorkle in its original place.
What i find is that while cruising, every x amount of seconds, the car starts to drift a little lean (say from 14.7 to 15.5) because its getting more air than its measuring, so it then corrects itself by pushing extra fuel through, then it ends up drifting a little rich and so on.
The lean out at 2-3k may be because at that particular partial throttle at low load, the maf volts DO need scaling as it is getting more air and its AFR effects are being amplified (this is beebles insight into the matter)
What i find is that while cruising, every x amount of seconds, the car starts to drift a little lean (say from 14.7 to 15.5) because its getting more air than its measuring, so it then corrects itself by pushing extra fuel through, then it ends up drifting a little rich and so on.
The lean out at 2-3k may be because at that particular partial throttle at low load, the maf volts DO need scaling as it is getting more air and its AFR effects are being amplified (this is beebles insight into the matter)
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
Got some interesting logs this evening on 3dman1's Evo 10. A log of the stumble is shown below. Besides going lean, check out the ign timing being pulled. It really stumbles badly when this happens. What we found is that if it goes lean enough, then it triggers a misfire, and the ECU pulls timing even with no knock. I think the timing pull is taking place as part of the way the ECU is programmed to respond during a P0300 event. We aren't getting a P0300 CEL though, so I'm not sure by any means that this is it. I've got no time to go digging through Evo 10 code that I don't look at much any more, so hopefully tephra or others can find some time to take a look.
So we ended up dropping the open loop load down to 90. That pretty much took care of it. There is definitely something whacked with the way the ECU is calculating IPW in that ~100-120 load range during closed loop. It doesn't look to be MAF scaling because the MAF voltage rises steadily in this RPM range even though AFR is leaning out badly.
EDIT: Guess I'll mention that we changed MIVEC exhaust from -20 deg to -10 deg (-10 is stock) in the load/rpm range where the stumble occurs, and it did not lessen the stumble at all. We also logged MIVEC to make sure it was following the MIVEC maps, and it was.
So we ended up dropping the open loop load down to 90. That pretty much took care of it. There is definitely something whacked with the way the ECU is calculating IPW in that ~100-120 load range during closed loop. It doesn't look to be MAF scaling because the MAF voltage rises steadily in this RPM range even though AFR is leaning out badly.
EDIT: Guess I'll mention that we changed MIVEC exhaust from -20 deg to -10 deg (-10 is stock) in the load/rpm range where the stumble occurs, and it did not lessen the stumble at all. We also logged MIVEC to make sure it was following the MIVEC maps, and it was.
Code:
Time TPS RPM Load MAFV Boost Ign IPW AFR KS STFT 22.444 53 2000 100 2.33 1.9 18 6.33 13.44 0 0 22.544 53 2000 101 2.38 1.94 18 6.38 13.23 0 0 22.643 53 2063 102 2.26 2.04 17 6.53 12.74 0 0 22.758 53 2094 103 2.34 2.24 17 6.65 12.48 0 0 22.859 53 2125 104 2.27 2.48 17 6.72 12.63 0 0 22.959 53 2156 105 2.39 2.57 17 6.84 12.69 0 0 23.059 53 2188 107 2.49 2.72 17 7.01 12.83 0 0 23.159 53 2219 108 2.6 3.06 18 7.14 12.86 0 0 23.272 53 2250 110 2.41 3.11 17 7.28 12.97 0 0 23.367 53 2281 111 2.43 3.4 17 6.72 13.01 0 -0.2 23.467 53 2344 113 2.48 3.54 17 6.3 13.02 0 -0.59 23.567 53 2344 114 2.62 3.73 17 5.86 12.82 0 -0.98 23.667 53 2406 116 2.69 3.98 17 5.39 12.95 0 -1.37 23.767 54 2406 118 2.67 4.22 17 5.39 13.48 0 -1.76 23.876 54 2406 120 2.84 4.61 17 5.72 14.3 0 -1.37 23.976 54 2438 122 2.81 4.65 17 5.78 16.41 0 -0.98 24.076 53 2531 122 2.62 4.8 13 5.86 17.14 0 -0.59 24.176 54 2406 123 2.47 5.04 16 5.97 17.33 0 0 24.276 54 2531 124 2.76 5.09 5 6.03 18.08 0 0.39 24.386 54 2375 109 2.7 4.31 17 5.86 17.42 0 0.78 24.486 50 2563 123 2.89 5.14 3 5.89 17.64 0 1.17 24.589 55 2594 119 2.52 5.62 17 6.46 16.8 0 1.76 24.685 59 2313 124 2.78 5.28 15 6.22 16.93 0 2.15 24.785 58 2688 127 2.95 5.86 1 6.5 16.86 0 2.54 24.900 58 2375 133 2.3 6.49 15 6.44 17.18 0 2.93 25.000 58 2594 131 3.05 5.52 1 5.68 16.61 0 3.52 25.099 57 2719 125 2.77 6.39 14 6.49 16.93 0 3.32 25.199 58 2281 134 1.88 6.06 14 6.23 14.92 0 3.32 25.299 59 2625 130 2.8 6.54 3 6.64 14.01 0 3.91 25.414 50 2781 134 2.69 7.02 17 9.59 17.65 1 0 25.514 47 2594 132 2.83 6.98 8 8.39 15.71 0 0 25.614 46 2688 132 2.69 7.36 17 7.9 13.8 0 -0.39 25.714 49 2688 133 2.76 7.56 17 9.35 12.88 0 0 25.813 49 2750 135 2.95 7.99 17 9.22 13.51 0 0 25.929 49 2781 137 2.95 8.47 17 9.38 14.19 0 0 26.029 48 2906 140 3.04 8.91 17 9.68 13.52 0 0 26.128 47 2875 143 2.96 9.25 17 9.8 13.48 0 0 26.229 47 3000 145 2.93 9.73 19 10.05 13.58 0 0
Last edited by mrfred; Oct 12, 2011 at 11:50 AM.
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
I agree. IPW should be increasing. STFT shows the ECU switching from open to closed loop at 23.367 seconds. That's when fuel starts getting pulled despite increasing MAF voltage and load.
I know this is technically different but the timing being pulled comment sparked something I have noticed on upshifts and such which actually end up in this area we see lean issues also... I've been meaning to get some data to post up and see if others have seen it also. Anyway, I get a loud pop sound from my exhaust randomly. I originally thought it was a rich condition until I caught the ECU throwing negative timing at the car in this range when the condition happens.
I am guessing this is the fireball people seem to love so much but I dont care about and find it annoying as hell. I've seen up to -10 to -12 degrees in my logs when this occurs on the couple times I have caught it.
Anyone else ever experience this? I am thinking the max negative timing map will be the correction but I dont understand why the ECU would do this in the first place on a simple upshift... The map by default I think is -20 degrees across the board so I am thinking replacing it with the low octane timing map is the key.
I am guessing this is the fireball people seem to love so much but I dont care about and find it annoying as hell. I've seen up to -10 to -12 degrees in my logs when this occurs on the couple times I have caught it.
Anyone else ever experience this? I am thinking the max negative timing map will be the correction but I dont understand why the ECU would do this in the first place on a simple upshift... The map by default I think is -20 degrees across the board so I am thinking replacing it with the low octane timing map is the key.
Last edited by itzwolf; Oct 12, 2011 at 08:13 AM.
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
ok, so here is the plot of the lean stumble. Easy to see that IPW starts dropping at the moment the ECU goes into closed loop. The ignition timing pull looks quite dramatic in the plot.
I continue to think the lean issue lies with the IPW calculation. Below is 3dman1's fuel map. Many people don't know that the fuel map is only used during open loop. During closed loop, the ECU automatically uses 14.7:1 as the target AFR no matter what is in the fuel map. So, what's interesting about 3dman1's fuel map is that in the 2000-3500 rpm range, I have to use very rich AFRs to keep it from going lean. There seems to be something funky the calculation in this rpm range. I'm not sure if people have noticed it, but the Evo 8/9 also has this same issue, but not as bad as the Evo 10.
