APS 70mm CAI and MAF

ravenhil · 11-18-2007, 12:12 AM

Tuned about 7 cars now, mostly euro models WRX with fairly good results so far for the past 3 months of getting that Tactrix cables. I'm not a professional tuner though I enjoy putting a big wide smile for my friends who do not have the time or patience to use openecu. However I hit a stumbling block recently.

My buddy is driving a MY03 WRX euro domestic model (EDM) and he has very little mods save for an APS cold air intake with a 70mm diameter MAF sensor housing.

Once it is bolted on, the car seems to die on idle and behaved erratically on low rpm. tried a lot of tricks on it and decided that the MAF Sensor Scaling is the solution to the bad idling.

The stock MAF sensor housing diameter is 65mm and thus the extra 5mm diameter roughly equates to 17% more area of a circle. So my guess is, the MAF sensor is a hot wire system, and measures the amount of air going thru the MAF housing by the rate which the hot wire cools down. In other words, it measures the amount of air going thru the MAF sensor housing based on air velocity. This is my assumption.

Thus if the area of the circle is larger, and the rate of which air enters the engine at idle is the same, air flow would be slower. Since the stock ECu do not know that a 70mm MAF housing is used, it assumes that the houing is still 65mm. This screws up the ecu and thus it can't idle well!

So the next thing I did was to increase the g/sec (the second column) in the Maf Sensor Scaling table by 17% all across and I hit the max figure of 300 g/sec at 4.6 volts. There are 4 more cells to go so the figure stayed at 300 g/sec.

While idling has been resolved, I realised that the injector duty cycle maxes out as early as 4600 rpm on 1 bar boost (stock turbo TD04)! My buddy did mentioned that his fuel economy has gone out of the window and it doesn't take a genuis to realise that the new MAF g/sec values has made the AFR way too rich.

i tried a compromised set of MAF Sensor Scaling values that allowed his car to idle by increasing the g/sec value by 17% at idling range (observed no more than 1.8 volts using Enginuity datalogger) and smoothing down the value so that it hits 300 g/sec at the last cell.

Seems that I hit jackpot for a while, that is until the P0171 CEL comes out. Too lean. While it is easier to swap the 70 mm MAF housing to a 65mm one, it seems somewhat foolish to give up at this point.

Any ideas gentlemen?

* Registered users of the site do not see these ads.

HndaTch627 · 11-18-2007, 12:20 AM

on a stock TD-04 that APS CAI is more work then it's worth. the K&N Typhoon and SPT CAI's seem to be the best overall units.

Personally i hate Big MAF housings as they are damn near impossible to tune just right.

ravenhil · 11-18-2007, 12:36 AM

Apparently his car was running fine on Ecutek with the aforementioned set up. i refuse to believe that Ecutek can do what openecu can't. And yes, big maf is tough. my head hurt a lot that day.

WeldingHank · 11-18-2007, 01:00 AM

it seems you multiplied his MAF scalar a bit too much.its only about a 5% change. also, the stock MAF diam. is 66.5mm.

66.5MM*1.052=69.958mm

try that out and log his long term trims to fine tune. the top end may need a wideband.

I got this value from one of cobb's spread sheets.

Broxma · 11-18-2007, 01:37 AM

The resolution doesn't stay consistent across the entire MAF Scaling range, but like Hank said that figure is good place to start but it will only work for a very limited range in the MAF Sensor Scaling table. You could probably do that increase across the board and it would at least idle and if anything run slightly lean at WOT.

I know the Injen CAI starts off around 15% increase at .94 but by the time you get up towards 3 volts it's down to 3% or so. I have an Injen CAI and my AF Learning is around 5% at idle so I know it's not correct, I just haven't done anything about it. I am slightly lean at idle but not by much, maybe half a point or so.

You can log your AF learning, flow and voltage numbers with a Wideband and get a good idea on how much you need to adjust at specific MAF Voltages, but you need alot of data points at very consistent throttle positions.

/Brox

ravenhil · 11-18-2007, 09:08 AM

Thanks, now I kno clearly where to go. I suspected that a blanket multuplier wouldn't work. Already purchased a PLX WBO2 sensor, now waiting for that exhuast clamp to set things right.

And another thing, know that low detonation fuel map? It seems that the actual AFR (as captured in datalog) is anything but what you set on that table. And when it goes into open loop, the AFR is alwaya 11.25 regardless of make and model, from MY02 "bugeye" to MY07 "hawkeye" moels. What gives?

clemsonteg · 11-18-2007, 10:31 AM

Quote:

Originally Posted by ravenhil

And another thing, know that low detonation fuel map? It seems that the actual AFR (as captured in datalog) is anything but what you set on that table. And when it goes into open loop, the AFR is alwaya 11.25 regardless of make and model, from MY02 "bugeye" to MY07 "hawkeye" moels. What gives?

Thats because thats the limit on how low the stock O2 sensor can read

WeldingHank · 11-18-2007, 10:33 AM

the stock a/f sensor doesn't have enough resolution to be accurate in those a/f ranges. you have the injectors scaled properly and the MAF as well, you can make your low-det fuel maps match what comes out your tailpipe. i know mine is only about .1-.2 A/F difference right now on a stock intake of course.

dynamix · 11-20-2007, 03:43 PM

i run with an APS 70mm CAI and whilst it has taken a while for the MAF curve to be right it does get there in the end. It is substantialy different to the stock curve though.

This is for a 2.5 STI UK 32 bit ecu.

Not sure if the maf table is the same size.

ravenhil · 11-21-2007, 05:42 AM

I heard that the STI allows a higher ceiling for MAF g/sec values. While the WRX max out at 300 g/sec (in EcuFlash, it was 299.98), the STI can supposedly go up to 350 g/sec.

And when you said "taken a while for the MAF curve to be right it does get there in the end", do you mean that the ECU simply learns it by itself?

dynamix · 11-21-2007, 06:15 AM

Not by itself no, it is a very gradual process of monitoring afr's, correction, ltft and then applying that to the maf curve to get it right. Mine has run to 336 grams/sec last week.

Wrinkleboi · 11-21-2007, 12:51 PM

just wanted to add that 66.5mm to 70mm is more than 5% because flow doesnt concern diameter, it concerns area. the difference in area of the two sizes is 10.8% (pi * r^2)... with that said, i dont think that the maf scaling is quite as simple as that.

Freon · 11-21-2007, 01:12 PM

I've found MAF scaling doesn't scale with cross section area. That does make sense, and it was the first assumption I made as well. It just doesn't seem to hold. It's closer to being related to diameter even if that doesn't make sense.

I haven't dealth with the APS 70mm yet, but I'd try just multiplying all the g/s values in the MAF table by about 6-7%. That will get you started.

You really need to log your fuel trims (AFCorrection#1, AFLearning#1), drive around, and then when you stop you can also check your stored AF learning A/B/C/D. Keep in mind sometimes D won't move a lot if your open/closed loop delay is set really short. The upper part of MAF scaling really has to be tuned with a wideband.

dynamix · 11-21-2007, 06:08 PM

from experience, i dont think the way it flows at one point compared to std is the same all the way up the range.

bboy · 11-21-2007, 06:56 PM

Flow thru a tube scales with the quadratic of the radius. r^4 Poiselle's Law.

I've tuned exactly one APS CAI. It took a little time, but all intakes do.

The stock injectors are too small from the factory here in the US. They have basically nothing left in them for more power. So, cut the ends off or get bigger ones.

ravenhil · 11-24-2007, 10:38 AM

Now I know why AFR always show a static figure even though the values in the open loop look up table shows otherwise.

It is because of this!

Front Oxygen Sensor Rich Limit

ravenhil · 11-24-2007, 10:53 AM

Okay, with the advice from everyone, this is what I have done.

This is the stock MAF sensor scaling

and this is the modified maf scaling

Still get P0171 once in a while,

jays05 · 11-24-2007, 11:09 AM

Try this

http://www.iwsti.com/forums/cobb-str...libration.html

dynamix · 11-24-2007, 12:44 PM

Code:

[Table2D]
0.94	0.98	1.02	1.05	1.09	1.13	1.17	1.21	1.25	1.29	1.33	1.37	1.41	1.48	1.56	1.64	1.72	1.80	1.88	1.95	2.03	2.11	2.19	2.27	2.34	2.42	2.54	2.66	2.77	2.89	3.01	3.12	3.24	3.36	3.48	3.59	3.71	3.83	3.95	4.06	4.18	4.30	4.41	4.49	4.57	4.61	4.65	4.69	4.73	4.77	4.80	4.84	4.88	4.92	4.96
2.35	2.55	2.80	3.15	3.60	4.05	4.55	5.10	5.50	6.10	6.95	7.85	8.80	10.10	11.60	13.62	15.05	17.15	19.25	21.44	24.02	27.05	30.11	33.20	36.30	39.85	44.88	50.16	56.30	62.95	71.50	80.30	91.50	103.02	116.10	129.78	144.91	160.47	176.65	195.15	216.80	239.82	263.34	285.55	310.13	323.70	339.60	354.55	372.00	389.80	407.00	425.30	444.00	465.00	480.00

These is my maf scale

Gives me 0.8 /-1.6% learning.

Ignore the really top figures as I have never been above 338 gram/sec with it

ravenhil · 11-25-2007, 09:04 AM

I am touched. Thanx guys

dynamix · 11-28-2007, 03:40 AM

How has it gone with it?

ravenhil · 11-29-2007, 06:55 AM

tried to bring down the maf g/sec values with reference to a datalog that shows Air/Fuel Correction #1 (%) 25% across the 2.28 to 2.52 V range. My guess that the ecu is trying to add as much fuel as possible. Didn't help. Deceased most of the g/sec values by 20% (avg) but idle hunt was bad.

The MAF is tough. I am not sure what's the next step to go. Increasing the MAF too much makes the IDC max out too early, going too low makes idling terrible. Just can't sem to find the perfect middle ground.

Will consider tweaking fuel latency next!

dynamix · 12-01-2007, 02:18 PM

If it is adding fuel you should RAISE the grams/sec at a voltage.

DJIMPREZA · 12-01-2007, 09:28 PM

How you guys edit the maf table using the enginuity software? I am getting this estrange bug that I can`t type in decimal points and the up and down arrows are not making any change to selected cell..

WTF???

Thanks

ravenhil · 12-03-2007, 12:14 AM

Erm, have you tried keying in the values you want using the '=' key? You just have to click on the cell you want to edit, type in the value you want, then press 'enter'. It should enter that value and go to the next cell.

I prefer to use EcuFlash myself for the cool interpolating and mutiply function.

11-18-2007, 01:00 AM	#4
WeldingHank Scooby Specialist Member#: 98645 Join Date: Oct 2005 Chapter/Region: NESIC Location: Lawrence,ma Vehicle: 2002 WRX Wagon PSM	it seems you multiplied his MAF scalar a bit too much.its only about a 5% change. also, the stock MAF diam. is 66.5mm. 66.5MM1.052=69.958mm try that out and log his long term trims to fine tune. the top end may need a wideband. I got this value from one of cobb's spread sheets. Last edited by WeldingHank; 11-18-2007 at 01:11 AM.*

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
FS (MD): GT52, Perrin inlet, APS 70mm CAI, Turbosmart Dual Port BPV, Samco MAF hose	boost junkie	MAIC Private Classifieds	29	02-21-2010 03:12 PM
FS: PA - APS 70mm CAI and an Amsoil oilless air filter	Mycues1982	Engine/Power/Exhaust	7	01-05-2009 01:29 PM
WTB/WTT K&N CAI FOR APS 70mm/Perrin Big MAF	GRUMPY514	Private 'Wanted' Classifieds	5	06-02-2008 09:10 PM
SoCal: UR EWG Uppipe, APS 70mm CAI and 3" perrin turbo hose	BlueFish	Engine/Power/Exhaust	10	03-12-2008 11:49 PM
WTT: APS 65mm CAI for APS 70mm CAI... gotta be crinkle red and great condition	Wrinkleboi	Engine/Power/Exhaust	4	11-30-2007 01:58 AM

11-18-2007, 12:12 AM	#1
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	APS 70mm CAI and MAF Tuned about 7 cars now, mostly euro models WRX with fairly good results so far for the past 3 months of getting that Tactrix cables. I'm not a professional tuner though I enjoy putting a big wide smile for my friends who do not have the time or patience to use openecu. However I hit a stumbling block recently. My buddy is driving a MY03 WRX euro domestic model (EDM) and he has very little mods save for an APS cold air intake with a 70mm diameter MAF sensor housing. Once it is bolted on, the car seems to die on idle and behaved erratically on low rpm. tried a lot of tricks on it and decided that the MAF Sensor Scaling is the solution to the bad idling. The stock MAF sensor housing diameter is 65mm and thus the extra 5mm diameter roughly equates to 17% more area of a circle. So my guess is, the MAF sensor is a hot wire system, and measures the amount of air going thru the MAF housing by the rate which the hot wire cools down. In other words, it measures the amount of air going thru the MAF sensor housing based on air velocity. This is my assumption. Thus if the area of the circle is larger, and the rate of which air enters the engine at idle is the same, air flow would be slower. Since the stock ECu do not know that a 70mm MAF housing is used, it assumes that the houing is still 65mm. This screws up the ecu and thus it can't idle well! So the next thing I did was to increase the g/sec (the second column) in the Maf Sensor Scaling table by 17% all across and I hit the max figure of 300 g/sec at 4.6 volts. There are 4 more cells to go so the figure stayed at 300 g/sec. While idling has been resolved, I realised that the injector duty cycle maxes out as early as 4600 rpm on 1 bar boost (stock turbo TD04)! My buddy did mentioned that his fuel economy has gone out of the window and it doesn't take a genuis to realise that the new MAF g/sec values has made the AFR way too rich. i tried a compromised set of MAF Sensor Scaling values that allowed his car to idle by increasing the g/sec value by 17% at idling range (observed no more than 1.8 volts using Enginuity datalogger) and smoothing down the value so that it hits 300 g/sec at the last cell. Seems that I hit jackpot for a while, that is until the P0171 CEL comes out. Too lean. While it is easier to swap the 70 mm MAF housing to a 65mm one, it seems somewhat foolish to give up at this point. Any ideas gentlemen? * Registered users of the site do not see these ads.

11-18-2007, 12:20 AM	#2
HndaTch627 Scooby Guru Member#: 6551 Join Date: May 2001 Chapter/Region: MWSOC Location: Carol Stream, IL Vehicle: '01 GC8 Dinged STM '09 Concours 14 ABS Black	on a stock TD-04 that APS CAI is more work then it's worth. the K&N Typhoon and SPT CAI's seem to be the best overall units. Personally i hate Big MAF housings as they are damn near impossible to tune just right.

11-18-2007, 12:36 AM	#3
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	Apparently his car was running fine on Ecutek with the aforementioned set up. i refuse to believe that Ecutek can do what openecu can't. And yes, big maf is tough. my head hurt a lot that day.

11-18-2007, 01:37 AM	#5
Broxma Scooby Specialist Member#: 158006 Join Date: Aug 2007 Chapter/Region: TXIC Location: San Antonio Vehicle: 2003 WRX IdleWagon 1992 GVR4 379/1000	The resolution doesn't stay consistent across the entire MAF Scaling range, but like Hank said that figure is good place to start but it will only work for a very limited range in the MAF Sensor Scaling table. You could probably do that increase across the board and it would at least idle and if anything run slightly lean at WOT. I know the Injen CAI starts off around 15% increase at .94 but by the time you get up towards 3 volts it's down to 3% or so. I have an Injen CAI and my AF Learning is around 5% at idle so I know it's not correct, I just haven't done anything about it. I am slightly lean at idle but not by much, maybe half a point or so. You can log your AF learning, flow and voltage numbers with a Wideband and get a good idea on how much you need to adjust at specific MAF Voltages, but you need alot of data points at very consistent throttle positions. /Brox

11-18-2007, 09:08 AM	#6
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	Thanks, now I kno clearly where to go. I suspected that a blanket multuplier wouldn't work. Already purchased a PLX WBO2 sensor, now waiting for that exhuast clamp to set things right. And another thing, know that low detonation fuel map? It seems that the actual AFR (as captured in datalog) is anything but what you set on that table. And when it goes into open loop, the AFR is alwaya 11.25 regardless of make and model, from MY02 "bugeye" to MY07 "hawkeye" moels. What gives?

11-18-2007, 10:33 AM	#8
WeldingHank Scooby Specialist Member#: 98645 Join Date: Oct 2005 Chapter/Region: NESIC Location: Lawrence,ma Vehicle: 2002 WRX Wagon PSM	the stock a/f sensor doesn't have enough resolution to be accurate in those a/f ranges. you have the injectors scaled properly and the MAF as well, you can make your low-det fuel maps match what comes out your tailpipe. i know mine is only about .1-.2 A/F difference right now on a stock intake of course.

11-20-2007, 03:43 PM	#9
dynamix Scooby Newbie Member#: 122603 Join Date: Aug 2006	i run with an APS 70mm CAI and whilst it has taken a while for the MAF curve to be right it does get there in the end. It is substantialy different to the stock curve though. This is for a 2.5 STI UK 32 bit ecu. Not sure if the maf table is the same size.

11-21-2007, 05:42 AM	#10
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	I heard that the STI allows a higher ceiling for MAF g/sec values. While the WRX max out at 300 g/sec (in EcuFlash, it was 299.98), the STI can supposedly go up to 350 g/sec. And when you said "taken a while for the MAF curve to be right it does get there in the end", do you mean that the ECU simply learns it by itself?

11-21-2007, 06:15 AM	#11
dynamix Scooby Newbie Member#: 122603 Join Date: Aug 2006	Not by itself no, it is a very gradual process of monitoring afr's, correction, ltft and then applying that to the maf curve to get it right. Mine has run to 336 grams/sec last week.

11-21-2007, 12:51 PM	#12
Wrinkleboi Top Scoob 006 Member#: 14424 Join Date: Jan 2002 Chapter/Region: South East Location: Stuart, FL Vehicle: BugI w/ 07 STi Swap www.TriStateTuners.com	just wanted to add that 66.5mm to 70mm is more than 5% because flow doesnt concern diameter, it concerns area. the difference in area of the two sizes is 10.8% (pi * r^2)... with that said, i dont think that the maf scaling is quite as simple as that.

11-21-2007, 01:12 PM	#13
Freon Scooby Specialist Member#: 88322 Join Date: Jun 2005 Chapter/Region: MWSOC Location: Indianapolis, IN Vehicle: 2009 BMW 135i	I've found MAF scaling doesn't scale with cross section area. That does make sense, and it was the first assumption I made as well. It just doesn't seem to hold. It's closer to being related to diameter even if that doesn't make sense. I haven't dealth with the APS 70mm yet, but I'd try just multiplying all the g/s values in the MAF table by about 6-7%. That will get you started. You really need to log your fuel trims (AFCorrection#1, AFLearning#1), drive around, and then when you stop you can also check your stored AF learning A/B/C/D. Keep in mind sometimes D won't move a lot if your open/closed loop delay is set really short. The upper part of MAF scaling really has to be tuned with a wideband.

11-21-2007, 06:08 PM	#14
dynamix Scooby Newbie Member#: 122603 Join Date: Aug 2006	from experience, i dont think the way it flows at one point compared to std is the same all the way up the range.

11-21-2007, 06:56 PM	#15
bboy Scooby Specialist Member#: 56468 Join Date: Mar 2004 Chapter/Region: NWIC Location: Seattle, WA Vehicle: 04 Improved STI Dirty White	Flow thru a tube scales with the quadratic of the radius. r^4 Poiselle's Law. I've tuned exactly one APS CAI. It took a little time, but all intakes do. The stock injectors are too small from the factory here in the US. They have basically nothing left in them for more power. So, cut the ends off or get bigger ones.

11-24-2007, 10:38 AM	#16
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	Now I know why AFR always show a static figure even though the values in the open loop look up table shows otherwise. It is because of this! Front Oxygen Sensor Rich Limit

11-24-2007, 10:53 AM	#17
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	Okay, with the advice from everyone, this is what I have done. This is the stock MAF sensor scaling and this is the modified maf scaling Still get P0171 once in a while,

11-24-2007, 11:09 AM	#18
jays05 Scooby Specialist Member#: 78165 Join Date: Dec 2004 Chapter/Region: South East Location: SC Vehicle: 05 STI 35r + e85	Try this http://www.iwsti.com/forums/cobb-str...libration.html

11-25-2007, 09:04 AM	#20
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	I am touched. Thanx guys

11-28-2007, 03:40 AM	#21
dynamix Scooby Newbie Member#: 122603 Join Date: Aug 2006	How has it gone with it?

11-29-2007, 06:55 AM	#22
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	tried to bring down the maf g/sec values with reference to a datalog that shows Air/Fuel Correction #1 (%) 25% across the 2.28 to 2.52 V range. My guess that the ecu is trying to add as much fuel as possible. Didn't help. Deceased most of the g/sec values by 20% (avg) but idle hunt was bad. The MAF is tough. I am not sure what's the next step to go. Increasing the MAF too much makes the IDC max out too early, going too low makes idling terrible. Just can't sem to find the perfect middle ground. Will consider tweaking fuel latency next!

12-01-2007, 02:18 PM	#23
dynamix Scooby Newbie Member#: 122603 Join Date: Aug 2006	If it is adding fuel you should RAISE the grams/sec at a voltage.

12-01-2007, 09:28 PM	#24
DJIMPREZA Scooby Specialist Member#: 1199 Join Date: Apr 2000 Location: Caracas, Venezuela Vehicle: 1994 Supra+1K, 22b alike, 07STI	How you guys edit the maf table using the enginuity software? I am getting this estrange bug that I can`t type in decimal points and the up and down arrows are not making any change to selected cell.. WTF??? Thanks

12-03-2007, 12:14 AM	#25
ravenhil Scooby Newbie Member#: 113774 Join Date: Apr 2006 Chapter/Region: SCIC Location: Long Beach, CA Vehicle: 2002 Impreza WRX White	Erm, have you tried keying in the values you want using the '=' key? You just have to click on the cell you want to edit, type in the value you want, then press 'enter'. It should enter that value and go to the next cell. I prefer to use EcuFlash myself for the cool interpolating and mutiply function.