WREXY

As you know from my previous posts, I seem to have lean 02 sensor readings on my car with the Delta Dash. I'm trying to work out what the cause is and I'm thinking it could be the APS induction kit, but I need your thoughts as well. I have done logs on my car and two other ones at stock boost levels, for comparison and I'm posting the, inter related log bits, (not the actual log pics), that have to do with fuelling only. They are the boost, airflow sensor readings, injection pulse widths and 02 sensor readings. The logs I have quoted from the three cars have been taken from the same gear, at the same speed with similar boost.

My car has the AE 802 ECU, APS cold air induction kit, midsection and backbox 5Zigen exhaust, Blitz boost controller, FCD and Blitz BOV. I lowered my boost to standard to compare to these other two cars.

The other cars were a MY00 with only PPP, and similar boost to mine. The second car a MY99 has an AE801 ECU, a downpipe and magnex exhaust, (I think), with similar boost to mine. In other words I lowered my boost to have similar boost to these cars, which incidently are Jim Theo's MY00 PPP, and Dmell's MY99, so that we can see if I am really running lean at similar boost to these cars.

Mine had 13.1 psi boost at 116 kph, airflow sensor voltage was 4.1volts, and fuel injection pulse was 13.57. The 02 sensor was 0.865v. This was in 30degree celsius day.

In the same gear at the same speed the PPP car had 12.5psi boost, airflow sensor was 4.24 volts and fuel injection pulse was 16.1. The 02 sensor was 0.94volts.This log was at night around 15 degrees, celsius

A second car with an AE801 ECU but with twin downpipe and magnex exhaust at the same speed had 13.5 psi boost, airflow sensor voltage was 4.2 volts and fuel injection pulse was 15.1. The 02 sensor was 0.89volts. This was in 26 degree celsius day

I couldn't get the boost exactly like theirs, but it is very close. Too hard to program exactly on the boost controller.

I feel I am running lean from these results.

What do you think? If I am running lean, do you have any ideas why this is so?

Could it be my APS induction kit causing this?

Any input would be gladly appreciated.

Cheers,

Wrexy.

Beastie

iTrader: (2)

Wrexy
I can bring out my AE800 Black label for you to try and log.
If that is helpful, as then you can compare results and eliminate items on your car depending on log results - if you follow me.
Beastie

WREXY

Beastie,

Excellent! top Idea!

I'll be up for that thanks m8.

Cheers,

Wrexy.

WREXY

Beastie,

I just remembered. Your AE800 ECU will not bolt into my floor cos you have the RHD ECU and the mounting bracket is the opposite shape to my LHD.

Does the ECU have to bolted to the floor to work?

Anyone know?

If it doesn't need bolting to the floor we can go ahead with our trials.

Cheers,

Wrexy.

[Edited by WREXY - 6/6/2002 11:12:00 AM]

mrkimpreza

Did you already try another (new) MAF sensor ?

Under full load my MAF sensor gave 4,1V as well and low O2 readings (0,83). Fixing the MAF sensor helped (4,4 / 0,9+)

I probably blown my previous shortblock due to the old maf sensor. So be carefull.


Mark.

Stelios

Dont forget that Jim Theo's car has a Jun (like) manifold as well

WREXY

Mark,

I'll be careful m8. I'm driving the car around like a wuss and on low boost too. Especially from Monday, as the temperatures in Greece are going to be in the 40s for the whole week. F****n h**

My stock MAF and the second MAF which is one month old gave the same results. The second MAF gave the same results from brand new. Theo (Evil Bevel) has the same problem, with a brand new MAF. I have ordered another MAF which should be coming next week cos Subaru ran out of them and they are waiting to unpack the container. Maybe the second MAF was from a faulty batch. I'm thinking this because Theo's brand new one is doing the same thing and his readings are identical to mine. If I get the same readings, I'll be removing the APS induction kit to see if it is the problem, but first when Beastie arrives with the AE 800, I'll try his ECU, just to see the difference, power wise and mixture wise. Although they run a little leaner than the AE802/782 I'll still be interested to check it out.

Stelios,

The logs I posted from Jim Theo's car are before the PE manifold was fitted to his car.

Cheers,

Wrexy.

PS I still need to know if the ECU needs to be bolted to the floor to work. I'm thinking is there some type of ground issue for the ECU to work. Would anyone answer please?

EvilBevel

To back up Wrexy here... you may have read my previous MAF blowing - I'm now 100% certain it has to do with shock loading BTW rather than with the filter itself.

Edit: by blowing I mean: stopping to work altogether, i.e. no signal at all.

Before the MAF died, the readings on both DeltaDash and the Lambdalink were exactly where they should have been... around 7-8 % CO on full load (sometimes 9), MAF sensor giving 4.3 volts max. (this is with Unichip setup, remember)

OK, MAF dies, fit in brand now one from the box. Strange, readings seem to be lower ??? After a few days, this was confirmed, the max voltage I could see now was 4.2 volts, and the car really ran lean in several spots (5 CO where I was expecting 7-8 ). I could feel the ECU retarding.
This really really puzzled me.

So I got fed up, and just finished refitting my standard airfilter / resonator (only took me 4 hours and 2 wounded fingers LOL). I will now go for a ride and see if this changes anything to the readings.

If that fixes it then... am *REALLY* shafted, as I ordered the Pace FMIC. That puppy ... erm... doesn't fit with the standard filter housing

If it doesn't fix it, it is even more worrying. Why would a new MAF give a lower reading than another MAF (and yes, it was fitted correctly etc...).
I feel that in the end this is gonna get expensive, and I might give BRD a call. I have lost all faith in MAF sensors to be honest.

Will report back after the drive.

[Edited by EvilBevel - 6/6/2002 1:54:36 PM]

WREXY

I'll be waiting for the report Theo.

Wrexy.

Beastie

iTrader: (2)

Wrexy
There is no sign of it needing to be bolted to the floor.It has one U shaped cut out for securing only plus the plug connector. However as long as we do we should have no problem. Are you sure the ECU has a different casing from LHD to RHD?
Beastie

WREXY

The casing is the same. It's the mounting bracket on the bottom of the ECU that's opposite, so it won't bolt to my floor. The 3 sockets on the ECU where the wiring plugs into is the same in LHD and RHD so we should be right.

Cheers,

Wrexy.

dowser

The boards are easily interchangeable....4 screws from memory

Richard

WREXY

Great point Richard. I never thought of that option.

Cheers,

Wrexy.

EvilBevel

Right... interesting.

Did a drive/log with standard jobbie:

Compared to a run 2 weeks ago:

Highest Injector DC : 17,41 ms now, was 16,38
Highest Airflow : 4.31v, was 4.18v

This is both from 30 minute runs, and car has been through all gears/speeds etc

Overall observations:

Driving off boost feels a lot better.
Throttle response is MUCH improved.
LambdaLink is a lot less "nerveous"
Full throttle now keeps 7%CO light rock steady (even when fully heated up), whereas it dropped to 5%CO previously.
Annoying noises have all gone
No real difference noticable in the higher rev range (VF23, currently 1.25 bar max).
Happier driver overall

BUT:

The values are still slightly lower than on a run/log I did in April, with the previous MAF (RIP). There I saw values up to 4.4 volts. This may however been due to cooler weather ? (Today was about 20°C)

Will report again when I received a MAF shipment from Greece ( ), to see if there are any differences still.

My overall advice would be to not bother with cone filters, as long as you have a MAF to protect. I personally am now back to the standard filter housing/resonator/Subaru paper jobbie.

What I'll do when the Pace arrives is yet to be decided ...

HTH,

Theo

WREXY

Thanks a lot Theo!

Ah yes, looks like we have found the problem. Looks like it's these induction kits, which I kinda hoped it wasn't.

Looks like I'll have to remove mine as well. I'll be using the Delta Dash logs to see if I will lose performance by looking at the in gear times.

On a dyno here, there was an 11hp gain with the APS induction kit on a friends car, Stelios's MY00. I'd be sorry to have to remove it. I'll have to remove it and then the induction kit will only go back on after a remap. I'll leave it on till I try the mew MAF and Beasties ECU, then I'll remove it with Beasties help.

Theo your injection pulse duty is quite rich compared to mine, so your MAF could be faulty giving low 02 readings, but your fuelling may be OK. Even with your previous log from 2 weeks ago that you mention, your injector duty cycle is still rich at 16.38 compared to mine and still more that the other two cars I tested.

I noticed on the EcuTek site, that with the Delta ECU remapping software there is an option in the mapping sofware called, Air Flow Sensor Scaling. I wonder what it does and would it help my case.

Cheers,

Wrexy.



[Edited by WREXY - 6/6/2002 4:41:50 PM]

EvilBevel

>>Ah yes, looks like we have found the problem. Looks like it's these induction kits, which I kinda hoped it wasn't.

Well, the last thing I want to do is provide false information, so I'm still careful. What I can say 100% sure is that:

. off boost driving is a lot better with OEM setup
. same for throttle response
. MAF voltages overall seem to be a bit higher
. MAF voltages seem more "consistent"

I have looked inside the PiperCross, and from what my amateur eyes can tell, the air seems to mostly come from the "front" (of the car).

What I also noticed: when booting it from say 3000 RPM, with cone filter I would see the CO% rise (as expected), then ... fall back all the way (<1 CO%), then rise again. Car would feel "sluggish" at that point. My amateur gut feeling tells me this may have to do something with the "shape" of the airflow, "bypassing" the MAF as it were.

>>Looks like I'll have to remove mine as well.

It's a good idea anyway: when having a problem, go back to last known situation. Works in IT, works on cars I think. I'm not saying that is your problem, I'm saying it may be easier to find the problem by elimination.

>>I'll be using the Delta Dash logs to see if I will lose performance by looking at the in gear times.

I'll bet you a free MAF sensor they won't change You have to account for the weather though, especially in Greece now !

>>On a dyno here, there was an 11hp gain with the APS induction kit on a friends car, Stelios's MY00.

Those 11 BHP would be gained in the last 1000 RPM right (6000 to 7000) ? It's there where the standard setup could become "restrictive", and it's there where you typically see the max BHP for a Scoob. But it might well be that you *lose* horsepower in lower revs. In fact, I'm rather sure of that now.

>>Theo your injection pulse duty is quite rich compared to mine, so you should be right.

Don't forget that a VF23 sucks a lot more air mass than a TD04L at the same boost. And also don't forget that the MAF voltages I report are "scewed" by the Unichip (you richen/lean the mixture by altering the reported MAF voltage to the JECS)

>>I noticed on the EcuTek site, that with the Delta ECU remapping software there is an option in the mapping sofware called, Air Flow Sensor Scaling. I wonder what it does and would it help my case.

I guess you could do a general "correction %" on it. But I'm not sure if that would solve the "lean" spots ? You may end up with places where the car is "OK", and places where the car is way to rich.

John, could you enlighten us on this one ?

john banks

I suggested WREXY post for other opinions as I suggested it was the MAF sensor misreading from the looks of the logs, so I am glad you have reached the same conclusion independently.

I think everyone Ecutek mapping has only adjusted the main fuel map rather than the MAF scaling, but it can be done. There are rather impressive lean spots in the standard maps anyway during spool up, and adding 6 degrees of advance and a shed load of fuel literally makes it spool up wonderfully well and it feels much better off boost. There is a MAF sensor scaling map:



...you don't need to be told they are 16bit hex values The map input is the sensor voltage in 10 mV steps. 0 to 512 in the left column of the map corresponding to 0 to 5.12 volts.

Apologies for the lurid green from converting to .gif and the large fonts not displaying correctly on this screen dump. Note the saturation point of the MAF sensor also. I have seen up to 4.46V no higher and there is a difference in hot conditions. Also the lambda sensor apparently can underread in hot weather (I would have thought the EGT was far more significant!), so this in combination with an underreading MAF sensor leads to the impression that WREXY's car is very lean - the injector pulse widths being so low back this up for the boost he runs, so it seems his car is only slightly lean on full load, if at all, it might be that you would tune it to be there, but it is perhaps a bit odd to be there by accident. Certainly with it on you are likely to be losing a lot of low down/lower midrange torque. Thankfully your car is not running a large amount of knock correction. There is a sweet spot for fuelling and timing at a given load, and you are unlikely to be sitting on it low down even if you are nearer to it higher up the range.


[Edited by john banks - 6/6/2002 5:37:16 PM]

WREXY

Cheers Theo.

I doubt very much that you have provided false info. I can relate to your findings. I have a feeling the new MAFs we have ordered are going to be used as spares for the future.

OK, I understand about the VF 23 and your airflow.

LOL at this, "I'll bet you a free MAF sensor they won't change!"

You are probably right, cos for the 11hp gain at high revs with the APS, I spoke to Stelios and he said that it was 12hp, but it was over 5500rpm, and he had less power under 5500.

OK, APS must come off till I do something about a map.



John,

Thanks for your help also m8. For looking at my e mails and for suggesting I put up a post here. It looks like we have found the problem.

Also thanks for putting up the MAF sensor scalling map, although I must say I have no idea what it all means.


Beastie,

I still want ot try that ECU m8! Even without the APS.

Cheers,

Wrexy.

WREXY

John,

Just read the rest of your post. And I thank you very much for your analysis. Makes sense to me.

Cheers,

Wrexy.

JIM THEO

Wrexy
I'll send you my new log files after the PE exhaust manifold in order to compare with yours (I like Stelio's "JUN like manifold" approach!).
The ECU needs at least one bolt for grounding, just for safety.
I think the problem is your APS kit and not the MAF, but as I told you before I know a guy who changed six MAFs (the 5 of them brand new)!!! Except the dust, the oil and the abnormal flow of air with an aftermarket cone filter, vibration and wrong installation procedure is the main reasons for a blown MAF.
I could explan you how to convert hexadecimal numbers to decimals just to understand John's posts but isn't so easy.
By the way I have also leaner mixture than before the manifold from 1200 to 3000 rpm, and have to fit the downpipe, so, would see what will change after all. Is what I bought a Walbro GSS341 fuel pump from MarkA (very helpful), just for help thinks go better (Ecutek 3 forthcoming!)
Cheers
JIM

PS : Forgive my poor English.
PS1: Beastie when are you going to come here, it would be nice to meet you!

WREXY

Hi Jim,

I think I may have to order one of those fuel pumps too. It may richen the mix a bit and I may be able to get away with the induction kit then.

When we meet up again you can explain the hexadecimal numbers to me.



John,

just a thought, could the cold air induction kit have helped in keeping my intake temps lower, resulting in the little knock correction that I have been seeing in these higher temperatures and leaner mix?

Cheers,

Wrexy.



[Edited by WREXY - 6/6/2002 10:29:38 PM]

john banks

Yes. There are so many factors at play here:

Intake temp
Lambda sensor differences (they are inaccurate)
Environmental temperatures
Manifold
Downpipe, rest of exhaust
Fuel pump
Intercooler and its hoses
Induction kit placement, hoses, intake pipes
Resonator, airbox etc etc.
Fuel octane, additives?

Plus a few others at least which I haven't thought of.

Very difficult to know how much each compensates and where in the rev range and at what loads without doing a controlled study of lots of parts alone and in combination - ie not feasible.

[Edited by john banks - 6/6/2002 11:11:55 PM]

WREXY

Cheers John!

They reckon in 15 days time we will be getting the proper Shell Optimax fuel like you guys have in the UK. Apparently it will be imported from the UK and will be called V Power Super. We already have V Power unleaded, but it is only 95 RON. They also reckon BP is going to come up with an equivalant fuel to compete. I hope it happens.

Cheers,

Wrexy.

dowser

Theo...... *whisper*....remove the resonator, lol!

Richard

RT

On the DeltaDash, whats the "knock correction" units? Is +ve retard and -ve advance, or vice versa?

Also, is the "ignition timing" the actual timing you get? Or is it the "ignition timing + knock correction" the actual timing you get?

tks.

dowser

RT,

The knock correction value is degrees, either '+' or '-' (advance/retard respectively). The ignition value is the actual value after factoring knock correction.

Richard

RT

Thank you! That figues out the convention for me.

john banks

+ is advance, - is retard. On MY01 and after it routinely runs +8 or so during boost. The MY00 and before typically run +-0.5 if the timing map is about right in good conditions, in poor conditions the standard car can get occasional -2, more frequent -1. The timing reported is the final value, but it has an offset (10 degrees I believe), but the offsets between the models are rumoured to be different.

EvilBevel

OK, drove the car again today, and what I felt/read with DeltaDash yesterday was confirmed with the pants dyno today.

My car runs a *lot* better with the OEM setup, no doubt about it.

I'm not saying that is the case with every induction kit, but it certainly is on mine.

Oh, and even the EGT at idling drops quicker than before Where it never used to drop below 400°C after a hard run, it now went to 370-ish °C after 4 minutes.

Richard, LOL ... do you think I would gain power with that ?

EvilBevel

PS: John, my car seems to run 0 knock correction all the time... is that good/bad or in between ?