Lean @ WOT

Discussion of tuning specific to MAZDASPEED3/MAZDASPEED Axela/Mazda 3 MPS vehicles
Dolfius
Posts: 64
Joined: May 31st, 2017, 2:54 am

Lean @ WOT

Post by Dolfius »

Hello and welcome to another episode of WTF. I hope someone can help shed some light on this. Yes, I have searched and tried reading up on this problem, but have yet to come across any solution.

No matter hoe much I calibrate/scale the MAF, I keep missing AFR targets (on the lean side) when going WOT. By a lot. I'm at this point 99.99999% sure it's not hardware related. Had a tiny exhaust leak on the external wastegate - fixed that. No boost leaks. Certainly no vacuum leaks.

So what am I missing? Map and a log attached for reference. The map is a work in progress and farfrom perfect, as I need to get MAF and target AFR down before I can continue with tuning.


Some history:

Car previously had a built motor with a Holset HX40 turbo and Cobb custom map (tuned by myself) - no issues. Life was peachy until the motor shit the bed (long story).

Sold the AP while the car was down. Eventuallyu got to building it again last year, and decided to give VT a go. Had the lean WOT problem back then already. But there were so many small issues to sort out, that I never got to tuning it properly.

The built motor eventually ran bearings to such an extent that the mtor had to be scrapped (even longer story). Replaced that with a stock mk2 motor, and a smaller T3/T4 turbo. And lo....still the same problem. The interesting thing is, I get next to no KR while it's running this lean. And by lean, I mean 12.7+ AFR when it should be 11.4.

The other interesting thing is a friend of mine is having the exact same issue on his BT Mk2. Never had sw before - went straight VT. Installed a custom BT kit after his k04 decided to grenade itself. Also running lean at WOT. Increasing MAF values has little to no impact in honing in on AFR targets.

What are we "doing wrong"?

My mods list:
Custom twin scroll manifold.
T3/T4 china charger
Turbosmart External Wastegate, with screamer pipe
Custom 4" intake
Custom FMIC
El cheapo (but good) BOV, set to VTA
VTCS delete
Custom 3" Exhaust from turbo to tailpipe

Any advice, thoughts, suggestions - even questions - will be greatly appreciated.
Attachments
Pump 2018-07 Rev 2 v1.00.vtune
Map used for WOT Log
(125.13 KiB) Downloaded 503 times
Tuning Logging - 2018-07-29 21.24.28.csv
WOT Log
(10.25 KiB) Downloaded 441 times
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: Lean @ WOT

Post by mituc »

Weird indeed. The ECU seems to be adjusting for something and is doing it wrong.

So try this: under the Absolute Limits section you have Closed Loop Load Limit, 1 to 3. Set all the values there to 1 - 1.1 - 1.2 or something instead of the default 2.5. This should prevent the ECU make any fuel trim adjustments past that load value and let your MAF cal dictate the AFRs, assuming it's good.
2008 Cosmic Blue Mazda 3MPS
Built engine + WMI + GTX3071 gen2, ~509BHP @34PSI
2008 Icy Blue Mazda CX7
Stock engine and exhaust (YES!!), JBR3" + GTX2867 gen2 + Autotech HPFP, self-tuned to 330-ish BHP
Dolfius
Posts: 64
Joined: May 31st, 2017, 2:54 am

Re: Lean @ WOT

Post by Dolfius »

Thanks Mituc, I'll try that. I'm pretty sure my MAF cal is out by quite a bit because of my trying to get it dailed in to no avail.

Hopefully this have the effect you described and I can calibrate the MAF from there.

Sent from my LG-H870 using Tapatalk
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: Lean @ WOT

Post by mituc »

Well, log the MAF Voltage, desired AFR and actual AFR. Forget about STFTs, LTFTs and all known MAF calibration procedures, with these 3 PIDs you can get an accurate MAF calibration for the OpenLoop section in 1-2 iterations by multiplying the corresponding MAF g/s in your MAF cal with 1-(desired-actual)/14.7. Or eventually take smaller steps with 1-0.75*(desired-actual)/14.7.
So since you're doing this for the open loop section of your MAF cal, setting lower load values in the above mentioned tables prevent the ECU stop pursuing automatic fuel trims adjustments as soon as you enter boost basically. This way there is going to be a natural/real deviation between the actual and desired values, deviation which you will be able to adjust in your MAF cal.

It is very important not to have any leaks, because with this approach you can start tuning around defects quite easily.
2008 Cosmic Blue Mazda 3MPS
Built engine + WMI + GTX3071 gen2, ~509BHP @34PSI
2008 Icy Blue Mazda CX7
Stock engine and exhaust (YES!!), JBR3" + GTX2867 gen2 + Autotech HPFP, self-tuned to 330-ish BHP
Dolfius
Posts: 64
Joined: May 31st, 2017, 2:54 am

Re: Lean @ WOT

Post by Dolfius »

Made the recommended adjustments, loaded the map this morning and on my way to my first client, did a quick 4th gear WOT log. This was more just out of sheer curiosity to see what the AFR would show. I have yet to log this adjustment officially (maybe later tonight), but it seems the lean condition worsened.
Can MAF really be that far out? Impromptu log attached for reference.
Attachments
Tuning Logging - 2018-07-31 11.01.51.csv
(6.28 KiB) Downloaded 383 times
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: Lean @ WOT

Post by mituc »

The MAF cal is pretty bad, and that I think may be the good news. the lean condition worsened because the ECU did not compensate for the missing fuel past load 1, but apart from wanting to see how much from that MAF cal is missing I never imagined it's going to be that bad, and I apologise for that because I could have anticipated if I paid more attention.

So for 17PSI of boost around 5500rpm your current MAF cal says there are about 240g/s of air entering the engine. In my opinion there should be close to 250-255g/s of air (depending on weather) on a car with stock intake manifold (no VCTS delete). With deleted VCTS I think you can easily add at least 10 more g/s of air at this boost, but it should be a bit more (like 15 or so).

So 265/240 =~ 1.10, so about 10% difference which yes, I agree it's less than the difference you see in the AFRs (about 18-20%).
Also, those 240g/s of air are registered at 3.6V of MAF voltage. This is an air flow value typical for a 3.5" intake at that voltage, not for a 4".

Based on all these facts and I'd do now is take a good known 3.5" MAF cal, multiply the whole thing by 1.3 (which is the cross section difference between a 3.5 and a 4" both ID) and adjust from there.
If the ECU will have to adjust down by 10% (let's say your intake is not exactly 4" ID, or the reference intake is not exactly 3.5" ID) - not a problem, a lot better if 20% are actually missing.
Here is a good JBR MAF cal I perfected in the last few months for the 3.5I silicone path intake:
http://www.versatune.net/forum/viewtopi ... 8319#p8319
Use that one, add as many percents as you think (I'd multiply that by with 1.20-1.22-1.25 for starters) and take it from there.

I hope that now you're going to be on track to finally fix that tune. Keep us posted!

P.S.: when you have such situations take logs in a lower gear, such as 3 if it can keep the wheels on the ground and not spin, it will keep the engine under load in less than ideal conditions for less time than a 4th gear pull.
2008 Cosmic Blue Mazda 3MPS
Built engine + WMI + GTX3071 gen2, ~509BHP @34PSI
2008 Icy Blue Mazda CX7
Stock engine and exhaust (YES!!), JBR3" + GTX2867 gen2 + Autotech HPFP, self-tuned to 330-ish BHP
Dolfius
Posts: 64
Joined: May 31st, 2017, 2:54 am

Re: Lean @ WOT

Post by Dolfius »

Very valuable info there Mituc, thank you! I'll give feedback on this real soon.
Dolfius
Posts: 64
Joined: May 31st, 2017, 2:54 am

Re: Lean @ WOT

Post by Dolfius »

Done the MAF cal....and almost no change.....


Logs attached. All 3 4th gear.
Attachments
Tuning Logging - 2018-07-31 23.39.18.csv
Post MAF cal WOT log 3
(8.44 KiB) Downloaded 391 times
Tuning Logging - 2018-07-31 23.38.25.csv
Post MAF cal WOT log 2
(12.86 KiB) Downloaded 407 times
Tuning Logging - 2018-07-31 23.37.19.csv
Post MAF cal WOT log 1
(9.64 KiB) Downloaded 345 times
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: Lean @ WOT

Post by mituc »

There's still a difference of almost 12%... however the MAF g/s values match better what that 4" intake should flow.

So at this point if you do not have a leak between the MAF and the turbo somewhere, then you may have a leak on the exhaust side between the head and primary O2 sensor somewhere, and it can be quite a lot of places (manifold to head, manifold to turbo, the turbo itself, turbo to downpipe), or the primary O2 is done.
2008 Cosmic Blue Mazda 3MPS
Built engine + WMI + GTX3071 gen2, ~509BHP @34PSI
2008 Icy Blue Mazda CX7
Stock engine and exhaust (YES!!), JBR3" + GTX2867 gen2 + Autotech HPFP, self-tuned to 330-ish BHP
Dolfius
Posts: 64
Joined: May 31st, 2017, 2:54 am

Re: Lean @ WOT

Post by Dolfius »

Can a dodgy primary o2 work during PT driving, yet go screwey when going WOT?
Post Reply