AFR going to low ALWAYS without KR

Discussion of tuning specific to MAZDASPEED6/MAZDASPEED Atenza/Mazda 6 MPS vehicles
Post Reply
User avatar
keridil
Posts: 58
Joined: March 2nd, 2012, 6:05 am
Location: Asturias / Spain

AFR going to low ALWAYS without KR

Post by keridil »

Hi there,

Since a while my AFR is always going pretty rich to 0.715 lambda which is my Min Desired Lambda value without any reason for it. I'm commanding 0.7347 in Open Loop at 6,000 rpm. and I'm not seeing any knock, my boost (21.5 psi) pressure is what I command and even my ignition advance (10º) is perfect.

What in the heck is happening? Any ideas or advise will be very welcome.

Cheers guys!
Mazda 6 MPS, BPV GFB, RX-8 wheels, B8 + Eibach Pro-Kit, CS 3" SRI, GS EBCS (interrupt), CS 3" TMIC, EGR off, HPFP internals, CS catted downpipe, 3 bar MAP, BNRS3, forged engine and stuff
User avatar
Steve @ VersaTune
Lead tuner
Posts: 1157
Joined: March 29th, 2010, 12:58 pm

Re: AFR going to low ALWAYS without KR

Post by Steve @ VersaTune »

Please post or send a copy of the tune and a log showing the behavior.
User avatar
keridil
Posts: 58
Joined: March 2nd, 2012, 6:05 am
Location: Asturias / Spain

Re: AFR going to low ALWAYS without KR

Post by keridil »

Steve @ VersaTune wrote:Please post or send a copy of the tune and a log showing the behavior.
Yeah, of course, my bad Steve.

Find attached the map, please realise that WGDC is 5% lower than in the logs I'm sharing because I just gave some extra pressure to my BNR spring, the rest is exactly the same.

Full log: http://goo.gl/LL4RBM

Image

Thanks in advance.
Attachments
Keridil v1.69 eco ITV 24psi v1.69.vtune.zip
MS6 BNRS3 CS 3" CS DP catted 23.5 psi 6,700 rpm
(83.24 KiB) Downloaded 424 times
Mazda 6 MPS, BPV GFB, RX-8 wheels, B8 + Eibach Pro-Kit, CS 3" SRI, GS EBCS (interrupt), CS 3" TMIC, EGR off, HPFP internals, CS catted downpipe, 3 bar MAP, BNRS3, forged engine and stuff
User avatar
keridil
Posts: 58
Joined: March 2nd, 2012, 6:05 am
Location: Asturias / Spain

Re: AFR going to low ALWAYS without KR

Post by keridil »

Just flashed a map with all the fuel tables including knocking an overload sharing the same values, same outcome, ECU is demanding 0.715 lambda (10.5:1 AFR). There is no knock on the logs. Are High BAT/ECT Retard No Knock & Knocking forcing Min Desired lambda once a cell is not 0 for a given BAT and ECT?
Mazda 6 MPS, BPV GFB, RX-8 wheels, B8 + Eibach Pro-Kit, CS 3" SRI, GS EBCS (interrupt), CS 3" TMIC, EGR off, HPFP internals, CS catted downpipe, 3 bar MAP, BNRS3, forged engine and stuff
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: AFR going to low ALWAYS without KR

Post by mituc »

For those who want to take a look jump directly to like 2772, 5351, 5774, 6785, 6816, 7154.

Once you step into the section of the ECT/BAT retard (knock or no knock) tables where the timing retards are not zero you will also get this AFR enrichment as an unwanted bonus. The ECU is trying to protect the engine as on this bath you're going to crack the ring lands on the middle pistons. Get a meth kit, even a small DO5 nozzle (even though a DO10 or DO12 is recommended in this case at these air flow values) will make wonders when it comes to efficiency, knock protection and internals cooling.
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
User avatar
keridil
Posts: 58
Joined: March 2nd, 2012, 6:05 am
Location: Asturias / Spain

Re: AFR going to low ALWAYS without KR

Post by keridil »

mituc wrote:For those who want to take a look jump directly to like 2772, 5351, 5774, 6785, 6816, 7154.
Sorry @mituc; what is that?
mituc wrote:Once you step into the section of the ECT/BAT retard (knock or no knock) tables where the timing retards are not zero you will also get this AFR enrichment as an unwanted bonus. The ECU is trying to protect the engine as on this bath you're going to crack the ring lands on the middle pistons. Get a meth kit, even a small DO5 nozzle (even though a DO10 or DO12 is recommended in this case at these air flow values) will make wonders when it comes to efficiency, knock protection and internals cooling.
Yeah, we spoke about it some months ago. I even ordered a DevilsOwn kit and had some trouble so I forgot about it.

The thing is I'm not seeing knock and even under 60ºC around 50ºC I'm seeing the mixture enrichment. So your logic seems correct but I believe we are missing something in the picture here. I was configuring Min Desired Lambda to 0.715 because that mixture is the richest one in knock tables on a stock tune, my knock tables are stock, just in case, I believe is the thing to do.

BTW, if I crack ring lands in a forged engine I'll dump this car for good. ;-P

Cheers.
Mazda 6 MPS, BPV GFB, RX-8 wheels, B8 + Eibach Pro-Kit, CS 3" SRI, GS EBCS (interrupt), CS 3" TMIC, EGR off, HPFP internals, CS catted downpipe, 3 bar MAP, BNRS3, forged engine and stuff
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: AFR going to low ALWAYS without KR

Post by mituc »

keridil wrote:
mituc wrote:For those who want to take a look jump directly to like 2772, 5351, 5774, 6785, 6816, 7154.
Sorry @mituc; what is that?
Those are the lines in that huge logs where you have some WOT pulls :)
When looking at a huge log like that it's pretty inconvenient to locate the relevant sections which are relatively small.
keridil wrote: The thing is I'm not seeing knock and even under 60ºC around 50ºC I'm seeing the mixture enrichment. So your logic seems correct but I believe we are missing something in the picture here. I was configuring Min Desired Lambda to 0.715 because that mixture is the richest one in knock tables on a stock tune, my knock tables are stock, just in case, I believe is the thing to do.
You're not seeing knock because there is no knock, but there is some timing retard because of the high BATs at least, not sure about the ECTs because it is not logged. The table relevant in this discussion is Ignition -> High BAT/ECT Retard No Knock. I'm not sure if there's another table similar to this which instead of the ignition retard introduces enrichment or if the ECU simply reacts with the enrichment amount based on the non-zero values in this table (remember they are interpolated, so at 95C ECT you still get like 0.35 degrees less timing based in the values in your table).

Something else that I've seen affecting the AFRs is reaching the boost limits (throttle reduction int his case). You've set the load limit to 2.3 across the board and you reach the throttle reduction limits (Absolute Limits -> Overboost Protection Throttle Reduction) below 5000rpm.
Now you have a pretty large set of logs, you know where your load is based on the limits of the turbo or the boost you want, you should change the per gear load targets/limits accordingly and let the load limit logic to deal with limiting load/boost instead of a protection system (overboost protection).
keridil wrote: BTW, if I crack ring lands in a forged engine I'll dump this car for good. ;-P
Oh, I noticed that just now :) , it's at the end of your signature. Well, no, on a forged engine you will not get cracked ringlands with the load and timing you're pushing, but some extra cooling and more octane for more timing would help quite a lot.
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
User avatar
keridil
Posts: 58
Joined: March 2nd, 2012, 6:05 am
Location: Asturias / Spain

Re: AFR going to low ALWAYS without KR

Post by keridil »

mituc wrote:Those are the lines in that huge logs where you have some WOT pulls :)
When looking at a huge log like that it's pretty inconvenient to locate the relevant sections which are relatively small.
That's why my link was zoomed in just to a precise section. :-P
mituc wrote:You're not seeing knock because there is no knock, but there is some timing retard because of the high BATs at least, not sure about the ECTs because it is not logged. The table relevant in this discussion is Ignition -> High BAT/ECT Retard No Knock. I'm not sure if there's another table similar to this which instead of the ignition retard introduces enrichment or if the ECU simply reacts with the enrichment amount based on the non-zero values in this table (remember they are interpolated, so at 95C ECT you still get like 0.35 degrees less timing based in the values in your table).
The thing to note here is that I get the commanded timing. Which puzzles me even further.
mituc wrote:Something else that I've seen affecting the AFRs is reaching the boost limits (throttle reduction int his case). You've set the load limit to 2.3 across the board and you reach the throttle reduction limits (Absolute Limits -> Overboost Protection Throttle Reduction) below 5000rpm.
That is expected for those logs with Requested Load per gear tables not filled up.
mituc wrote:Now you have a pretty large set of logs, you know where your load is based on the limits of the turbo or the boost you want, you should change the per gear load targets/limits accordingly and let the load limit logic to deal with limiting load/boost instead of a protection system (overboost protection).
No worries, car is running now with Requested Load per gear limits, Load Target limits, Baro Vs. RPM limits and APP Requested Load values in place. All of them around 0.05 higher than seen in logs. So I believe I'm good to go. ;-)
Mazda 6 MPS, BPV GFB, RX-8 wheels, B8 + Eibach Pro-Kit, CS 3" SRI, GS EBCS (interrupt), CS 3" TMIC, EGR off, HPFP internals, CS catted downpipe, 3 bar MAP, BNRS3, forged engine and stuff
mituc
VersaTuner guru
Posts: 1324
Joined: December 17th, 2011, 2:47 pm
Location: Iasi/Romania

Re: AFR going to low ALWAYS without KR

Post by mituc »

keridil wrote:The thing to note here is that I get the commanded timing. Which puzzles me even further.
No you don't, the actual timing is lower :)

By example look at like 5357. You get the commanded timing and AFR by 6000rpm. Well, you command 10.8 and you get 10.7, that's pretty close, but the ECU also commands that value which you hit perfectly.
From that point on instead of climbing towards 10 degrees if timing commanded at 6500rpm and AFR also 10.8 the AFRs dip towards 10.5 and the timing is still about one degree behind. All this time the MAF g/s remains around 330g/s where it also was at 6000rpm where the AFR was pretty close to the commanded one. So the MAF cal you have is also perfect, or probably it's because you benefit from adjustability there because the closed loop load limits are set to 3 in the Absolute Limits section which I think makes the STFTs work continuously for you (but it doesn't really matter how this is achieved, as the commanded and actual are very close).

That is even more obvious in the pull starting around line 5774 in your log.

Personally instead of raising the lower limit of the high bat flag from 70C to 80C I'd lower it from 70 to 60-65C. Anyway, spraying meth will fix you the high BAT issues especially in the hot weather in Spain, and will also add a tremendous amount of protection and cooling to everything.
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
User avatar
keridil
Posts: 58
Joined: March 2nd, 2012, 6:05 am
Location: Asturias / Spain

Re: AFR going to low ALWAYS without KR

Post by keridil »

mituc wrote:
keridil wrote:The thing to note here is that I get the commanded timing. Which puzzles me even further.
No you don't, the actual timing is lower :)
Yeah, you are right just, checked again the section that I linked in the OP and I'm around 1º less than commanded, it makes sense after crossing BAT with 90º ECT I'm finding 0.9º retard in that cell around 6,600 rpm.
mituc wrote:So the MAF cal you have is also perfect, or probably it's because you benefit from adjustability there because the closed loop load limits are set to 3 in the Absolute Limits section which I think makes the STFTs work continuously for you (but it doesn't really matter how this is achieved, as the commanded and actual are very close).
It's a pretty good MAF calibration, yeah, and my Closed Loop Load Limits are set to 3 to assure that STFT in OL are doing it's job even in loads over 2.5, it was in purpose.
mituc wrote:Personally instead of raising the lower limit of the high bat flag from 70C to 80C I'd lower it from 70 to 60-65C. Anyway, spraying meth will fix you the high BAT issues especially in the hot weather in Spain, and will also add a tremendous amount of protection and cooling to everything.
It was for that logs, I have it set to 70º C by default and meth is a no go for now, probably I will be moving to Australia after the Summer and I don't feel like modifying the car further, I'm sure I will be selling it indeed. Anyway, I will be filling up my tank with E15 from now on and that will help a bit, I believe.

Thanks, mituc.
Mazda 6 MPS, BPV GFB, RX-8 wheels, B8 + Eibach Pro-Kit, CS 3" SRI, GS EBCS (interrupt), CS 3" TMIC, EGR off, HPFP internals, CS catted downpipe, 3 bar MAP, BNRS3, forged engine and stuff
Post Reply