Random Lean @ WOT

[ms6mil]

I had a decent map on for the last couple of days. I was trying to work out the dip in boost in the 5k RPM range, but other than that, it was running fine.

This morning, did a single log on the way to work and at around 4600 RPM my AFR's went from the 11.7(tgt) to 13's. Nothing on the map was changed from one day to the next. No change in fuel, or e85.

It is currently on boost tune, in the process of trying to collect the load info in all gears to put my hybrid tune back on.

Going to check for leaks, after work. Attached below is yesterdays(datalog3&5) and todays(log1) logs. Also attaching the VD

Curious as to why my car is being a bitch!

[ms6mil]

I got rid of the boost oscillation, AFR's are fine now. I don't understand why this happened. Anyone have any ideas? (Checked for leaks), maybe a sensor going bad or Karma...

[ms3blackmica]

Did you change the value of the Injector Phasing Multiplier from 1.00?

If so, to what value?

[ms6mil]

Originally Posted by ms3blackmica Did you change the value of the Injector Phasing Multiplier from 1.00? If so, to what value?

First off I would like to apologize for resurrecting this thread, also for not replying.. I deployed shortly after and am now back home trying to get this car figured out.

At the time Injector scaling was as follows




~29% e85

Fast Forward to now.

Same settings, it was running fine, then the random lean @ WOT occurred again, this time a bit more often. After reading a few threads and talking to some other folks I did the following.

1. Leak test #1
2. Changed Inj. Scaling tables to account for summer/winter blend.
3. Changed Inj. Scaling back to OTS.
4. Ran out the remaining e85.
5. Removed Pump, cleaned thoroughly, to incl. spill valve. wasn't bad but wasn't great, pic's attached...
6. Boost Leak test #2
7. Full tank of 93, started and loaded a new map.
8. MAF/WOT calibration complete.. LTFT's +/-4

Was still having occasional occurrence of Lean during a pull.

9. Swapped out new NGK's gapped @ .026
10. Boost Leak test #3
11. Drove to AL Meet. AFR's were stable
12. Compression Test done 190/190/190/190
13. Did a log this morning and FUCK ME AFR went 11.9>12.1>13.1
14. Did a log on the way home and everything seemed better.
15. I have also done a smoke/vac test with nothing remarkable to note.

Attaching a couple of logs, and another screenshot of the the VD for reference..



Of note the most recent mod to go in is the Guardian Angel. Haven't tried removing yet. But this did happen prior to deployment before I even got the GA. What's really got me that even an OTS w/HTP maf scaled map that was working before is giving me same results.

Could this be an O2 sensor issue?
Any input is appreciated.

[Agent_Orange]

Try cleaning the MAF sensor?

[ms6mil]

Originally Posted by Agent_Orange Try cleaning the MAF sensor?

sorry forgot that one, but yes that was done as well..
I'm hoping it's just something that obvious.

[SarcasticOne]

On phone, can't view logs, but did you go over load values of 2.5?

Edit: @Lex; just posted this: http://www.mazdaspeedforum.org/forum/foru....php?p=2412336

Sent from my Nexus 5 using Tapatalk

[ms6mil]

Originally Posted by SarcasticOne On phone, can't view logs, but did you go over load values of 2.5? Sent from my Nexus 5 using Tapatalk

Yes, and I thought about that too, but current map load tables are all beyond. 2.5. and I have hit beyond 2.5 with previous maps with no issues..

open to suggestions

EDIT: thanks for the link looking through the thread now

[atvfreek]

Originally Posted by ms6mil Yes, and I though about that too, but current map load tables are all beyond. 2.5. and I have hit beyond 2.5 with previous maps with no issues.. open to suggestions EDIT: thanks for the link looking through the thread now

What is your throttle requested load A,B and C tables set to?

[ms6mil]

Originally Posted by atvfreek What is your throttle requested load A,B and C tables set to?

Currently OTS @ 3.0, Looking over Lex's thread about this very issue.

[ms6mil]

Update:
I went out around 4am, I figured these cooler temps in the morning would help be replicate my 2.5 load a bit easier.

The changes were made according to @Lex; thread. AFR targets at loads higher than 2.50 ATR/ATP fix






As you can see in the logs attached the very moment it hits 2.5 load, it is still going lean then evens out by the end. As long as I stay under 2.5, AFR's are good.




Here's another VD just to reference the AFR's going lean compared to the log 3 which stayed under 2.5



I am not sure what I've missed, any input is appreciated. @atvfreek; any luck leaving the toggle unchecked

I'm going to try a boost only, hybrid and then load only map and try to toggle/untoggle the check box, also gonna try the increasing load limit a cunt hair.

@rfinkle2; @Mike@Stratified;

[rfinkle2]

I am not tuning anything that is experiencing this issue, but would you be willing to try scaling the load axis for the fueling tables?

If you do decide to, I have noticed that there is a chance that your ignition tables will be affected so please take a look @ those to make sure everything jives before flashing.

[ms6mil]

Originally Posted by rfinkle2 I am not tuning anything that is experiencing this issue, but would you be willing to try scaling the load axis for the fueling tables? If you do decide to, I have noticed that there is a chance that your ignition tables will be affected so please take a look @ those to make sure everything jives before flashing.

working on that right now, will post results later today. Thanks man.

[ms6mil]

Well I have tried a few different ways and all results are the same. Once 2.5 hits, AFR's go lean. I can successfully limit load below 2.5 but the fact that I can't figure out why, when strictly boost tuning, this issue keeps popping up. what am I missing here?



These aren't as bad as before bust still see it go up to 12.4 in some spots

[rfinkle2]

Originally Posted by ms6mil Well I have tried a few different ways and all results are the same. Once 2.5 hits, AFR's go lean. I can successfully limit load below 2.5 but the fact that I can't figure out why, when strictly boost tuning, this issue keeps popping up. what am I missing here? These aren't as bad as before bust still see it go up to 12.4 in some spots

Can you post a log that includes all calculated loads from your closed loop max load and redline when you get a chance?

[ms6mil]

Originally Posted by rfinkle2 Can you post a log that includes all calculated loads from your closed loop max load and redline when you get a chance?

will do, gonna go grab some logs. muchas gracias

[ms6mil]

Originally Posted by rfinkle2 Can you post a log that includes all calculated loads from your closed loop max load and redline when you get a chance?

per your request...

[rfinkle2]

If you would like to try this maf curve, it is for voltages 2.97 -5.0. You will need to do some smoothing in the closed loop / lower voltages.

105.05	106.61	108.98	112.13	116.02	119.40	124.50	130.25	136.62	143.56	151.04	159.03	167.50	176.41	185.73	192.97	202.91	213.17	223.71	234.49	245.49	256.65	267.96	279.38	290.87	302.40	311.05	322.56	334.02	345.38	356.63	367.71	378.60	389.27	399.67	409.78	419.56	426.66	435.77	444.47	452.71	460.45	467.67	474.32	480.39	485.82	490.59	493.71	497.23	499.99	501.97	503.11	503.40

[ms6mil]

doing it now, thanks. Will post up results in a few..

Edit: just viewing the curve on this, if this addresses the lean afrs at the point of 2.5 load, what is likely to be seen if load remains under 2.5 throughout WOT? About to go get some more logs now just curious what is to be expected..

Thanks again

[rfinkle2]

I saw that you were targeting 11.7 afr's, so it is just a corrected curve based on your fuel targets.

[ms6mil]

Took some logs last night. I don't know what to make of this load I'm hitting now. Assuming its due to the maf cal being a bit off, and its relation to calculating load?



What I don't understand is, I did the tips suggested by @Lex; for the 2.5 load issue, why is it still happening? I've haven't seen an issue like this until this past August, when I started seeing the random spikes in AFR's @ 2.5 load

In reference to the MAF cal, I had been using Lex's MAF calibrations for bigger intakes for the HTP 3.5" so I'm curious as to the remarkable difference in the curve between the two. Historically I haven't had to make too many changes after loading the HTP scale from that thread, attaching a side by side to view the difference in in g/s in relation to the maf voltage.



I'm probably just geeking out over this, but would like to find the source of my problem.

I appreciate any insight on this from you pro's.

[rfinkle2]

Correct, but now the afr's are correct (or @ least closer).

Multiplying the stage 0 fuel curve to get a base maf curve is absolutely a good starting point, but the base maf curve is almost a perfect 3rd order polynomial.

In the real world, maf curves are not perfect in shape due to hardware configuration and a million and 1 variables.

Any time you change the maf curve, you will see a change in calculated load... In this instance, your load was short a few 10ths because your maf cal was a touch off..

Speculation on my part, but the latest version of ATR (as atvfreek mentioned) must be less forgiving in terms of phantom open loop trimming.

One thing of note:

If you correct the maf curve using a 3rd or 4th order curve, there is a chance you will run into a portion that curves back on itself, so you will want to make sure that your corrected curve makes sense in terms of flow before flashing.

[Monotonous ONE]

[rfinkle2]

Mono!

[ms6mil]

Originally Posted by rfinkle2 Correct, but now the afr's are correct (or @ least closer). Multiplying the stage 0 fuel curve to get a base maf curve is absolutely a good starting point, but the base maf curve is almost a perfect 3rd order polynomial. In the real world, maf curves are not perfect in shape due to hardware configuration and a million and 1 variables. Any time you change the maf curve, you will see a change in calculated load... In this instance, your load was short a few 10ths because your maf cal was a touch off.. Speculation on my part, but the latest version of ATR (as atvfreek mentioned) must be less forgiving in terms of phantom open loop trimming. One thing of note: If you correct the maf curve using a 3rd or 4th order curve, there is a chance you will run into a portion that curves back on itself, so you will want to make sure that your corrected curve makes sense in terms of flow before flashing.

Got it! thanks again, gonna make some adjustments now and go grab some more logs.

[chaser27]

Originally Posted by ms6mil Got it! thanks again, gonna make some adjustments now and go grab some more logs.

I'm curious to see what will happen if you don't hit 2.5 loads after adjusting your MAF curve to correct the issue.

[ms6mil]

Update: Went out and logged after corrections to the MAF.... AFR's are much better. Load is now reaching.... well you'll see.

I had @leeroy38; in the car so he can see these loads I was hitting.

So at this point my load and g/s are more than I have ever seen. Until I understand what is going on I'm gonna load limit the car to under 2.5 and read some more. Once again I appreciate the help so far..

Gonna go do some more logs. had to lay low for a bit. cops almost had me earlier

[rfinkle2]

Originally Posted by chaser27 I'm curious to see what will happen if you don't hit 2.5 loads after adjusting your MAF curve to correct the issue.

My $ says they will be dead on.

[chaser27]

Originally Posted by rfinkle2 My $ says they will be dead on.

I wasn't doubting you Haha that's why you're a tuner and I'm not. I'm trying to figure out how all this shit works

Sent from the lead

[Agent_Orange]

Originally Posted by rfinkle2 My $ says they will be dead on.

Fly by night tuning! FTW!

[rfinkle2]

Originally Posted by chaser27 I wasn't doubting you Haha that's why you're a tuner and I'm not. I'm trying to figure out how all this shit works Sent from the lead

I absolutely didn't mean to discourage you in any way.

To adjust open loop fueling, (all fueling after the ecu has passed what closed loop max load you choose), you use: air fuel ratio logged / desired air fuel ratio.

To create an maf curve for a larger diameter maf housing, you multiply the area of the maf sensor housing by the % difference over the stock diameter.

Because the stock diameter maf curve that is often multiplied against is not always the shape of the maf curve that your car follows, the shape of the curve needs to be changed to hit fueling targets.

If you look @ many of the areas of the OTS curve posted above, and the curve that was a bit closer to targets, you'll see some areas are actually very similar in terms of voltage per grams/ second.

Not bad for a fly by night, non-engineer tuner huh?

[chaser27]

None offense taken. I have my OL MAF curve dialed in. It just acted really weird when I hit 2.5 load this morning. Maybe it's a gen pu thing? Idk. I posted another log after it warmed up outside in the other thread.

Sent from the lead

[ms6mil]

Originally Posted by chaser27 None offense taken. I have my OL MAF curve dialed in. It just acted really weird when I hit 2.5 load this morning. Maybe it's a gen pu thing? Idk. I posted another log after it warmed up outside in the other thread. Sent from the lead

I don't think its just a genpu thing. Back in August I was, or so I thought, pretty much dialed in as well. (except to the one lean log) if you look at the beginning of the thread you can see the car was running great, never seeing a load above 2.5.
Then I deployed came back in the Winter. ran the same map, and because of the cooler temps started seeing higher loads. +2.5 loads brought on the erattic AFR's. Some seem to be seeing AFR's go rich, others, myself included are seeing them go lean. This was an anomaly that I'm sure could be explained eventually, but seems to be under control now. (thanks to @rfinkle2

[chaser27]

Originally Posted by ms6mil I don't think its just a genpu thing. Back in August I was, or so I thought, pretty much dialed in as well. (except to the one lean log) if you look at the beginning of the thread you can see the car was running great, never seeing a load above 2.5. Then I deployed came back in the Winter. ran the same map, and because of the cooler temps started seeing higher loads. +2.5 loads brought on the erattic AFR's. Some seem to be seeing AFR's go rich, others, myself included are seeing them go lean. This was an anomaly that I'm sure could be explained, but seems to be under control now. (thanks to @rfinkle2

But what did you end up doing Rob? You recalibrated that spot in your MAF curve using the log where you hit 2.5+? I did another log today. Hit 2.42 or so and everything looked great.

Sent from the lead

[rfinkle2]

Originally Posted by chaser27 But what did you end up doing Rob? You recalibrated that spot in your MAF curve using the log where you hit 2.5+? I did another log today. Hit 2.42 or so and everything looked great. Sent from the lead

I think the general consensus is that in past versions of ATR, there was a phantom open loop trim that must've had the ability to correct the fuel curve by a larger %.

Some guys seem to have trouble @ 2.3 load, some 2.5, We'll figure it out in time.

This is just my opinion, but if correcting the maf curve works, that is what we have available to us @ this moment... until we can get Cobb to confirm that there was a change made or reveal something about the trims that we don't currently know about.
@chaser27; I edited the the entire open loop fuel curve using excel.

[atvfreek]

I've got both cars situated that were going lean. It definitely has to do with OL trims saying FU over 2.5 load (could vary from car to car) Although the maf cal appears to be perfect due to Afr being spot on below 2.4 load. It could in fact be off by a few percent. One car was only off by 3% but that is the difference between 11.8 and 12.2.

[ms6mil]

Originally Posted by chaser27 But what did you end up doing Rob? You recalibrated that spot in your MAF curve using the log where you hit 2.5+? I did another log today. Hit 2.42 or so and everything looked great. Sent from the lead

Ha both Robs, yeah... pretty much what he said, I took his re-calibrated curve and made some normal WOT adjustments now AFR's are within limits. I'm gonna be posting up some more logs in the morning. Along with screenshots of LEC, &TRL X tables and anything else relevent. Thanks again Rob.

[chaser27]

So just a little update. I corrected my MAF curve using the log where I went really rich and acted like there was a phantom OL fuel trim that disappears at 2.5 load. Dropping it to ~87% of where it was. (~10/11.4) Smoothed everything out and drove around for 50 miles or so to make sure my LTFT's were still in check before taking a 4th gear log.

AFR's were spot on just like @rfinkle2; said they would be

So if I'm able to move my MAF curve > 10% and AFR's stay good, I'm thinking there are definitely some OL fuel trims that we can't see working. Which to me means our current methodology for calibrating the OL MAF curve is wrong.

Originally Posted by rfinkle2 To adjust open loop fueling, (all fueling after the ecu has passed what closed loop max load you choose), you use: air fuel ratio logged / desired air fuel ratio.

This won't really do anything if the ECU is trying to do the same thing.

However, you can see something peculiar about the tip-in, let-off points in the image attached. Before I scaled down my MAF curve by ~13%, during tip in the AFR's would go rich for a split second. and then when I let off they would go straight up to 29.95 or whatever.
After my adjustment, I'm seeing no real drop in AFRs on tip-in, but a moment of richness when I let off. This is leading me to believe that I've over corrected my MAF curve. I'm going to try scaling it back up a little and see if this is fixed.

Again, I'm the farthest from an expert on this subject so @Lex; @rfinkle2; @atvfreek; @superskaterxes; @phate; ,etc. feel free to offer your opinion.

[chaser27]

forgot to add this

[Lex]

Originally Posted by ms6mil Took some logs last night. I don't know what to make of this load I'm hitting now. Assuming its due to the maf cal being a bit off, and its relation to calculating load? What I don't understand is, I did the tips suggested by @Lex; for the 2.5 load issue, why is it still happening? I've haven't seen an issue like this until this past August, when I started seeing the random spikes in AFR's @ 2.5 load In reference to the MAF cal, I had been using Lex's MAF calibrations for bigger intakes for the HTP 3.5" so I'm curious as to the remarkable difference in the curve between the two. Historically I haven't had to make too many changes after loading the HTP scale from that thread, attaching a side by side to view the difference in in g/s in relation to the maf voltage. I'm probably just geeking out over this, but would like to find the source of my problem. I appreciate any insight on this from you pro's.

@ms6mil;
It may be true that the OL fuel trims that we can't see are saying FU and going to 0 above 2.50 ... or that an altogether different fueling strategy is being used. We can assume the MAF curve is off in the upper RPM and we are scaling it to the correct values or the ECU is using a different set of targets and (like in the old days of using a piggyback) we are scaling the MAF curve to get the desired AFRs although they are not the actual targets.

We certainly don't know what exactly is going on since we don't have the visibility of all tables or the logic.

The strange "s" shape to the MAF curve is not representative of actual flow. Actually the fact that we have to add 8-10% to the MAF only in the upper ranges and LTFTs stay close to 0 in the lower ranges are a stronger indication that different fuel targets are used in the upper RPMs. Further the actual power numbers and fuel consumption we are seeing are not lining up with the "inflated" MAF values. In light of this, instead of having the s shaped MAF curve, I would scale the entire thing by the 8-10% needed in the high load area. The higher you go up the MAF curve the more likely you are to be above 2.50 load so there is no reason to flatten it out in the top end.