My car is just crazy, bad ECU?

bartun

bartun

Greenie N00B Member
Hello, I want to explain my actual situation with my Mazda 3 MPS Gen1 (European Mazdaspeed3).

Like 4 months ago, I saw my car smoking on cold (finally was just condensation), but because of that, I saw the afr idling was between 14.3 and 14.4 so I recorded a few logs, there we can see a 3rd gear WOT log and idling log:

#LOG 3RD GEAR WOT: https://datazap.me/u/bartun/idle-first-4-gears-normal-use-and-3rd-gear-wot?log=1&data=7
#LOG IDLING: https://datazap.me/u/bartun/idle-first-4-gears-normal-use-and-3rd-gear-wot?log=2&data=3-7

I was runing a Cobb intake, autotech hpfp, catless dp, catless tp, bigger tmic, one step colder spark plugs and a custom tune (Versatuner) that always work'd good.
In the logs, absolute load is low, desired afr is wrong, actual afr is lean, ignition timing is high, boost pressure is very low, and wastegate duty cycle is stuck at 60.3%, basically everything was working bad.

I installed the stock intake airbox, installed stock testpipe and loaded "High flow versatuner map", supposedly its like stock map for stock airbox, but I can keep my catless downpipe installed without getting error and fuel cuts. (THIS IS THE MAF CALIBRATION OF THAT "STOCK TUNE" https://i.ibb.co/1mLzPpS/maf-cal.jpg < Is this the MAF calibration for the stock airbox?)

After a nearly 40km's of all type of drivings, with the stock intake, stock testpipe and "stock map", I recorded log again.

#LOG 3RD GEAR WOT: https://datazap.me/u/bartun/log-1629046533?log=0&data=6
#LOG IDLING: https://datazap.me/u/bartun/log-1629046533?log=2&data=6

Wot values looks good to me I think, but LTFT is maxing at -8.6. If you see the idling log, now the big problem is the afr idling is oscillating so much, normally between 14.3 and 15.0.

After 2 or 3 days more of using the car, I made a few logs more (Now, LTFT is maxing at -12.6):
#LOG IDLING: https://datazap.me/u/bartun/turn-small-accel-idle-3k-constant?log=0&data=6-12

After 2 or 3 days more of using the car, I made a few logs more (Now, LTFT is maxing at -14.8):
#LOG CRUISING 2ND + 3RD GEAR: https://datazap.me/u/bartun/3-logs-cold-small-throttle-low-rpm?log=2&data=7-13

At this moment I have 2 problems, LTFT very negative, and very oscilating afr idling, between 14.3 and 15.0.
After the last log, I cleaned the first O2 sensor connectors + connectors coming from ECU and the ECU conectors + the wiring connectors to ECU with 0 residual cleaner, let them dry, started the car and magically, the afr idling was PERFECT, between 14.67/14.70/14.73

I used the car 3 or 4 days and I think It was solved, but after 4 days using it, CEL P2188 System rich at idle, that means LTFT were over -25%, and car started to give small jerks / pulls at low rpm, P2188 disappeared after 2 days of use.

1 or 2 days later, appeared new error code, High Voltage in second O2 Sensor, and idling afr was again stuck near 14.5

Yesterday, I installed new second O2 sensor, I reseted the ECU and fuel trims, and I tested it.

New Second O2 Sensor Denso installed yesterday.
New First O2 Sensor NTK installed 2000 km's ago.
New Denso MAF Sensor installed 2000 km's ago.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Is my ECU broken? Are my ECU wiring pins / connectors broken? Why my car is crazy? Should I burn my car and turn the page? I am fucking tired.

PD: since all this started, I feel like my brake pedal is very spongy and I have to use a lot of force to brake, It could be because vacuum leak but I read vacuum leak would make possitive fuel trims, not negative.
 
Last edited:
bartun said:
In the logs, absolute load is low, desired afr is wrong, actual afr is lean, ignition timing is high, boost pressure is very low, and wastegate duty cycle is stuck at 60.3%, basically everything was working bad.
Click to expand...

All of these could be linked, as Versatune is load tuning only.

First simple test: Start the car, pop the hood, gently lift up on the MAF wiring. Does the engine sound change? if so, you need to wire in a new MAF plug like I did or fix the pins/wiring in the existing one. Also, need to see MAFv instead of KR for idle, and you should just post the raw CSVs rather than using that site to do graphs. All the tuners around here just look at the raw data instead of the graphs, which do nothing for you but reduce how much pure information you can see at once.

That said, go back to the Cobb intake and your normal tune (who tuned you, by the way?), and the stock test pipe isn't really doing anything for you, but being on a catless downpipe with a stock like tune is *not* a good idea.

Focus on one issue at a time (make a checklist), pull new hot engine logs depending on what issue you want to troubleshoot first. It's crazy how much can change between a cold start log and a hot engine log, especially at idle.

That said, this should be your to-do list:
1. Create list of issues in order of perceived importance (we will update order based on symptoms later)
2. Put car back to how it was with the boltons and tune (make sure your VT software is updated on all systems you use it on, and then re-save the map before reflashing)
3. New concise logs, posted here via direct upload as .csv files (individual logs for idle, cruise, and WOT if it's safe to do a WOT pull based on symptoms)

Godspeed.
 
Enki said:
All of these could be linked, as Versatune is load tuning only.

First simple test: Start the car, pop the hood, gently lift up on the MAF wiring. Does the engine sound change? if so, you need to wire in a new MAF plug like I did or fix the pins/wiring in the existing one. Also, need to see MAFv instead of KR for idle, and you should just post the raw CSVs rather than using that site to do graphs. All the tuners around here just look at the raw data instead of the graphs, which do nothing for you but reduce how much pure information you can see at once.

That said, go back to the Cobb intake and your normal tune (who tuned you, by the way?), and the stock test pipe isn't really doing anything for you, but being on a catless downpipe with a stock like tune is *not* a good idea.

Focus on one issue at a time (make a checklist), pull new hot engine logs depending on what issue you want to troubleshoot first. It's crazy how much can change between a cold start log and a hot engine log, especially at idle.

That said, this should be your to-do list:
1. Create list of issues in order of perceived importance (we will update order based on symptoms later)
2. Put car back to how it was with the boltons and tune (make sure your VT software is updated on all systems you use it on, and then re-save the map before reflashing)
3. New concise logs, posted here via direct upload as .csv files (individual logs for idle, cruise, and WOT if it's safe to do a WOT pull based on symptoms)

Godspeed.
Click to expand...
Thanks for answering and for all the advice.

Last day, just before resetting the ecu/fuel trims, I tried to move the MAF wiring at the same time I was checking the AFR with Torque app. AFR was all the time 14.4-14.5 but in one moment when I moved the wiring, I thought I saw 15.5 or 15.8 AFR just only during 1 second, then 14.4-14.5 again, I tried more times to move the MAF wiring and this didn't repeat, It stayed at 14.4-14.5 all the time, so maybe is something related with the MAF wiring or MAF connector.

Okay, next logs, I'll only use KR for WOT logs, idle and cruising logs I will delete KR and will put MAF Voltage, and I will upload directly the .csv files, I won't upload them to datazap.

Next saturday (in 2 days) I will install again my Cobb Intake and and I will install my custom tune (It was made by a known Spanish guy who tuned few Mazda MPS here). I will drive it 40 or 50km and I will record logs as soon as I can and I will post here.

Thank you Enki.
 
Easter Bunny said:
How long was the tune on the car and did it ever run fine on the tune?
Click to expand...
I had that custom map since 2018 more or less, I had in the past some problems too with the AFR, or with small jerks/pulls at low RPM's, but in the end it ended up being solved after doing something (Tightening the bolt of the blow off valve, replacing the bolts on the maf sensor because they never they tighten, infinite spin, replacing the maf sensor, replacing first o2 sensor, etc.) either because I solved it, or because it fixed itself or because it only failed randomly.

You can see here a WOT log with that map when my car was working good (sorry for datazap again, but is where I have recorded all the logs, since now, I will save them in .csv)

3rd wot: https://datazap.me/u/bartun/wot-mapa-3?log=1&data=7
4th wot: https://datazap.me/u/bartun/wot-mapa-3?log=0&data=7
 
Last edited:
@Enki but what happens with catless downpipe and stock map? I think the only problem is the CEL and the possible fuel cuts when going WOT on 4rd, 5th or 6th gear, right? Nothing related about AFR/fuel trims.

Because my tuner told me he thinks I have a faulty MAF wirings or MAF connectors, he told me that he had to re-touch the tune few times because the afr or fuel trims were different sometimes between different maps.

So wouldn't be the best choice the stock map with the stock intake airbox? This way we make sure that the MAF calibration is perfect to the fitted intake (stock)
 
Overboosting, which risks a blown engine. *ALWAYS* run the proper tune for your mods.
 
Enki said:
Overboosting, which risks a blown engine. *ALWAYS* run the proper tune for your mods.
Click to expand...
I already ran catless dp with the stock map and fuel cuts prevents overboost, by the way, i would use the car 99,9% very soft use, not boosting it, only for test afr/fuel trims.. but as you want, I will install the custom map.
 
Are you running a different MAP sensor? The stock one doesn't read very high at all and a higher reading one is recommended. You may not be hitting fuel cut because of boost, it could be that the sensor itself is maxing out and the car is going full stop to prevent damage.

What Enki said is right. Don't be that guy. Dial your WGDC down so you're within the limits of your map sensor and what your mods are prepared for.
 
L337TurboZ said:
Are you running a different MAP sensor? The stock one doesn't read very high at all and a higher reading one is recommended. You may not be hitting fuel cut because of boost, it could be that the sensor itself is maxing out and the car is going full stop to prevent damage.

What Enki said is right. Don't be that guy. Dial your WGDC down so you're within the limits of your map sensor and what your mods are prepared for.
Click to expand...
Stock MAP sensor. I just want to install stock map like 30 minutes and very very soft use (lower than 2,5k rpm and lower than 25% accel pedal) to check if the desired afr idling is 14,70 or 14,60/14,50 and test how the car works with 100% stock intake airbox.



Enki said:
Utilizing fuel cut to prevent overboost is like wearing steel toed boots so you can fire bullets at your feet safely. Don't be that guy.
Click to expand...
I said I wouldn't hit boost with the stock map, so you can use steel toed boots if you are not pretending to fire bullets to your feets xD (lower than 2,5k rpm and lower than 25% accel pedal)

This saturday I am gonna make some tests, first of all I'll disconnect ECU and will connect it again. I remember 2 years ago, I had some AFR problems and hesitation/jerks/pulls at lows rpm, and If I remember good, I disconnected the ECU, reconnected it and problem solved during some time. Plus 1 month ago like I say in the 1st post, I cleaned the 1st O2 Sensor and ECU connectors and something changed, idling afr stopped dancing between 14,0 and 15,0 and stopped perfect in 14,70 for few days.
I will install 100% stock map, very very very soft use with no boost around 20-30 min, check afr and fuel trims, if desired afr idling is under 14.70 and fuel trims goes very negative, I will install the Cobb Intake and install the custom map and make the same checks.
I will move up and down the MAF wirings while I look the AFR at the same time, and I will check if some fluctuation appears.

If disconnecting and reconnecting the ECU doesn't make any change, My hunch is that the problem maybe are MAF connector or wirings, that was your first option, plus I saw how the AFR idling dropped from 14,6 to 15,5 or 15,8 for an instant moving the afr wiring, plus my tuner said me today he thinks I have some problem with the MAF reads, because the connector or wirings while he was tunning my car, so I have to unassemble all the wiring from maf to ecu to check.
 
Last edited:
Yesterday first I did is closing a bit the pins of the MAF wirings, so they grab better and make better contact to the MAF pins.

I upload 4 idle logs (all logs were recorded after 53km driven from last fuel trims reset), I was touching the MAF plug and the MAF wirings in all the logs, but no changes, so MAF connector and MAF wirings are working good, or are working permanently bad all the time.

I resetted the fuel trims last saturday, and after 53km (yesterday) I made these logs, LTFT looks good after 53km.

After driving to my home, were totally 73km driven after fuel trims reset and LTFT still around -5%, so maybe the new Second O2 sensor that I installed last week, solved something.

The only weird thing is the "desired afr", when recorded the logs, desired afr idling was between 14.61 and 14.68 (14.68 permanent should be perfect) but after driving home, it was 14.59, and everytime I drive the car, it lows a bit more, so maybe after 200km's the desired idle AFR is 14.30 and obviously, the afr idling would be 14.30.

Maybe the MAF calibration of this "nearly stock map" it's not 100% perfect, or maybe should I try to relearn the throttle body calibration.

I have no idea why the "Desired AFR" changes and which part/component of the car makes that.

Opinions?
 

Attachments

Last edited:
First off, desired AFR is what the ecu wants, so don't bother watching it. Secondly, a slight change in the desired AFR could be the result of any number of tests, including testing the narrowband O2 sensor calibration. This can be tested by disconnecting the secondary O2 and seeing if the AFR still dips.

Basically, you're looking for problems in every bit of information logged, when the car has ways of telling you something is wrong outside of the logs.
 
Enki said:
First off, desired AFR is what the ecu wants, so don't bother watching it. Secondly, a slight change in the desired AFR could be the result of any number of tests, including testing the narrowband O2 sensor calibration. This can be tested by disconnecting the secondary O2 and seeing if the AFR still dips.

Basically, you're looking for problems in every bit of information logged, when the car has ways of telling you something is wrong outside of the logs.
Click to expand...
I know, you're right, I'm obsessed, but when my car was running fine, all the values/parameters were fine, that's why I compare it so much, and because I have pending to install a BNR S4, FMIC, 3 port EBC, 3.5inch intake, and before that, I want to be sure the car is 100% perfect.

I am just trying to gather some information and hopefully someone appears who knows what's happening in my car.

Today I used the car again, I finally made 100km's since the fuel trims and I recorded a log **My bad I didn't record KR** (all the log is in 3rd gear, slowly going to 3k RPM, WOT and then driving through city still in 3rd gear all the time)

AFR in WOT is lean, I made one month ago a 3rd gear WOT log with the same map and the AFR was perfect (https://datazap.me/u/bartun/log-1629046533?log=0&data=2-6)
So 1 month ago, the WOT AFR and the Desired AFR were perfect, but LTFT was increasing negative to -20/-25, today LTFT stays under -6 but desired AFR idling going rich and WOT AFR is lean.

I disconnected and reconnected the ECU connectors after recording the log, and Desired AFR idling rised up to 14.69 (from 14.59 it was today) then slowly raised up to 14.78/14.81.

After driving 25km to home, desired AFR idling raised down again to 14.63 (afr to 14.63 too) so apparently reconnecting the ECU didn't make nothing, but it's strange just reconnecting the ECU the desired AFR changed.

PD: the engine oil level seems to have risen up, and I feel that it smells a bit to gasoline
 

Attachments

It will never be 100% perfect. Ever. There are so many moving parts, each with their own tolerances, not to mention day to day air / fuel quality and other variables.

As for the WOT AFR, you've said your tuner has adjusted it, right? Well if you fixed the MAF issue by working on the pins, then it stands to reason your calibration is now off.

STOP LOGGING DESIRED AFR. It doesn't mean anything to anyone.
 
Enki said:
It will never be 100% perfect. Ever. There are so many moving parts, each with their own tolerances, not to mention day to day air / fuel quality and other variables.

As for the WOT AFR, you've said your tuner has adjusted it, right? Well if you fixed the MAF issue by working on the pins, then it stands to reason your calibration is now off.

STOP LOGGING DESIRED AFR. It doesn't mean anything to anyone.
Click to expand...
I am not running my custom map, I am running an OTS Versatuner map (it says its stock map but with a bit more boost until redline, so MAF CAL is for stock airbox) so nothing changed in the MAF CAL of this datazap 3rd WOT log, same map, same MAF CAL (https://datazap.me/u/bartun/log-1629046533?log=0&data=2-6)

MAF pins didn't make a change, Most probably LTFT stopped to going to -20/-25 because new second O2 Sensor installed (Due P0138 O2 Sensor Circuit High Voltage)
 
Enki said:
Unless that secondary O2 is OEM, I wouldn't trust it. Unplug that fucker and try again.
Click to expand...
It's a Denso one, I'll try. Unplug it and let it unplug? and what should change? AFR in WOT should be good again? :s
 
You must log in or register to reply here.
Back
Top