AFR difference between the banks
 

I find my AFR is always 0.5 richer on bank1, is this common?

this happens in open loop only, they are pretty close to each other in closed loop.

also found that MAF reading is slightly higher on bank 1 than bank 2 under open loop.

anyone with some ideas? :bowrofl:

My tuner found a discrepancy as well. Apparently Uprev can only adjust fuel globally, so bank-to-bank variation means you have to une to the leanest bank. I'd assume there's a certain amount of variance that can't be avoided.

thanks for the info :tup:

Ecutek apparently can adjust fueling by bank. Mine wasn't a half point off, so it didn't make a huge difference in my tune. A half point is a pretty big difference.

Uprev can adjust fueling on each individual cylinder... just saying ;)

I really depends on how far the tuner is willing to go to adjust/set your AFR targets.

Mine is about .5 off toward bank 2. Bank one being the richer bank. Omar has been playing with my car for a bit since it has been at UPREV for the past few months. I will report in once I pick her up.

Sent from my SGH-T889 using Tapatalk

Interesting.

top right :)

http://www.the370z.com/members/megan...ure55763-1.jpg

Shouldnt the engine have long and short term trims that should handle these inconsistencies?

you mean he ECU....

short answer yes,
but not all sensors are the same. There is tolerance that are acceptable in everything, OEM included.

I had that type of variance .5 to 1.0afr when I did my custom CAI and I fixed it with tuning.

yes I did mean ECU sorry. Right but my concern would be once you start manually messing with the duty cycle of each cylinder how do you know that say you arent running lean on 2 cylinders and rich on the other one? I mean I guess you dont really know with the trim either, but atleast it is not a static value that you input and the ECU runs with as I would imagine setting the trim manually would be.

I am fairly new to the uprev side of things but I have been tuning engines for 4+ years with EFI Live and that would be one of the concerns I would have with manually setting trim values.

I agree that there is tolerances, its why you will only get rich and lean codes after the engine has exhausted its available trim values, but to me this is exactly what the trims should be doing, making sure each bank is running as closely as possible to the target AFR (whatever it is you are calling for).

Just my thoughts...

just to give a sample example on this issue with Bank 1 & 2 not being equal.

adjust the whole bank that is leaner to the other side.
on the table it would be like
1:100%
2:105%
3:100%
4:105%
5:100%
6:105%


as for tuning for each cylinder.
I cant help you on this. I have a theory but I wont go on that subject.

Just did this cylinder adjustment to 104.69, A/F ratio is exactly the same, before it was about .50 different, also i can hear the engine running more smooth rather than like a tractor.

Yep. There's always potential for some variance between the banks -- not sure if .5 would be considered unusually high deviance tho'.

Can you monitor fuel trims for each bank -- how close are they?

Also, one MAF sensor may need to be cleaned or you could have one extra dirty filter.

Yeah i thought it was the filters as well so i looked at them they looked barely dirty and i have found a few tree leaves, cleaning that barely did anything. There was notice, but u can still hear that something was off. The A/F logging showed it was off by~-/+.50 from other bank. Found this post and adjusted it, now its right on the money they r both the same ty very much, if it wasn't for this thread and ppl's shearing i don't know when i would have figure it out.:tiphat:

I also used cylinder trim and got both banks to similar A/F now.

actually I reduced whole lot by around 3% and that put the car to match up very closely to the target A/F, without touching the fuel correction table. :D

I spent a good part of 5 hours logging my car today and noticed a few things. Bank 1 would consistently run 0.5-1AFR richer than bank 2. The kicker is the ECU was commanding bank one to run at ~115% which was clearly causing the rich issue. Bank 2 trims would sit at about 97-104%.

Those trim values posted here are only good for open loop conditions. Under closed looped the ECU is going to do everything it can to make each bank meet the target afr.

That being said the ECU is for whatever reason commanding one bank to run richer than the target. I have done some research on this tonight and it looks like with a CAT delete guys have been running into issues where if they reset the trims its dam near perfect, however after a bit one of the banks will richen up on its own as the trims come back into play. Some of the 350Z guys have just unplugged the post cat 02s as it seems like the ECU does listen somewhat to these cats as well.

I have tried everything to try and dial these values in with all 4 cats connected to the car. No luck, eventually in closed loop the car will deviate. While on the road for awhile one bank afr will read 14.5-14.9 just bouncing around slightly, the other bank afr reads 13.6-14.0. My wideband which is post turbo after the exhaust gasses mix sits at about 14.1-14.3.

Has anyone had this issue? I guess the next thing im going to try is unplugging the 2 post cat 02s. I just hope this doesnt put the computer into a permanent open loop state.

Any suggestions would be appreciated.

Thanks

Mitch

Afr
 

I actually have that same problem thou it is on a G37. I found out by a Uprev e-tune. Randy said to look for a air leak that was the size of a pin hole...

As for the two problems I noticed the first was at idle. I need to idle around 95-105% on both banks. Bank one idles correct, but bank two will idle between 90-115%.

The second problem is when I am WOT I have deavations of 1.0 AFR on bank 2 from bank one. If bank 1 was 11.00 bank 2 would be 11.80-12.00.

The only thing I found that temp "fixed" it was tighten up the exhaust, but when it came back I had the same problem on both banks now....

I believe the problem is in the exhaust somewhere. If it is in the seals, blown out hfc or broken shorty's I am not sure. My next plan of attack is to go to LTH and get rid of the seal between the shorty's and HFC. While doing that I will look at the 02 sensors and other connections.

Well I hope I could help and subbed for info. GL

HC_416,

I dont think any exhaust leak would cause this. The ECU knows its running rich but yet its still trimming like its lean. Even if you had a big leak the ECU should do whatever it can to meet the target AFR, not push itself away from it.

Do you have a cat delete? are both downstream O2s roughly in the location they would be stock? I have a BP kit and both of my downstream O2s are actually in one bank. I think this might be causing my issue, however if I can find someone else who has both his 02s approximately in their stock location with this issue then that would rule this out.

The way I see it aside from the possible secondary cats causing the issue there is nothing else on the car that could possibly contribute to this. A pin hole leak or even a big leak would make the ECU work overtime to meet the target AFR, not push itself away from the target. If that makes sense.

Well I hope you at least get it figured out. I have had this problem since January and have been nit picking at it cause of time and life. More than likely wont get it fixed till the winter.... Also I think I should read better, but since you were seeing the kind of of the same problem I figured I would put in my 2 cents just to try to help out. GL man hope you find it. I know I made 309 whp on a mustang dyno and that was with no parts and this air leak. Once I figure it out I'm hoping for 330-340 WHP. I did a lot of improvements.

I don't think it's ever in open loop. The target maps cover all load ranges, and that means there is always a STFT and LTFT. It's just that beyond a certain load range, you can set the targets to whatever you want without the ECU trying to wrestle the engine to run close to stoich (assuming it does that -- I seem to recall uprev commenting on this, but no idea if its accurate).

The cylinder trims vary most likely due to flow variation -- massive correction implies either the sensors are off more than one would like or your set-up is just flowing unevenly, which to some extent is unavoidable.

The car is in open loop under load. You can watch the trims sit at 100% while the car does its thing (which for people trying to follow a 100% means the ECU is not making any adjustments to the correction tables). Then once the load drops down the trim starts again. Uprev however doesn't have the tables mapped for when the ECM switches from open to closed.

It does make sense that there will be a hint of variation per bank. However these trim values everyone are talking about are only active during open loop power enrichment mode. When you are in closed loop the ECU will over ride whatever "trims" you set in when it starts listening to the 02.

For example, lets say you take bank 1 and cut the fuel down by 25%. So on 1,3, and 5 you set the trim to 75%. The fuel relearns are then cleared and the car is set to sit at a constant load under closed loop. At first you will see the trims on bank one sit at 25% as it tries to take your changes into account and yet keep the 02 on that bank happy. Now as you drive you will see the LTFT (long term fuel trim) start to react to the constant high STFT and it will increase to help bring the STFTs back into a better range.

Once the LTFT stabilize you will then see car will run at or around whatever the AFR is set at now completely disregarding your bank trims. Now under WOT this is not the case at all.

I agree that these trims are great under WOT runs, you can clearly see that if one trim bank is constantly higher than the other bank you can then see and adjust these bank trims to compensate for that and ensure that when you do go WOT both banks are running very close to each other in terms of AFR.


If I am wrong I would like to know it. I have been tuning vehicles for 4 to 5 years now and this is the same logic used on the cars I tune. This UPrev setup might be different, but in my 6 or so hours of log time on the weekend this seems to be exactly what is happening in this case as well.

Mitch

That all sound spot on -- I wasn't entirely clear on whether there was a real open or closed loop like there was on narrow band O2 controlled AFR's, because, yeah, as you noted, uprev doesn't clearly indicate it. I thought since it had wide band (5 volt?) sensors it just maintained a close loop throughout, maybe forcing the ECU to trim to stoich under low load.

So is there an aggregated LTFT from the closed loop sections of the map that carries into open loop?

Also, what approximate TPS or MAF voltage switches it over? I'm guessing it corresponds to whatever point in the main AFR target map it goes richer than 14.7?

Is that much variance common on an OEM set up or is this strictly due to the plumbing differences for your set-up?

The only other thing I can think of is that one of sensors may just be reading very differently. Have you swapped the sensors to see if the bank difference moves with it?

Yes I have found that manually setting the correction tables to some very wrong values (lets say removing 30 percent to the constant throttle section) caused the STFTs to constantly sit high until the LTFTs started to take that into account and change to meet the needs of the STFTs. Needless to say you can be way way out to lunch on your correction tables and the ECU will bring it back into check, it will just take longer to do so. And I guess if you bottom out your LFTs then you will get the rich or lean bank codes from there.

It seems like anything under 14afr commanded will put the car into a closed loop situation although its really hard to pinpoint weather its load based or just afr based. With the BP kit as soon as the boost starts to come on its already in open loop and then that is where you can see how your correction tables really come into play because now the ECU is not looking to adjust these values and it just takes it as "correct".

I have looked on many other forums (including 350z and titan forums) and it seems like its a fairly common issue. I am surprised infact that more people have not noticed it. On the forums I did find people were describing the exact same situation however most of the replies were uneducated and telling the OP to check for boost leaks ect. No boost leak or exhaust leak is going to tell the ECU to inject more fuel if it is already running rich.

As far as switching the sensors on the bank, that wont fix anything because the ECU clearly sees bank 1 is running rich. I can see it through uprev. Its sitting at around 13.8-14.2, yet the ECU is trimming like it is running lean.

I have come across a post where some guys have said disconnecting their downstream 02s have helped to fix the problem. The thought is that the ECU is looking to see a certain AFR post cat, since the cats are gone its getting fooled and I would suspect that the ECU may be injecting more fuel to try and light off the cat or heat it up to bring #2 O2 into whatever reading the ECU would expect to see on it. Unfortunately we cannot see exactly what these 02s are doing. I am going to try unplugging both O2s next time I am under the car and remove the associated codes and see what that does to the tune. Some have said it will force the ECU into a open loop situation under all conditions. One way to find out.

Switching to open loop is probably both TPS and load based -- the boost just reads as high load (i.e., higher voltage on the MAFs), so the only concern would be if you skew lean under part load with moderate boost. If it's not happening, then I guess its either just load based or weights load as more important than TPS?

As to the 2ndary O2's, I thought if you turned off the DTC for it that effectively got around the ECU taking notice of their presence and attempting to correct -- meaning, you first have to work your way through the the trip detection logic algorithm and then the ECU starts correcting, so if the detection logic is never invoked, the problem is circumnavigated.

If not, and if there's no way to just shut them off or change their threshold, then I bet a lot of us are driving around with weird random fuel trims... :icon14:

As to switching the MAF sensors -- are you saying voltage readings are the same? If not, it could still be referencing different cells in the fueling map, or if it just aggregates them when accessing the tables, one might be skewing things off more than it should.

On that note, if the MAFs are fine, what about the primary O2's?

Good luck with the 2ndary O2 fix!

Jordo,

Your probably correct, I do know that it is very touchy as to when it switches from open to closed loop and just a small amount of load will trigger open loop. If anything as soon as my car goes open loop it runs very rich (which was the main point of me starting to look at my own tune, I wanted to cut back on the fuel a bit as it does at time drop below 10:1).

With my experience in dealing with this automotive ECUs just turning off the codes associated with a sensor will not stop the ECU from reading and using this sensor. If it is there it is going to use the data. I have actually ran into issues where just simply turning the diagnostics off and assuming it was dead caused issues. You need to physically unplug or cut power to it to stop the ECU from using it. I think my particular issue is that both my downstream 02s are side by side on one bank which could be causing this whole mess. Ill know more once I find time to crawl up under the car and unplug them.

There is no way that a MAF is causing the ECU to do this. If the MAF was reading high or low you would see it in the trims and the corresponding AFRs. Same with the primary 02. Nothing like that explains why the ECU is seeing rich, yet adding fuel.

I wonder if any of the Uprev tuners have ever seen this issue? I have seen it across multiple message boards all the time no one with any real affiliation to Uprev has commented on any of the threads. I might try logging a few other 370Zs and see if its just me or if everyone is having similar issues.

Mitco,

If you need a base to read and its free, let me know, will volunteer :)

I will take you up on that when I have time. Do you still have the cats and everything on your car?

Bone stock except have K&N typhoons CAI. I can convert to stock at any time

Hmm. So that doesn't affect the detection logic, it just turns off the CEL... that's probably randomly screwing with things for a lot of folks :shakes head:

Having the two O2's on one bank sounds like it very well might be the problem! Actually, if they are located on the same downpipe, one after the other rather than on separate but parallel paths, they are probably reading latency in voltage change from the primaries quite differently, and that could result in wonky corrections being applied.

One more thought if it isn't the O2's: Are you using bigger injectors? The k multiplier (I can't recall if that scales for size, injector latency or both) might need to be tweaked if it was left at stock values.

Jordo, thats how it is in a bunch of other ECUs I've played with. I would expect this one to be no different. I have not tested out that fact but I would be very surprised if this one did something other than the Bosche/Delphi ECUs I have played with.

The double O2's could most definitely be the issue and right now thats the next thing I am going to do (been a bit busy as of late so I haven't played with the car at all as of yet.

I do have larger injectors and the K multiplier was changed accordingly. Even if that wasn't dialed in properly you would just see that in the correction factors as well as at WOT you would go lean or rich depending on how far out your K factor is. Essentially any of those changes will absolutely affect the AFRs but under closed loop the computer will do its best to account for any errors in the tune itself and if your to far out you will get lean or rich bank codes.

Tomorrow I am installing CJMs fuel return system which will warrant a full redo of the fuel tables so I will be playing with UpRev alot then. As I type this I am soldering together my own device for listening to the knock sensor with headphones that they have posted on their website. :tiphat:

:tiphat:

To anyone following this thread unplugging the O2s fixed the trim issue. The car is now doing a much better job at meeting its targets. Haven't had extensive time on it yet, but it does seem to be much better.

That makes me sad. How am I supposed to pass emissions testing if I have to disable my O2 sensors to get the car to not flip out?

Serendepity solved this one for me.
 

Not sure if this issue is still circulating about the forum. I was having the same issue. Bank 1 and bank 2 were always off by 0.5-0.7. An accidental wipe of my rom file lead me to this series of adjustments which virtually eliminated the disparity.

1) I downloaded a different rom from uprev by accident. I flashed the 2013 370 NISMO file 1A40D instead of the one my ecu was supposed to match:1A39D.

2) I began running all uprev programs as administrator.

3) I edged up my injector latency value at 14V a bit (added 0.3ms).

That's it. Like the majority of useful medications on the market, this was a serendipitous discovery. I hope this helps at least one other person.

Peace.
-JMC

unplugged the secondary or the primary -- I'm guessing the former?

Your point 3 might be the reason you're not seeing this happen anymore. Are both your secondary O2s on the same bank?

Secondary, they been unplugged all summer with 0 issues.

This is on my todo list for the V1.75 upgrade. Bought the plugs to be able to pull them completely!

yeah.... i been lazy... :rofl2:

Reading this thread I am looking also in my engine a little differences between banks: 0.5-0.8 AFR, but only in 1800-3600RPM Full Load area!

In this case, it's better to tune all richer maintaining all Bank1 & Bank2 MAX AFR below 12.7, or I must consider a mid value?

...after 4000RPM everything the AFR was balanced...

Some positive feedbacks removing the O2s?

Thanks.

The two banks are pretty close, but in the 3250RPM area, there is a difference :confused::confused:

Two examples:


http://www.psm1999.net/images/share/1diff.png

http://www.psm1999.net/images/share/2diff.png