My fuel trims consistently run positive, sometimes with the LTFT low running as high as 10%, so I decided to try scaling the injectors a little bit lower (from 513 to 504). After lowering the injector scaling, my trims are a bit better. I had read that the fuel trims will affect open loop fueling. That never made much sense to me, but I assumed it was true and expected that after my fuel trims had settled with the revised injector scaling, my AFRs and IDCs would be pretty much as they were before revising the scaling. That wasn't the case. AFRs are richer and IDCs are up across the entire rpm range during open loop runs. Seems like I'll have to revise my fuel maps if I want to stay at this revised injector scaling. Other people's experiences?

 

I knew for a fact that fuel trims had no effect on open loop fueling for DSMs. I have said the same for the Evo, since the ECUs seem nearly identicaly in many routines, but others have said otherwise. Since I haven't had much time to really sit down and tune my Evo a lot yet nor could I disassemble the routines, I couldn't argue either way. I think nj1266 had some sort of data to show that they did affect open loop...maybe we can ask him to post some data and see if his logic was sound.

I guess disassembling the routines is the only way to know for sure, but in the DSM ECU, and I would image the Evo ECU, fuel trims were only for closed loop fueling. They did not affect open loop fueling at all, which makes logical sense if you think about it.

If you rescaled your injectors, you would definitely have to adjust your open loop fuel maps to get back to the same AFR. Actually, that is a bit of a proof that fuel trims aren't applied during open loop fueling.


Eric

 

You changed the injector scaling but didn't change the actual fuel map to compensate?

If so, that is why the open loop AFR changed.

Or are you saying that you changed injector scaling, changed the fuel maps to compensate, and then after your fuel trims came down, then the open loop AFR changed?

If that is the case, it would do it after every flash as the Evo 9 fuel trims reset to zero after every flash.

 

That's correct. However, from what I read here, closed loop fuel trims affect open loop fueling, and within the range of what fuel trims can do (to maintain stoich at cruisse), any changes to the AFR due to changes in the injector scaling will be affected by subsequent changes in the fuel trims. I'm having a hard time explaining this. nj has a thread on this somewhere. He said that each time he would flash his car, WOT AFR logs several hours after the flash would be different than WOT AFR logs immediately after apply the flash. He determined that the fuel trims were changing and affecting open loop fueling. After tweaking his injector scaling and latencies so that the trims would stay pretty much on zero all the time, he said the AFRs stopped drifting.

 

I understand that mrfred, but if you changed injector scaling (not latency) and did not touch the fuel maps at all that is the main reason your open loop fueling changed.

Every time you change the injector scaling the ECU thinks they are a different size. So if you do not change the fuel map to compensate the AFR will change even with the same number in the fuel map.

Example:

Denso 720cc injectors
Scaling: 636

Fuel map is tuned to run 11.1 afr

Denso 720cc injectors
Scaling: 597

Fuel map UNTOUCHED from above

The car will run richer.

Now if you changed it to higher than 636 scaling it would run leaner

 

In your example, what if the injector scaling was changed from 636 to 630 (I know 630 is not possible, but say it was). With that small of a change, the question is whether or not the resulting change in the closed loop fuel trims would compensate for the injector size change and bring the open loop AFR back to the value at 636?

As l2r99gst says, I think its more reasonable that closed loop fuel trims have no effect on open loop fueling, but I haven't looked at the ROM code, and so I'm wondering if anyone has any examples of where closed loop fuel trims did affect open loop fueling.

BTW, I have a set of FIC 1050s on my desk next to me. I've now got everything I need to at least run straight E85. :-) I'd like to set it up this weekend, but no time. :-(

 

One of the points in fuel trims is to compensate for degrading fuel system performance. Example is concerning a stock car: Say after 60k miles, the fuel system is starting to wear down, injectors getting dirty, pump losing efficiency, filter getting clogged, etc. This would end up leaning the mix out. During WOT or open loop driving, fueling could be too lean and cause some issues. But by having a trimming system in place, the ECU can compensate for the degradation.

https://www.evolutionm.net/forums/sh...=282398&page=8

Naji's problem was improperly scaled injectors that cause his fuel trims to go +20 something, which in turn would cause differences in WOT runs from initial reflashing and after a few drive cycles. Trims DO effect open loop.

 

I remember reading those posts by Naji and I even had PMs with him back and forth a little and recommended what to do to fix his issue.

But, still, his results didn't make much sense to me. I would still love to see some other data or disassembly around this. His trims could have been drifting for other reasons, which would cause his AFR to fluctuate, but that fact that he says it didn't happen right after a flash and did a few days later consistently is pointing towards trims affecting open loop as well.

If this is the case, though, the example mrfred gave should hold true. You should be able to rescale your injectors to make your trims from +10 to -10 and your WOT AFRs shouldn't change, if these trims are also applied to open loop. In my experience that just isn't the case. Every time you rescale your injectors, your open loop AFR will change accordingly, and the trims won't get it back to the original value. Although, as I mentioned, 95% of my experience is with the DSM ROM as I have yet to sit down for a lengthy amount of time to tune my Evo.

I'm thinking that maybe something else was at play here, maybe the STFT or some other enrichment that nj1266 was experiencing. I would love to see some more logged data and evidence for this, though. Eventually, disassembly will prove it.

mixmaster, you are completely correct though in the fact that fuel trims are there to help with degrading fuel systems and changes in conditions, etc, but the fact that Mitsu runs their open loops so rick stock is an argument for fuel trims not affecting WOT.

Also, I wonder if there was anything changed from VIII's to IXs. Whatever it is, I hope we get it resolved sooner or later.


Eric

 

Just so you know the Evo doesnt have a fuel filter fyi

 

Yes, it does, but it is built into the pump and is a non serviceable filter. Thanks for playing, though.

 

a +/- 10% fuel trim is barely going to be seen while logging WOT runs. Its when trims are greater than 20% that it becomes noticeable.

If you want to test it, then change your MAF scaling table only in the lower areas to induce large idle and cruising fuel trim changes. After LTFT is set higher than 20%, log some WOT runs and note the difference.

 

I thing what MrFred is saying is this ...

If fuel trims affected open loop fueling and your WOT AFR was correct with the trims in place, then you wouldn't need to change the fuel map when the scaling/latency was changed. The trims would make up the difference just like in closed loop.

Is that basically it?

Since we do have to change the maps to make up for scaling changes, that basically means that the trims don't affect open loop.

 

Here's the thing, though. Everytime you flash the car, fuel trims are reset to 0. If your injector scaling is off, or MAF scaling is off, or you have a small vacuum leak, it won't show up right away. That is why it is important to ensure STFT is +/-3% at idle and cruise before you even begin tuning open loop tables.

Take this example: You install bigger injectors, but scale them too small. Because of this, the ECU is sending too long of an IPW. You disregard the STFT and start tuning it right away. You set your open loop tables to achieve an 11.5:1 and call it a day. Since the injectors were scaled too small to start with, STFT starts going negative to compensate for the rich condition during idle and cruise. In turn, LTFT begins to go negative since it is a constant correction in the STFT. This cause the ECU to shorten the overall IPW, including during open loop, which could cause your once perfect tune to now be too lean.

DO NOT count on fuel trims to compensate for poor injector sizing or MAF scaling.

 

mmm,

You mentioned no effect of 10% fuel trim on AFR. Are you using the EvoScan fuel trim formula, or the OBD-II fuel trim formula? Kinda curious because I would assume that with the proper scaling, a 2% increase in a fuel trim would equate to a 2% increase in fueling during closed loop, and then if fuel trims apply to open loop, then there would be a 2% increase in fueling there too.

 

I have my car configured to be in open loop all the time.

my fuel trims seem to be turned off, I have no indication in my logs that they are even alive.

I guess the test would be to configure your car to be in open loop all the time, have the afr where you want it, change the latencies, of course the fuel will be not right. But wait and see if the car tries to correct.

It may, I don't know, even in my configuration the car seems to flick back to another afr for a moment, maybe it is checking the status of the fuel trims