Is LTFT a value or a table?

[tomm.brz]

Originally Posted by Kodename47

I'm not going to keep arguing with you. But yet again, you have just contradicted yourself.

The delay settings shouldn't matter, the fact they do suggests that the LTFT is still active as it's allowing the car to hold STFT into higher g/s. I know it says that those should stop the LTFT from activating but in my experience they don't. I could probably dig out a load of logs as well, seeing as those have been set to 0 on my car since my early tunes, but I don't fancy wading through thousands of files to try and find some examples.

The only sure fire way is to set the closed loop fueling targets to 14.7, but because I like to use CL to enrich the fueling under load I use the fuel trim bands to ensure that there is no trims set for WOT OL operation and so far, so good.

i haven t thought till now we were arguing, sorry I was invadent or something like that
still I havent contradicted myself because i have logs too to show my point of view but since you arent interested we ll just go our own way
i meant that if you have 14.7 untill 0.7 on CL tables and something richer after that, LT min max zeroed out, and delay tables stock, you will see some ltft after 0.7 load then when the delays time out, ltft will be 0 again

[Kodename47]

Originally Posted by Tor

Just as another curiosity with LTFT. Since changing injector ratios to a uniform value above 2400 rpm, I only see the 20 g/s LTFT applied during OL. It's like it disregards the 60 g/s completely.

I have logs from cruising in CL, 5th gear, low rpm (exactly 20 g/s) and I see the LTFT changing. Later in the logs WOT and it has the exact trim throughout the full rev range that was set at 20 g/s.

I have multiple examples of this (including multiple new LTFTs set during CL and applied during WOT above 60 g/s in the same logs), so it makes me wonder if the descriptions of the of the brackets are correct. Maybe injector ratios play a part as well.

With the stock injector ratios, there is the classic split at 5000 rpm when port injection becomes active again. I've never seen this split with uniform injector ratios.

There are 2 A/F Learning Ranges that I can find in the ROMs, maybe there is some some of DI only/Combined type of setup or perhaps one LTFT defined for one of the injector setups? I have always had some form of suspicion that this happens but not enough time/bothered to prove it.

Originally Posted by tomm.brz

i meant that if you have 14.7 untill 0.7 on CL tables and something richer after that, LT min max zeroed out, and delay tables stock, you will see some ltft after 0.7 load then when the delays time out, ltft will be 0 again

EcuTek define those settings as:
The maximum/minimum fuel trim allowed to be applied by the Long Term Fuel Trim, set this map to zero to avoid long term fuel trim working
There is no reference to open loop or closed loop. If you change just these settings to zero on a stock car with no other changes, test what happens. If I have some time I'll go and revert on my car and report back.

[tomm.brz]

what happens when you zero out those 2 tables on a stock (actually also non stock) rom is that ltft stops working during open loop. in closed loop it still collects data from stft and after a while ltft works on closed loop anyway, just not applied to open loop. I tried many times that ( i m over 400 reflash on my my17 brz), if you different experience from that i ll be happy to hear!

[Kodename47]

Originally Posted by tomm.brz

what happens when you zero out those 2 tables on a stock (actually also non stock) rom is that ltft stops working during open loop.

The reason why I stipulated a stock ROM is that by changing the most common settings, like OL delays and both main and CL fuel maps you can influence when it drops into open loop and therefore has the knock on effect of preventing the upper LTFT from ever being reached. There is a chance that I took the fact I saw LTFT at all as the reason that I didn't think these parameters worked as I can't see a reason that the ECU would bother to differentiate between OL and CL LTFT.

[Tor]

Relevant to this topic, this is a freeze frame readout from Techstream. This is how Techstream appears to lists fuel trims:



From observations with how the trims behave with stock injector ratios vs. uniform injector ratios, as previously mentioned, I would suggest "dual" is referring to injection mode.

As such, I'm not sure the current understanding and the way the ranges are defined are correct. Maybe there are missing tables in our definitions?

[Kodename47]

Originally Posted by Tor

Relevant to this topic, this is a freeze frame readout from Techstream. This is how Techstream appears to lists fuel trims:

EcuTek also has these. If you log these it will show you all the stored LTFT. It also has A/F Learn Value Bank #1 (PFI).

[tomm.brz]

i noticed too ecutek has those value loggable, and they vary also with ltft max/min at 0, but it seemed to me that with ltft max/Min zeroed, those values were not applied to the open loop

[Kodename47]

Originally Posted by tomm.brz

i noticed too ecutek has those value loggable, and they vary also with ltft max/min at 0, but it seemed to me that with ltft max/Min zeroed, those values were not applied to the open loop

Only the high bands are likely to be applicable in OL.

[tomm.brz]

i used to log High dual and those were not applied in OL with ltft min/max zeroed even if those band were different to the usual 0.2 value

[steve99]

Other thing that appears odd is. There is only about 5 air flow bands but when you drive along through the rpm range the ltft seems to step up and down many more times than 5.


We o ly know waht romraider or ecutek guys have discovered theors hundreds or more tables and functions that are unknown or o ly partially understood

[Tor]

Originally Posted by steve99

Other thing that appears odd is. There is only about 5 air flow bands but when you drive along through the rpm range the ltft seems to step up and down many more times than 5.


We o ly know waht romraider or ecutek guys have discovered theors hundreds or more tables and functions that are unknown or o ly partially understood

I suppose you could be seeing the LTFT adjusting itself and making jumps. And it would appear that there are more ranges than there actually is.

Or I also suspect that some of the tables are just defined with the knowledge from previous Subaru engines and may as such not be complete due to the D4S. E.g. the MAF ranges for LTFT may be an example of that. I could well imagine that they should be called dual/single injection instead of A/B.

Edit:
Also, I'm not sure the min/max ranges are correct, the tables say there should be 6 ranges (A to F) where Techsteam says there are only 4 (idle, low, mid, high) - maybe there is no A and F? My ECU for sure set WOT LTFT at 20 g/s - not 60 as was previously commonly assumed.


It's, of course, all speculation and I can't see that it matters much anyway, as there should be little reason to adjust the ranges.