Fuel trim nosediving during hwy cruise?? :confused:
 

Ok, so I've been datalogging the hell out of my car and have run across something interesting/confusing today...

If I putz around town my fuel trims stay more or less in line. The only one that really changes according to Evoscan is the "FuelTrim_Middle", which based on what little I know is as it should be. The "FuelTrim_Low" never changes and neither does the "FuelTrim_High" ... no idea why.

I finally had to take the car out on the interstate today and after looking at the logs noticed that my "FuelTrim_Middle" had steadily deteriorated until it sat at a value of "50"!! The FuelTrim_High and Low never even budged.

Is this a bug? Shouldn't the FuelTrim_High change at some point in time?

The only change I made today is to bump up the "Injector Battery Voltage Latency Compensation" value. Before my first interstate run I had bumped the 14.06V value to .480 and the 11.72V value to .720. After seeing the first run I changed the 14.06V value back to the stock setting (.432) and I changed the 11.72V value to .696 (smallest possible change from stock). The reasoning behind the voltage change was that I was logging the "FuelTrim_Low" at ~110 (+10%) and so the thought was that if I bumped the latency compensation, it would lower the "FuelTrim_Low" closer to 100. However neither voltage setting affected the "FuelTrim_Low" in any way.

Even with the second settings the car did pretty much the same thing. It was fine around town, but then after a highway run the "FuelTrim_Middle" shot back down to 75.

This really has me bugged, since I wanted to rescale my injectors this weekend using the injector scaling feature...

Thoughts? Comments?

l8r)

PS: Please note that I do not have EGR on my car (dunno if that matters)

From reading the manny posts about scalling it seems the latency is mainly used to adjust idle trims and the scale factor is used to adjust the cruize trims, then the map would be adjusted for the desired openloop afr's.

Did you recnetly change your injectors?

I've got a lot of fueling work to do, i have my openloop afr's good, bu my trims are wacked out and the car runs way leaner than it should on the stock map. I have a set of 680's waiting to go in, but i have to get the base problem sorted out first, waiting on a fuel pressure gauge and a walbro pickup screen / filter, my stock one was nasty when i put in the walbro a few weeks ago.

Fuel trims are affected by three (technically four) things..

1) Injector Scale
2) Voltage Latency
3) Mapping value...

The injector scale should be used to get your fuel trims close, injector latency for 11 and 14v (At idle the car should show 12.5-12.9v, some are higher and some are lower) range should be used to adjust the fine tuning of the injector (If you find it swings back and forth between rich and lean very wildly with 1% or so injector duty, you would raise that number..
Finally, you would use the mapping table to alter the areas where the other two items don't trim correctly...

Finally, If if you cant get it sorted out, you would look at the different MAF calibration values and examine your MAF curve very closely, you'd likely have to compare it to a stock curve, but you would make airflow adjustments in the maf smoothing and maf scaling tables as a last resort.. Usually bigger turbos result in a slightly lower MAF reading at cruise speeds...

mad_VIII, I've run 780cc injectors for a year now ... scaled the "old fashioned way". This was the first time that I've been able to log fuel trims for an extended interstate run. Makes me wonder if not all cars that are rescaled this way exhibit the same symptoms when driven at 70+ mph for extended periods of time.

Has anyone observed the FuelTrim_Low and FuelTrim_High actually changing? The only time my FuelTrim_Low changes is when I reset the ECU. Fuel_Trim_High always sits at 100%. :dunno:

After the second highway run yesterday (FuelTrim_Middle down at 50%), upon restart my car idled like a$$ - targeting an AFR of 16-17. I reset the ECU and everything was back to normal.

One more note ... just to make sure that I have this right: If the fuel trim is positive (e.g. over 100), then the car is running lean and the ECU is adding fuel, correct?

l8r)

I think that is correct, over 100 equals adding fuel due to lean.

My trims are wacked, high is always 100 like you observed so I don't think it is even used on our cars. My low is stuck at like 140 and mid at 123 or so.

Not bashing Al or anything, but I tend to think the only way to tune is custome tune the individual car, my eflash was way off due to my fuel delivery problems, I included my utec logs with wbsfr's and boost logging, and the eflash was still way off.

It seems a lot of folks with reflashes (no particular brand or tuner) have never logged their basic parameters like timing and knock and when they do they find they are knocking and getting timing pulled.........

The purpose of the E FLASH! was to provide a basic base calibration which is well thought out and which in most cases will run the car very well - (better than the stock map)

I think the E FLASH! is mostly directed to the non tuner customer who is using the tactrix cable as a way to get a product similar to a mail in flash without mailing in the ecu

In many of those cases the customer lacks the tools or interest to tackle the project of tuning him / her self and is using the E FLASH! map as a final product

Experience over the last 3 years has shown the through the mail or "base" tune flashing has been very reliable, safe and effective in practice.

As far as your situation - Utec logs and data are really irrevant to E FLASH! mapping

On cars with modified injectors particulatrly and modified intakes also to a lesser extent it is NORMAL for the ecu to tune to the desired a/f target in closed loop operation by adding and subtracting fuel within its authority range. Of course you would like the adjustment to be as close to zero as possible, however the vehicle will operate normally until and unless the authority ranges are exceeded which triggers a CEL.

IT IS EXEPCTED AND ENCOURAGED for those users with E FLASH! to data log and monitor their vehicles under a wide range of operating conditions and to make any individualized minor adjustments which may be indicated to fine tune the map.

E FLASH! is an inexpensive base flash, set up to match the basic configuration of the end user's vehicle modifications, fuel used and geograhpic location. If there are any areas of the fuel trims which need adjustment the end user should simply adjust them.

:mitsu:

PS - although I love Evo Scan in total, I do not love the fuel trim logging feature

I prefer to use my MATCO determinator OBDII scan tool which is much more effective for adjusting fuel trims and gives a great graphing opions (of course its not a cheap tool) Other quality OBD II scan tools may give you more useable daat than the evo scan for fuel trims

On other values such as the knock sum evo scan is amazing

I think its just a matter of using the right tool for the right job

You really have to log the battery voltage and see how the change in voltage in a driving cycle is effecting fueling - IF it is

For example if when battery voltage drops when your a/c fan comes on and your car goes lean - for example - that is what this parameter is used for

It is not a global fuel trim

The general curve in this table is made to correct for physical response time of injectors to open which can vary as battery voltage varies

Many injector manufacturers supply the data for their injectors which can give you a decent map for your specific injector

Here is some data from Motec on this parameter

Injector Dead Time - Injector dead time refers to the latency of the injector in producing maximum flow rate. All injectors require a certain amount of time to open completely and produce maximum flow. The amount of time is dependant on several variables including; fuel pressure, battery voltage and physical characteristics of the injectors themselves. Typically higher fuel pressure or lower battery voltage tends to increase the dead time. This leads to a reduction in fuel flow in to the engine and as a result influences the engine's state of tune. Luckily MoTeC allows the user to define an injector Dead Time table if the values are known, or use a standard compensation which is known for a number of injectors. The ECU automatically adjusts the values as the battery voltage changes to ensure that the fuel curve remains constant. If a fuel pressure input is used on the system, MoTeC can compensate for variations in fuel pressure to achieve a consistent fueling even with varying fuel pressures.

Well again, no offense intended Al, but I don't see how logs that show the pre flash running condition with wide band afr's and boost and ecu timing are irrelevant to comming up with a base map. My logs were done with the utec in passthrough mode specifically so they would reflect the base state of the car running off the ecu.

anyway, off topic so....

I dont use data from ONE car in developing base maps

The idea is to use data from dozens of cars to make a map which is safe for all and which will make decent power gains and work across a wide range os similarly modified vehicles

BTW - I have found that UTECS in by pass mode have significant effects on the operation of the ecu - again off topic

Battery voltage doesn't really change much at all. The only time it significantly changes is if the idle stumbles and the car dips down to 500-600 RPM. Otherwise it is always between 13.5 V and 14.0x V.

I will use an OBD-II logger in an attempt to corroborate the Evoscan readings.

Also, to further confirm ... it is my understanding that the closed loop fuel trims are based on the load cell in the fuel map +/- whatever percentage of fuel the ECU calculates is needed to hit stoich, correct?

l8r)

Just to clarify further - the latency you would only adjust if you see the fuel delivery varying when the battery voltage is going up and down

It is an offset which deals with the fact that smaller injectors act faster and need less energy to open fully

An injector with larger mechanical parts needs more energy to get it open fully and will act slower on less voltage

The best way to handle this setting is the way AEM does it with a wizzard and you just click on your injector application and the ecu adtomaticaly loads the proper latency map for that specific injetor

Battery voltage does move about at various times like when you start the car or when you flip on your high beams or when the engine fans come on

Most injector compaines publish data on how their injectors react. The problem is that the actual numbers refected in Ecu Flash are just approximations of the hex values in the actual data.

As for the "closed loop" operation

First - changing the values in the fuel map does NOT alter the target

The target is in another map which Ecu Flash does not support yet

IF all your settings in the other active fuel trim maps are set properly then the car will produce an a/f which is near what the desired a/f target map

In some cases you have to manipulate the a/f target tables to tune the closed loop areas to dial in the targets so the car will not go beyond its authority range.

It is important to realize that the value in the a/f target table is not the actual closed loop target which the car is trying to achieve.

The a/f target map is just a fuel trim.

In regards to closed loop fueling tables:
_
This is what I also thought...

l8r)

Al is right though, the Closed loop target AFR is a different map though.. That is, the portion that makes the car hover at 14.7 afr's... I have not found a map specific to that yet, but its rare you would need or want to alter that.. The part I was referring to, and assumed you were referring to, is the minor adjustments needed to the fuel map to keep the fuel trims close enough to zero..

^^ that explains it. Thank you (both Al and MalibuJack) {thumbup}

l8r)