Fuel trim figures

[edbar...@googlemail.com]

Hi. I'm currently using mems rosco.
I'm having fueling issues so thought I would scan the ecu on my k series metro 1.4 mpi 16v.
The Long term fuel trim is currently 124. I understand the default '0' position is 138? Is that correct?
Also which side of 138 is 124. Is the car running rich and the ecu is leaning the mix or is it running lean and the ecu is adding more fuel to compensate?
Any information would be very much appreciated.

The program itself is fantastic.
Thank you
Ed

[leopold]

Hi normal fuel trim is 125 to128 according to this 124 is slightly lean.What about your AFR (air fuel ratio)?
With 128 the engine acts better on the throttle.
What about the short term fuel trim should be at 100.(when engine runs in closed loop and lambda status ok.
By the way, what issues do you have?
Leopold

[Martin Rubenstein]

128 not 138 is the mean position.     So yours at 124 means the ECU has backed off slightly on the fuelling compared to the factory-set position.     Mine is currently 117.   If everything is working as it should and the car is going into closed loop, then the ECU is actively controlling fuelling second by second based on information coming back from the oxygen sensor, and there’s nothing to worry about.   On the other hand, if your number was a lot higher than 128 (don’t ask me what a lot is), a lot more fuel would be being added than the factory envisaged and you might start looking for vacuum leaks.     If it was very much less than 128, you might start looking for a leaking injector or too much fuel pressure, say.    The important thing is keep your eye on that number especially once you have got the car running so you’re happy with it.    It should stay fairly constant.     You’re most interested in a sudden change rather than the absolute number, provided that number is reasonably near to 128.     After I renovated my cylinder head, I think mine was somewhere between 112 and 116.   So that gives you an idea about the magnitude of deviations from the factory setting - every engine is unique.     So it pays to get a feel for LTFT as well as normal STFT, so that you can recognise a sudden change from the norm.   But at the end of the day, in closed loop, the ECU is in charge, and the fuel map as burned in by the factory is just a guide to put the ECU in the right area to start off with.

(In a test a few years ago, I reduced my LTFT down to 60 to watch how the ECU would deal with it.    When the car went into closed loop, it was really running lean and the ECU ramped up the fuelling via the STFT and kept ramping it up until the STFT reached around 20% additional fuelling.    At that point, the ECU realised that such an amount of enrichment (with more needed) was abnormal, and the car went into open loop.      On the other hand, when I reduced the LTFT to 100, the STFT increased, to compensate, but didn’t quite hit the critical threshold for going into open loop.    The car was able to run at that setting, I left it alone, and after a few weeks, when I next checked, adaptation had taken place and the LTFT had gone back down to what it had been beforehand, and STFT was back down to pretty close to zero.)

[edbar...@googlemail.com]

Thank you for the reply. Stft is 100. It occasionally drops to 82 when I rev the engine to around 3000.
The car runs and idles fine but sometimes is a struggle to start. This seems to be when it's warm outside??
It goes to closed loop and the lambda is functioning I think as it should.
Air fuel ratio is 145 at idle and is generally stable. It doesn't go higher but has dropped to 135 under load.

I recently had the injectors cleaned and 2 failed static leak test and all had reduced flow rates. All came back cleaned and supposedly leak free with good flow rates. Cold starts improved.

The car is still using up a lot of fuel. I don't use it all that often and have only just made up a cable to use for diagnostics.

I had 3 fault codes that when I cleared the lambda sensor status instantly changed from 0 to 1. Up to that point stft had never reached 100 and was always under. Ranging from 91 to 99. Within a minute or so it changed to 100.

I haven't had a chance to put many miles on it since. That was at the weekend and I use a family car to take kids to school during the week.

I will go for a long drive this coming weekend and collect a load more data to see if anything changes. I fear I still have an injector leak all be it a smaller one than before but it could be it needs a longer run now that the stft has altered slightly.

Thanks again for your reply.

[leopold]

"Air fuel ratio is 145 at idle and is generally stable. It doesn't go higher but has dropped to 135 under load."

I would slightly increase fuel trim.

[Martin Rubenstein]

"Dropping" to 135 under load is I believe, perfectly normal.     (I think "increasing" to 135 might be more appropriate; I presume 145 indicates and air-fuel ratio (AFR) of 14.5:1 and 135 is an AFR of 13.5:1)..     13.5:1 is the figure I've seen in the literature for the AFR in open loop under load/acceleration, when you' want power, not fuel economy or emission control..

Do give Leopold's suggestion a try and see what happens over time (and let us know).     I would expect your LTFT to adapt slowly back to what it is now, but it would be an interesting exercise.

[leopold]

Of course, a logfile posted here would be a great help for all who want to assist you!

[edbar...@googlemail.com]

Absolutely. I have no idea how to add files on here?
Currently I only have a cvs for an android app called Rover mems.
I can create a file from mems rosco on my laptop this weekend.

Thank you again to you both.

[leopold]

What is difficult on adding a file here?
Send it to your email addres, save it ,then attach it here.

[edbar...@googlemail.com]

My computer isnt showing an attach button. I have tried on mobile as well. Sadly no dice.

[Ed Barnard]

I had to join the group to attach files.

 Easy.

This is all I have currently but will do a better file later. 

Thanks again

[leopold]

Either your temperature sensor is bad or the thermostat is not working right.When you started your engine again temperature fells down to 20 degrees.I think therefore ecu never switches to closed loop even lambdastatus is ok

[leopold]

[leopold]

this time with iac, after restart engine runs in limp mode

[leopold]

Did you change the temperature sensor and get one with a reversed temperature coefficient.
That would explain the temperature curve as well as that lambda status is ok when starting and off shortly after the restart. And never switching to closed loop.

[Ed Barnard]

Thank you for looking.

I may be miss reading my csv file or something is getting mixed up in the conversion to graphs. When I open the file as a spread sheet the coolant temperature is only at 19 degrees when I first turn the diagnostics on.Its then steadily rises to 70 degrees until I switch the car off.
I dont connect the diagnostics until the following day when the temperature is at 87 until it hovers around 93 degrees until again I switch the car off.

Its possible because the file is collected over 2 separate days the program you have used to create the graphs uses the time and doesnt allow for a change in date. The file starts on Sunday evening 7:30pm then re starts on Monday morning 8:30am. It looks like the graph has the days mixed around?

I could be wrong.

I will use mems rosco on my pc to do a new file that is just one complete run from cold start to getting up to temp while driving. I have been using android app rover mems because my cable was short. I have now extended it so I can have a laptop in the car and drive it.

Also, to answer the temperature sensor question, I have not changed it. I have replaced the o2 sensor recently while doing rust removal and new gasket on the exhaust manifold. Its a bosch one not a generic splice your connector on type.

Thanks again for helping me.

Ed

[leopold]

what about the new mems-rosco logfile?

[edbar...@googlemail.com]

Thank you for your continued help and interest. 

The file was while parked in the garage. Hopefully it shows something helpful. Unfortunately the laptop that I have access to belongs to my partner and is her work computer. I might get the time to get a file while driving the car if that one isn't in depth enough please let me know.

Thanks again Leopold

Ed

[leopold]

Hi Ed
I think the car is running fine.It runs a bit lean.Thats what I told you to the beginning.LTFT could be 128 but 124 is good too. IAC could be a bit less. But hot idle is 15 and the lowest IAC is 19.That is in the range.Lambda is good and working so I would drive it for a while and have a look what fuel consumption is now.Tell us again when you have wasted two or three tanks full of gas.

[Martin Rubenstein]

Ed,

After you've been on the long run, stick your finger up the tailpipe and see if it is sooty.     It ought to be quite clean.     Your lambdas sensor trace looks normal, as Leopold has said.

Your concern is that it seems to be using up a lot more fuel than normal, but your lambda trace seems to show the ECU is controlling fuelling exactly as it should.    (Of course, the sample rate of the ECU data is too low to measure the switch rate or response time of the sensor, so it's not possible to say if your lambda sensor has got old, or lazy or been contaminated - you'd need a 'scope for that).

Have you quantified how much more fuel than normal it's using or is it just a gut feeling?

Going back to basics, have you removed and looked at the colour of the spark plugs (good time to do that is after a long run - when the engine's cooled down, of course, unless you want to burn your hands and drop the plugs)?     And when were they last renewed?

When was your last MOT emissions test, or when's the next one due?    If you're on good terms with your MOT tester, you should find they'll do an emissions test for you at the end of the day for a nominal fee if that.     That should set your mind at rest.