Hello

I have been looking at some obd logs from my car, just the run into work which is about 5miles, traffic wasn't too bad but averaged 15mph!

Just wondering if anyone knows much about what I should expect from the fuel trim values (long and short term for each bank). The internets tell me that single figures are the norm, especially for the long term trim values.

I have
Short Trim bank 1 averaged 6.3% with max 24.9%, min -12.5%
Short Trim bank 2 averaged -2.34% with max 24.9%, min -24.9%

and

long term bank 1 averaged 17.5%
long term bank 2 averaged 10.8%

When I bought the car I was told it had a misfire when cold due to a faulty lambda sensor (following head gasket failure on bank 1), however so far I haven't seen anything in the codes to support this, and think it's more likely due to moisture getting in under the coil covers (new cover to be fitted this weekend).
For the record, the codes are misfire on 7 (& 8 since swapping coils 7&8), generic misfire, and a can bus error.

So:

-Should I be concerned by what looks to me like a pretty high long term fuel trim?

-Are these numbers supporting the lambda sensor theory at all?

-Should I (can I?) clear the long term trim values with a hard reset?

-Are these numbers any use at all, given the short run? I thought that due to the low speed run the ecu will be in closed-loop so more reliant on the 02 sensors.

(+ apologies to those xkec members reading a duplicate post)

Many thanks!

 

The Fuel trims will differ as a function of load, so Idle will be different from 30mph, will be different than 70mph, and there will be a very big variation as you accelerate and decelerate in traffic.

The ECU will go into closed loop pretty quickly after starting the car, but as the engine warms the numbers will change so they are not really comparable.

Rather than looking at averages and ranges across all of those things, you need to see what they do at a steady state.

Best is to drive the car 10 miles or so and be sure it is good and warmed up before making measurements and to take a snapshot on a flat road with the cruise control on so the throttle is constant.

If you want to read more, there is a thread that has a thorough explanation of how the fuel trims work:
https://www.jaguarforums.com/forum/g...ed-quiz-49317/

A hard reset will zero them out, and the car will learn again over the course of about 100miles of mixed driving.

 

And a bigger issue is where are they at when idling and high vacuum? As cars age they pick up vacuum leakage everywhere and this is a predominate reason for stft/ltft to maintain performance/mileage/emmissions as a engine ages. When new stft are pretty much very low single digits, as they age this changes. Ever smoke test a engine and you would be surprised all the areas that leak. And remember that crankcase sealing also affects afrs ratios. So oil leaks from old seals that are not covered with oil will leak vacuum in the crankcase. Cam covers, timing covers oil dipstick, etc

 

thanks chaps, that thread is brilliant, the docs give a very good primer to the situation. I'm going to start recording against engine load and rpm and temperature for each bank which will hopefully give me a bit of insight.
I hadn't realised how much fuel trims could tell you!

 

OK so I have more data and would be interested in your thoughts.
I have plotted for each bank RPM vs Engine load, with circle diameter indicating the amount of +ve or -ve trim on separate plots for each bank. Blue is +ve, red -ve.

here is bank one:

bank1 trims

and bank two:

bank2 trims

The area of interest is at fairly low engine loads - They mostly agree at higher loads and RPM. Bank one seems to add fuel where bank two is removing it.
Is this telling anyone anything? Could the o2 sensor on bank one be reading high, prompting the ecu to increase fuel?

Any feedback appreciated -

 

You must be an engineer, not sure who else would use a bubble chart for something like this.

What makes it hard to interpret is that even though the load is there, it isn't clear what you are actually doing with the throttle for each data point, which can make a huge difference.

I think this might be a situation where less data is more, so let's start with a simple and specific scenario:

Warm the car up, let it sit in the driveway idling in P until the trims stabilize. What are the STFT and LTFT numbers for both banks in this scenario. If you want, you can also record the Mass Air g/s.

 

+1 parked, engine hot
For me just LTFT at idle and rev to about 2500. 4 numbers.

 

OK properly warmed up, in park and idling:

Mass air flow rate 4.75

ST Bank 1: 5.4
ST bank 2: 0.78

LT Bank 1: 19.53
LT Bank 2: 7.82

Bank 1 looks a bit high to me.

When I increase revs (to about 2K from memory):

MAF rate 16.35
ST bank 1: 12.5
ST bank 2: 2.34

LT bank 1: 13.2
LT bank 2: 12.5

Penny for your thoughts!

 

19.5 is the max range for LTFT, so the ECU is having to add 5% more fuel on top of that even to keep bank1 in line.

A misfire can cause higher trims but you state that those are on 7 & 8, so bank2.

Air leaks are the most common problem that cause high trims. Yours are mainly on the bank1 side through, so the search would focus on things downstream of the supercharger where the airflow gets split between the two banks. Outlet duct seals or the seals around the fuel injector manifold.

A smoke test is best, but barring that you can use MAF cleaner and watch the STFT. MAF cleaner is better than carb cleaner or starter fluid for this as it won't eat the plastic parts the way the others will.

If you know the head was off for a gasket failure on the bank1 side, my first inclination is that they probably reused the gaskets around the injector manifold and didn't get it sealed properly.

Other possibilities are a misfire on bank1 or O2 sensor, but you don't mention any codes to support that.

 

thanks for that, sounds most helpful. I'll reveal my naivety by asking how to use the MAF cleaner. I suspect you mean spray it on to areas I suspect could be leaking, and if the STFT drops then the seal has been temporarily plugged?

 

You got it, Mike. Good luck.

 

What about an exhaust leak before the catalyst? / Bo

 

For me the LTFTs didn't drop enough to prove much of a leak and - worse for the leak theory - bank 2 got higher. That's not a leak as far as I can see.

Whether it's a bad MAF, bad coils or blocked cat(s), or indeed something else, I'm not sure.

 

To clean the MAFS you need to remove it and spray the probe with MAFS cleaner but do not touch it with a rag or your fingers let it drip dry. Reinstall and do a hard reset. The hard reset will reset all settings and that will include your windows, radio stations and radio security code. Drive the car for about 20 to 30 mi before you take another reading. If you encounter another P0171 & P0174 I would be looking for a vacuum leak first. When taking your readings use this guide and get the readings in the RPM range. I recommend that if you perform this test you do it outside the garage and no other objects in front of the car. You will understand when you perform this test.

Link to Guide JagRepair.com - Jaguar Repair Information Resource

I hope this helps.

 

thanks for the tips, although I don't understand the logic behind a faulty MAFS resulting in one bank running too lean. ccfulton seems to make sense in that if it's one side only, surely it the issue is after the intake is split?

I haven't had any codes telling me I'm too lean or rich btw, the only codes I have are misfire on 7&8 (bank 2 on my 2000 engine) which I now think I've solved by changing the coil covers.

I was suspect of the readings but have no performance problems whatsoever. So blocked cat seems unlikely to me. As I mention in the first post the bank1 02 sensor has been mentioned by the previous owner but there are no codes to support this, just the fuel trim data.

 

If you have no codes & no problems then I suggest letting it go. The difference in the readings could be caused by age or equipment that has been replaced.

 

you wouldn't be concerned that it's adding 25% more fuel to one bank?

I think I'd like to sort this. I'm going to try checking for leaks with maf cleaner and will try and get lambda sensor readings - the iphone app rev doesn't support them, but I think for android torque does.

 

I'd be concerned. If you've had coil trouble then a damaged cat is fairly likely and they can cause odd looking trims as you have.

 

Yes I would be concerned if I had a code.

 

When was the last time you used a fuel additive in the car? I would get a can of BG44k and add it to the fuel then do another test after that tank of fuel.