Understanding Fuel Trims?

[BahnStormer]

After a recent dyno pull where the AFR came in a little lean (discussed here), I decided to try logging some of the fuel trims using TorquePro and a little OBD2 reader... and I now know that I know nothing.

I have lots of numbers and very little idea of what any of them mean.

Apparently @shnazzle is the guru on these, but I'm happy just to be pointed to a "how to" guide to get these to make sense...

I grabbed all sensor logs that looked like they might be useful, but I think this is just adding to my confusion, amongst others, I seem to have values for these:
Engine RPM(rpm)
Fuel Trim Bank 1 Long Term(%)   
Fuel trim bank 1 sensor 1(%)   
Fuel trim bank 1 sensor 2(%)   
Fuel Trim Bank 1 Short Term(%)   
Fuel Trim Bank 2 Long Term(%)   
Fuel trim bank 2 sensor 1(%)
Fuel Trim Bank 2 Short Term(%) [edit: I missed this on the first log file gathering]   
Intake Air Temperature(°C)   
Mass Air Flow Rate(g/s)      
O2 Sensor8 Equivalence Ratio   
O2 Sensor8 wide-range Voltage   
O2 Volts Bank 1 sensor 1(V)   
O2 Volts Bank 1 sensor 2(V)   
O2 Volts Bank 2 sensor 1(V)   
Speed (GPS)(mph)   
Speed (OBD)(mph)   
Throttle Position(Manifold)(%)   
Timing Advance(°)   
Vacuum Gauge(psi)   
Volumetric Efficiency (Calculated)(%)

Now what do I do with them?!?

[shnazzle]

If it's been a while since your last ecu reset, all you're really interested in is the long term fuel trims across both banks. 

If you scroll through the log and don't see any values over, say 6 or 7,you're good.

Let's say your long term percentages across both banks are consistently positive and over 10; you're running lean. It means the car is adding fuel. So 8.27 for example means that for that particular RPM and load (not throttle, per se), the o2 sensors are reading too much oxygen (lean state) and your car is adding 8.27% fuel over and above what it has been mapped to.
Vice versa, negative values mean a rich state (too much fuel) and the ecu is subtracting fuel. 

In both cases, depending on how positive or negative, you may have a problem, or just an ageing car. If it's never more than 5-6%either way, I'd just put it down to age-related compression losses, sensor ageing, etc.

Let's say you find a problem; e.g. The car is consistently adding fuel across RPMs. Something is causing the car to think more air than expected is coming in. A vacuum leak for example. But also a manifold leak can cause this. 

Start simple;look at the long term fuel trim column. You want 0 or as close to as possible with a margin of 5-6%. Go from there. 

[BahnStormer]

Thanks for that Patrick - the only problem is that I fitted my battery cut out last weekend (a week AFTER the dyno pull), so pretty sure that would count as an ECU reset and that presumably means that these "long term" trims are only based on ~80 miles of previous driving...

Not sure if they're meaningful at all, but short term numbers only show Bank 1 (edit: I found that I missed ticking Bank2) and the "other" fuel trim (spot values??) has both banks... and  both of those show a few higher values...

I'll keep an eye on the Long Term numbers and log them after a few more drives.... current values show max values of ~6 and min values of ~minus 12... but this was a fairly abnormal run where I was working it up and down through the gears a fair bit, including a fair bit of shifting down while decellerating, which I think is where it's overfuelling slightly and i'm getting the little pops from the exhaust like a little AMG...  childish, but fun.

[BahnStormer]

here's the .zip with the excel file it in, if that makes more sense to you?

I've hidden the other columns - just left the long term fuel trims...

[Cap]

Lets Chat fuel trims..

Your AFR's are done at WOT..  so it's all fuel trims doing the work, the O2's are out of the Picture during this time..

What I have seen, is when the Fuel Trims are Positive, the ECU will tend to be LEAN in WOT conditions..  If your Able to 'Force' the Fuel Trims toward the Negative Side, the AFR's 'Tend' to be richer at WOT..

If you are able to Pull the Fuel Pressure Regulator out of the Fuel Tank, and 'SLIGHTLY' Squeeze the Flange that the Spring Sits on, you would be able to raise the Relief Pressure up a little, and force the Fuel pressure Higher..  This will force the Fuel Trims More Negative ( The ECU will be now pulling Fuel ), and the Trend I saw when I was Logging for the MAF Mod, was the Mixture will be richer in the Center of your Torque Curve..

For my Set up, a -7 Fuel Trim Long Term was my Golden Spot.. Your spot might be different..

I built a Fixture to hold the Fuel Regulator, and force air through it..  with a Pressure gauge showing the Air Pressure..  I then Slightly Compressed the Spring Flange end of the Regulator, while watching the Air Pressure..  This way I could 'Calibrate' the Fuel Pressure Regulator to what ever I wanted..

I did this Playing for Turbo Applications.. ( Not mine, I'm a NA sort of Guy ).. But I also did this for Playing with the MAF Mod.

Cap

[thetyrant]

Fuel trims look ok to me on your graphs so doesnt look like there is anything wrong in the closed loop side of things that may of caused the leaner AFR on dyno under WOT/open loop fuelling, whiles its not 100% sure its more than likely if the car was running 14 AFR at full chat like it shows on dyno something would be shown in the fuel trims for the lower rev/closed loop fueling.

As Cap says there is some carry over from closed loop fuel trimming into WOT/open loop so if you had big deviation from 0 trim (much more than +/- 10% which is accepable norm) in the long term trim under normal closed loop driving it might of pointed to a issue that contributed to lean WOT, i wouldnt worry too much more about it but would be good to get another AFR check at some point to see if its same as previous one.

[thetyrant]

Lets Chat fuel trims..

Your AFR's are done at WOT..  so it's all fuel trims doing the work, the O2's are out of the Picture during this time..

What I have seen, is when the Fuel Trims are Positive, the ECU will tend to be LEAN in WOT conditions..  If your Able to 'Force' the Fuel Trims toward the Negative Side, the AFR's 'Tend' to be richer at WOT..

If you are able to Pull the Fuel Pressure Regulator out of the Fuel Tank, and 'SLIGHTLY' Squeeze the Flange that the Spring Sits on, you would be able to raise the Relief Pressure up a little, and force the Fuel pressure Higher..  This will force the Fuel Trims More Negative ( The ECU will be now pulling Fuel ), and the Trend I saw when I was Logging for the MAF Mod, was the Mixture will be richer in the Center of your Torque Curve..

For my Set up, a -7 Fuel Trim Long Term was my Golden Spot.. Your spot might be different..

I built a Fixture to hold the Fuel Regulator, and force air through it..  with a Pressure gauge showing the Air Pressure..  I then Slightly Compressed the Spring Flange end of the Regulator, while watching the Air Pressure..  This way I could 'Calibrate' the Fuel Pressure Regulator to what ever I wanted..

I did this Playing for Turbo Applications.. ( Not mine, I'm a NA sort of Guy ).. But I also did this for Playing with the MAF Mod.

Cap

Cap what is your opinion of the AFR shown on the dyno pull on first post ?  i know youve seen a lot more AFR data on the 1zz than me

[shnazzle]

A) nothing worrying in those long term fuel trims, and they certainly don't need more than 80 miles to settle. Could argue that one bank is running slightly too rich. Slightly leaking injector or something. No action needed. 

B) @Cap are you thinking along the lines of MAF mod for -7 being a sweet spot? Or just plain stock?

I have seen also, on a fresh reset, that if fuel trims are negative, the first thing the ecu does in open loop is put a value of 4.7 for all cells. 
If the trims are positive, it puts a flat -4.7 in all open loop cells.
Then seasons to taste as it goes. 
So it does do a decent jump in trims in response to momentary short term trims. 

I've always wondered what the significance was of that specific number. It's actually more specific than that. Something like 4.7382 or whatever. Any clues? Or just randomly defined safety margin?

O2 sensors are ignored at WOT as you say, which is why I had my piggyback, so that I could take control of that area which the ecu was messing about with based on its closed loop findings. Easier said than done as it turned out. Especially as it adjusts long term trims on closed loop areas. 
Great for stock. Flippin miserable when you're trying to tinker!

But... All in all, with everything we've seen, it doesn't seem to tie in with what @BahnStormer saw on the dyno.
It seems to have carried over the average (which seems to be predominantly negative trims) into the open loop areas. 
Which makes sense. 
If the car is always rich during closed loop, it seems a safe assumption then to pull fuel at WOT. It just seems to have gone a bit overboard and pulled too much.

[BahnStormer]

Thanks for the initial input guys!

Slightly strange that the previous dyno reading said slightly too lean, but the O2 sensors are saying slightly too rich, but it sounds like I'm in the right range at least.

I'm not overly worried, but I will be using this as an excuse to run Momentum/Ultrima/V-Power fuel for now and let's continue the conversation.

Happy to keep on providing example data though as I've never really got into this side of tuning (or just checking the car's health), but I'm keen to learn!

I'll check the spacing of the sensors, but the Toyosport manifold runs unequal length headers, so presumably taking measurements at different points could give slightly different "adjacent" readings? Although I guess at 1 sample per sec, the extra few centimetres won't make any difference?

[thetyrant]

Thanks for the initial input guys!

Slightly strange that the previous dyno reading said slightly too lean, but the O2 sensors are saying slightly too rich, but it sounds like I'm in the right range at least.

I'm not overly worried, but I will be using this as an excuse to run Momentum/Ultrima/V-Power fuel for now and let's continue the conversation.

Happy to keep on providing example data though as I've never really got into this side of tuning (or just checking the car's health), but I'm keen to learn!

I'll check the spacing of the sensors, but the Toyosport manifold runs unequal length headers, so presumably taking measurements at different points could give slightly different "adjacent" readings? Although I guess at 1 sample per sec, the extra few centimetres won't make any difference?

Its an odd one for sure, the Dyno shown AFR is slightly to lean from what many consider optimal for na car (around 13 - 13.5 afr at full chat) probably why it made good power if that is accurate,  but its a long way from what a stock 1zz usually puts out from my research which is low 12's/high 11's and much much richer,  more turbo like AFR numbers than na cars usually put out so i think it was something to do with the dynos wideband or how it was fitted into tailpipe.

Shame you have since the dyno run reset ecu as maybe the fuel trims at time of dyno would of shown something, but at least its running ok now so i wouldnt worry just do the odd check on LTFT to make sure its not straying too far from +/- 10% area, its easy to get obsessed with fuel trims but they are there to compensate for changes in temps and pressures etc and should be moving around a little, as long as its not fully one way or another or much over/under 10% they are considered good