SiHethers

The NBout for the 2B0 is based around the Bosch LSM-11 NB sensor, and I've found that the simulated voltage is not quite perfect for closed loop control using either the OE ecu or Power FC on my MY99. This is because stoich ratio (14.7:1) is around 0.7V on the OEM NB but around 0.5V on the simulated LSM-11. The result of this on my car is that the closed loop control centres around 14.2:1, and swings from 13.5-14.8.

There is now a working WBconfig (v1.5.a) downloadable from the wbO2 site which I have been able to use to create a new look-up table for NBout and then flash the EEPROM that has resulted in much tighter control of closed loop fuelling (stoich now spot-on 0.7V with swings from 14.5-15.3:1 during cruise). In essence you can create/set your stoich voltage to whatever you want and also modify the gradient of the slope around this point thereby fine-tuning the control.

In brief, I first flashed my firmware and EEPROM to the latest update (dated 22/03/05) using wbflash-03 from the TE site. Then open up the wbconfig and select the correct COM port which you've attached to the 2B0 (wait till the red light is on solid on the 2B0 though otherwise you'll get an error). Click on the AFR tables tab and select the Output pin to NBout. Leave Ouput data as Linear parameters. Ensure that stoich is set to 14.7 AFR as the ouptut graph you produce uses lambda for the X axis. Then select 2 mid points.

Then select rich to be 11.2/1.0V, mid point 1 to 14.9/0.89V, mid point 2 to 15.5/0.15V and lean to 19.0/0V. Click on Generate to create the new table, then click the Graph button to display your new lookup table as a graph of NBout vs lambda. If you're happy then click flash. Turn your ignition, and hence 2B0, off then on again (not sure if this is really necessary) start your car and see where your idle now sits.

It is very likely you will need to adjust the 2 mid point AFR values to shift the steep part of the curve to where your car needs it (depending on how much offset you have between the true AFR/NBout and the apparent AFR/NBout that your ecu sees). Setting up as above, 0.7V should be 15:1 AFR but due to the offset in my car it actually sits at 14.7:1.

If you have Power FC you can do even more. Next turn off closed loop control in the etc menu of the commander. Then in Setting-Ign/inj use the overall injector trim to adjust your idle AFR to 14.7:1 (or whatever you want 0.7V to be), the go to etc-sensor check to see what the actual NBout voltage is. From this you can work out whether you need to shift your mid points up or down to hit the AFR that you want. You will need to experiment to get it exactly how you want (took me about an hour in total with around 10 different reflashes/tables)

Easy! Roll on better fuel economy and idle/cruise characteristics.

Simon

Andy.F

Nice one Simon
Did I mention to you the adjustable trip point between open/closed loop on the fuel map ?

Andy

john banks

Sounds good. I did wonder about making something up to take the WBLin signal and producing a simulated narrowband output with variable AFR depending on boost level (& possibly RPM). However, I'm a bit busy trying to get the PowerFC to run nicely off MAP at present.

SiHethers

Hi Andy and John

Andy I think you did mention it, but is that something you can adjust with the commander or do you need datalogit for that. Can you also adjust the feedback sensitivity within datalogit?

John, sounds like you're after closed loop fuel control on boost I'm not convinced the way I have the NBout set up would be suitable for that at the moment as the fluctuation in AFR is to great, but I would think that as you increase injector opening time the fluctuation should get less. My interpretation of the the way power FC works at the moment is that your map specified fuelling has some effect on the way the AFR fluctuates also, as if it tries to fuel the map specified AFR then adjusts to get the NBout less than 0.7V, then back again IYSWIM.

Currently the power FC swings the NBout voltage between 0.9 and 0.3V, but I'm not sure that that is by design, but rather a feature of the slope of the simulate NBout curve, the feedback sensitivity on closed loop fuelling specified within the PFC, and the map specified injector opening time. I found I could get tighter control by increasing the max voltage to say 1.3V, giving some extra resolution on the rich side of target, but I wasn't sure how the PFC would handle that. I didn't want to create a little fire in my footwell

I think more experimentation is needed, and datalogit is probably required to get the best out of it. What I'm thinking is mapping the cruise fuelling as close to stoich as you can get so you can run open loop there, but then switch to closed loop on boost with a NBout table that targets your ideal AFR. Do I dare risk it though...

Simon

PS None of this has made a difference to the LD01 indicated AFR Andy

AndrewC

LD01 is an averaged value based on a sampled version of the SVout signal it's nothing like as accurate as the value seen in TEWBLog or the later digital displays.

FWIW you can modify the response curves for all 3 outputs NBSim/WBLin and SVout.

Andrew...

Andy.F

No feedback sensitivity exists but the feedback loop naturally becomes faster as engine load/rpm increases anyway.
You can set an upper limit to the fuel map value which disables closed loop.
On yours it is currently 1.047
Only in the FCD I'm afraid.

I had considered running offset closed loop WOT when running the map convertor box (that Andrew C built and John B is now working on) but just to overcome not having temperature compensation. With a MAF, the AFR values are very repeatable and running CLC would only increase the risk to the engine WRT sensor failure/drift.

Andy

ScoobyDuck

iTrader: (1)

does it work with 2A0 versions ?

the TE website is so messy to navigate. have found the S/W and will give it a go at the w.end - if it works that is !

Pavlo

With the right software you can adjust the closed loop target voltage and parameters on the Power FC

ScoobyDuck

iTrader: (1)

yes, i know Paul, but all that does is change the point at which CL is initiated is it not ? it will not change the output of the WB which will still have its stoich point too rich for the ECU

also gona be OEM soon too

ScoobyDuck

iTrader: (1)

yes it DOES work on the 2A0 !

although now, its not all rosey!

had to scale things a bit differently : lowering the first mid point voltage to 0.7v (still @ 14.9 AFR) and get a fairly rock steady 0.7v on the PFC's O2 input.

turn on closed loop and it swings from 14.6 15.3 in about 2 seconds

however, the INJ -/+ setting doesnt change in the log. which i am expecting it to !
and the purge solenoid is ticking away like a noisy tappet.

turn off CL and it returns to 14.7 as set by the map & the tappety noise stops. have tried richening and leaning the map , its still copes with CL (until you go over 1.047) but all of the above issues still occur !

any ideas anyone?

Steve

Andy.F

turn on closed loop and it swings from 14.6 15.3 in about 2 seconds
>>> OK, this is normal.

however, the INJ -/+ setting doesnt change in the log. which i am expecting it to !
>>> This is a manual input, not a live ECU controlled value. Your IDC will be changing +/- 0.2-0.3 %

and the purge solenoid is ticking away like a noisy tappet.
>>> Yes, they do as long as it's in closed loop for a preset time.

turn off CL and it returns to 14.7 as set by the map & the tappety noise stops. have tried richening and leaning the map , its still copes with CL (until you go over 1.047) but all of the above issues still occur !
>>> If the noise bothers you, I'd suggest either run open loop at idle by setting above CL trip point or unplug the sol valve assuming you have disposed of the charcoal canister ? The early cars don't have a sol valve so it is easily eliminated.

Andy

ScoobyDuck

iTrader: (1)

thanks for the reply Andy,

ok, good


manual input ? as in the INJ/IGN 'run time' change settings ? I though this would be the INJ correction factor? guess not. INJ duties are as you say moving +/-

really ? thats crap. OEM didnt do this and it sounds really bad! i wonder if the solenoid is a bit duff...

humm..could do that i suppose, it wouldnt be too rich at >CL trip limit. Still got all OEM stuff, so cant really remove it.

cheers
Steve

Scott.T

iTrader: (2)

Thought I'd bump this post up to the top as I have just had a play with the NBSim. To get the stoich closed loop running a bit tighter.

I set mine to :

Rich 11.2/1.0
Mid Point 1 14.9/0.7
Mid Point 2 15.5/0.10
Lean 19/0

Seems to run between 13.8 and 14.9 at the moment.
Will see how it goes.....

Before the tweak is was running about 14.3 to 15.1

What are the default settings put in my TechEdge or Andrew ??