NefMoto

When my car decelerates and enters fuel-cut off condition (b_sa toggles), my car seems to still inject fuel with whatever current TEMIN value holds...

This results in wideband reading high limit for AFR (19.6:1 on gauge, 21:1 in zeitronix) since no ignition takes place but injectors appear to squirt with TEMIN value anyway it is just not visible on the gauge but would explain my lower than average MPG for allroad.
 
Is that how its supposed to work or legitimate b_sa set should result in 0.00 injector time?

teminva needs to be 0.00

It is stock at 0.00

The weird part is that the injector on time seems to be leaning on TEMIN value instead with b_sa set, so I should be getting fuel cut off.

The question is then: does stock unaltered bin go to 0.00 injector time or it also leans on TEMIN?

Reading the Me7.1 doc is like pulling teeth, even the translated one... or maybe specifically the translated one :)

However, as far as I can tell, TEMINVA whose usage is triggered by b_va bit, is a minimum injection time during warmup unless I am missing something. As such, the TEMINVA will be used until the engine reaches 60c and then it switches to TEMIN.

I heard comments recently (hi Nyet!) that you should see fuel cut off when decelerating but I just don't see it...

I would appreciate if somebody could post a short piece of log where it shows injectors activity or lack of thereof on decel.

I just checked a random log and it confirms what you are saying.  At 25% pedal/10% throttle, 3500 rpms in 3rd the fuel injector on time is 1.75-2.2 and desired AFR is 0.999755859375.  It dips to 1.1 range on initial throttle lift and desired afr goes to 1.046875, at 0% pedal/throttle and over-run it then hovers ate 2.3-2.5 as the engine should just be acting as an air pump....  hmmm.

this from my car, i do have a bov, yet it isnt open during most over run to my knowledge.  I swear I have logged 0.00 INJ before.  maybe its something we did in our files?

I will post logs from a stock m-box and modded in a little bit, on different laptop.

Jared: just had an interesting exchange with julex.

Julex if you get a chance, summarize it :)

I'll work on the wiki time permitting.

So here is the summary:

TEMIN appears to be used as a wall wetting/lubricating mechanism.
TEMINVA appears to be used for the same purpose when engine is cold. Since fuel doesn't evaporate even closely as violently on cold metal vs one that has several degrees, they keep it at zero.

When decelerating, depending on b_va set (b_va - warmup - TEMINVA used), the minimum injector pulse width will adhere either to TEMIN or TEMINVA. It will never fall below these values, again depending on b_va set.

It makes it then important to scale TEMIN by your injector size factor over stock size not just low enough to stop affecting idle. I believe that too large of TEMIN will give you weird rich off or low throttle, fuel dumped in overrun which will result in AFRs dwelling in rich area for few seconds ending in overshooting into lean once lambda control overshoots and popping when resuming throttle.

How I would calculate TEMIN:
[new injector flow] / [Stock Flow] = X
TEMIN / X = New TEMIN

For example:
1000cc / 350cc = 2.85
0.50 / 2.85 = 0.175 TEMIN

Care to comment on TVUB interactions with TEMIN... I assume if you have a "slow" injector, TEMIN might need to be boosted a tad?

TVUB should take care of all inadequacies of injector IMHO. You could use TEMIN to prevent engine stall but then you sacrifice low load drivability and some economy. Wiki already provides info on what can be done to prevent misfires etc, for some injectors you just have to raise idle RPMs/torque reserve due to their bad low pulse width performance.

Stupid question: I thought TVUB doesn't correct TEMIN?

TVUB is always added to final injector on time calculated for a given fuel mass needed in cylider.

So yes, TEMIN is corrected by TVUB. If your TEMIN is 0.5 and TVUB 0.8, then your final lowest injector on time will be 1.3ms.


Next time log te_w and te2_w variables, this is pre TVUB correction inside KRTI module. They are actual calculated injector on time values for perfect injector without any lag (0.00ms TVUB).

Fuel cutoff is applied after the calculation of ti_b1(2).

It is applied to ti_ev0...7 in ACIFI.

Also, ball park napkin calcs tells me that TEMIN during cutoff is still quite a bit of fuel.

I am skeptical that the injectors are still firing..

I like where this is going. Let me look it up.

Think about it this way, what if you go off throttle at 6000rpm?

Thats a SHITLOAD of fuel if you are still squrting TEMIN * 6 * 6000 / 2

func ESWE

Would this cause the revs to hang longer off-throttle?  Something I've noticed with my TEMIN where you have yours, though I'm not sure it is the cause.


Just from a feeling, that seems pretty low. Have you tried that value on your 1000cc's, and what were your observations if so?

i'm running .20 on ev14's 52# injectors which are operating as 625 cc/min @ 4-bar.  i think my tune has this a tad low and that his calculation of .50/(625/350) = .28 should be more accurate.

having calculated this at work, i got home and changed only this constant on my current tune.  i've never had the car this smooth while cold since it was stage 2.  however, i now need to adjust my tvub values - which i expected.  the car was running quite rich, which makes sense because the idle component of the (idle + tvub) went from 1.00ms to 1.08ms, an increase of 8%.

thanks to the knowledge here, i'm going to update my tvub calculation spreadsheet to reflect this information.

I also don't believe that fuel is dumped when off the gas pedal.
As phil said, fuel cutoff is done after ti got computed:

So you can still log ti > 0, but the injector is not opened in reality.
The injection driver code finally decides which cylinder really gets
injection and which one not.
The bitmask evz_austot is used to disable injection on cylinder basis.
If a bit is set to 1 in this bitmask, the coresponding cylinder doesn't get fuel.

You can log the following two variables in M-Box to see wether injection takes place or not:
redist          , {Inj-Reduction} , 0x380B92, 1, 0x0000, {#} , 0, 0, 1, 0, {Ist-Reduzierstufe}
evz_austot   , {Inj-Cut-Mask}  , 0x380B90, 1, 0x0000, {-} , 0, 0, 1, 0, {Einspritzausblendmuster total}

'redist' is just the number of 1-bits in evz_austot, so this is the current reduction step which
should read 6 if all cylinders see a fuel cut and 0 for normal operation.
'evz_austot' shows which cylinders are cut. Unfortunately you will have to convert this value in logfiles
by yourself from decimal to a hex value or a binary value to see which bits are set/cleared therein.

Thanks Setzi.

as usual setzi is awesome.

2nd 3rd and 4th!

This morning, when I went to update my TEMIN value to accommodate the increase in flow from my stock 1.8T AWP EV6 314cc injectors, to my TT225 BAM 386cc EV12's, I noticed that the stock bin for the 1.8T AWP and the 1.8T TT225 BAM have identical TEMIN values (0.75).  Is this an oversight from the factory?  Or am I missing something here?

I think the value is low enough on most tunes to never go into effect actually

Hi guys,

I would like to disable fuel-cut off on my ECU ( to make back-fires in exhaust when decelerating ).

I read ACIFI, but it's not clear how to do. Can somebody help me?

Few ways,

change rate of change of RPM for activation, or EGT threshold are a couple.

Hi Rick,

In ACIFI module ? or in MDRED ?

I found KFTVSA in BBSAWE Module :

KFTVSA Delay time for overrun fuel cut-f

I'll try  ;D