NefMoto

hello all,

I decided to redo my KFMIRL and KFMIOP tables as I was getting some part throttle jerkiness. I used Masterj's tool to calculate IRL and the appropriate IOP, then converted my KFZWOP/2 to suit the axis change of IOP. The table is still STOCK, just extended(interpolated) for the new load points on the axis.

Now my car has a bouncing idle. After first start it idles fine, but after about 5 mins it dips down and then catches itself. Do I need to reflect the new changes to IOP in the map KFMI_UM?

I should note, CWMDAPP is set to 8 currently for calibrating KFLDRL.

TIA.

Stick with stock values for the low load areas of IOP and IRL

Even for a BT 1.8t? The torque output of my turbo is drastically different than a k03. I don't make full boost till about 4.7k RPM.

man, you need to do some reading - no offense.  your car isn't using the turbo-charger for lower loads.  its just a variable air-pump controlled by the cams and ignition timing...  why would it change?

your turbos aren't doing squat during idle :P

You're probably right but it was my understanding that IRL is target filling via requested load and RPM. In my mind, that means a smaller turbo pushing more air at lower RPM requires a higher target filling than a large turbo?


I've read through berTTos' IRL and IOP thread a few times, can you suggest any others where I can learn more?



Yes, Nyet, I know my turbo doesn't do squat at idle, it doesn't make positive pressure until about 2.5k RPM. I will change my tables back to stock and only edit last 3 rows of IRL. Then shift axis of IOP accordingly.

check out the me7 tuner wizard in the tuning section.

also, for the sake of your sanity - do not modify any map that has _um after it...  you will permabrick your ecu, checksum tool or not, nobody has the _um maps figured out...

I used me7 tuner wizard to make my current IRL and IOP. As well as interpolating KFZWOP/2.

But I think I get what your saying. When load is <100%, the engine reacts as N/A rather than turbocharged?

There are a few spots where the stock IRL isn't the inverse of the stock IOP

Idle is one of those areas.. I can only assume audi spent some time hand calibrating that area.


A good rule of thumb, but not entirely true. Load can be well under 100% but the turbo(s) might still be doing work..

Understood. It makes a lot of sense now to keep the low load areas stock. I guess I was too hung up on the different turbo sizes.

Thanks!! :D

I would recoment checking ALL the maps THAT SHARE KFMIOP LOAD AXIS... There are quite a few...

I will give that a go when I have some time. I need to locate those maps.

I put IOP and IRL back to stock, and only changed the last 5 % rows of IRL. Then I adjusted the last few points on IOP axis to suit the changes in IRL. Fixed KFZWOP/2 for the changes and flashed the car. Seemed to work fine, but the throttle plate would not open past ~45% when I gave it the beans.

So I plugged my changed IRL into masterj's wizard, and got a new IOP. The lower half of the table is almost stock(by almost I mean off by <1), and the upper half is changed. Flashed that and all is nice now. Car makes boost as normal again, and part throttle is MUCH improved. Though my idle still bounces ever so slightly. I'll have to take a look at the IOP axis shared maps.

Really? Most of mine if not all are changed ie set to 99.1% and Ecu is fine... Think 1 or 2 might set to 0 can't remember but definitely not caused any issues...

But I did leave kfmi_um stock agreed...

Bringing this back up, still have a bit of an unstable idle. Lower half of IOP is stock, as well as IRL except for the last 5 columns, which I have scaled up. The last column now reads ~240.

I do have Deka 630s, and after doing some reading and consulting mr dillenger, I decided to bump up KFMRES/K by 15% at first, then another 10% after. This helped immensely after pushing the clutch when cruising, as the car would sometimes stall before. The car idles great when cold, and doesn't start to lump until it is warmed up.

Any ideas what to look for? Variables to log at idle? Any help is appreciated.

i assume your idle trims are stable throughout the warmup cycle, all the way through normal engine temps?

If not, maybe take a peek at KFFWL and try to get them stable? perhaps wall wetting tables as well?

the dekas just suck IMO. some batches just won't seem to idle right, no matter what you do.

Or, maybe we're all just missing something.

I've had some good results bumping the pressure up with the 630s, seems to smooth things out a bit. If you've seen used dekas on a flowbench, you'll understand they dribble more than they spray. More pressure seems to be a bandaid.

I think the problem is exacerbated by heat as well, which is why they idle WORSE after the motor is warmed up...

Yup, idle trims are within 3%. My next train of thought was indeed wall wetting, as I know these injectors are crap.

dd, I do have a 4bar FPR I can throw in. That was on my list to do before getting new injectors. I think ID1000s or EV14 biguns are my choice.


EDIT: I still think it could be a software issue. It idled fine on my OTS Uni tune that I ran up until a couple weeks ago.

What does wall wetting have to do with idle?

I thought wall wetting sometimes throws closed loop fueling off as the ecu adds/removes fuel, delaying the adjustment's effect and inducing wide swings in trims.

but that is just regurgitating somebody else told me a long time ago. Perhaps it was bs?

Wall wetting was just an idea. Haven't delved into the FR about it yet though.

The idea is to maintain target lambda during load changes.

If wall wetting is causing problems at idle, then the injector calibration is off elsewhere and tmot based changes in wall wetting is exposing it.

Any other ideas besides bumping up fuel pressure? Like I said it idled fine with my OTS Uni tune that I was running until a few weeks ago.