S_J_H

I decided to try tuning mafless on my Blown 98 TA a4.
I have some issues with my maf in it's new location.
First thing I did was disconnect the maf. I then made my first mistake. I turned off all the maf codes entirely. This will make the car start and then die right away. That's because there is no flag to the pcm to run in speed density mode. It starts fine because the pcm uses the cranking VE table to start, but will die a second later when it switches to maf mode once running and there is no maf and no DTC is set to tell it to switch to S.D. mode.
I turned the maf codes back on and just have it not set off the ses light. Now it starts and runs fine without the maf connected.
Since I'm using a stock cam and stock cubes, the factory tables are very close until it gets into boost. I multiplied the entire table by 107% up to map cell 85. This makes the L-trims near perfect. With a 10 mile cruise they averaged out at 0 and -2% for each bank. Of course some cells are rich and some are leaner to get at that average. But that can be tuned perfectly with more time. But right now it's well within the pcm's correction range.
VE Cells 90-105, I initially set very rich so to be able to tune from rich to lean under boost.
On my first programming try , giving it enough throttle to get into map 90+ made it run lower than 9-1 A-F and it just bogged. I then dropped those cells by 25%. Much better but still not there.
Still needs a lot more seat time to get it right but I can already give some feedback.
Absolutely no surging or start or idle issues. It actually starts and idles cleaner than with the maf! And I do not even have a idle bypass motor on my car.
There is really no need for a 2 bar map sensor with a belt driven supercharged car.
Boost is a function of throttle position, and rpm of the blower and not load.
This looks to be a very powerful tuning alternative to the maf method.
You have the VE table, fuel multiplier table based on temp for those cells, PE table and at what throttle % to bring in the PE mode table.
This gives you more tuning power than with the maf. And especially if the maf system is not doing all that well.
Oh, and it does run in closed loop and adjusts the L-trims in speed density mode as well. This is not some limp home mode. This is a full functiong S.D. mode.
I am using efilive to log everything including a wideband sensor output.
My night of tuning just got cut short as Just I blew off the mufflex flowmaster muffler out of the intermediate pipe.
But so far it looks very promising!
Steve

XLR8NSS

iTrader: (4)

So the computer still compensates for different air temps based on this table?

I guess the MAF is really just on these cars as a "correction factor". It seems it's absolutely not needed if you have the ability to tune in your VE tables and adjust some of the other tables properly.

Does unplugging the MAF have any effect on how the transmission shifts?

Keep us updated on your progress Steve.

MYTURBOT/A

I mentioned doing this about a year ago cause I read of a few Aussie cars doing it and getting great results but everyone said it couldn't be done. Glad to see your having good luck so far I knew somone with good knowledge of the VE tables would be able to figure this out eventually.

S_J_H

The thing is for me, the maf sensor has become THE PROBLEM.
it's sensitive to location,reversion,turbulence, range limited and size limited.
I could move the maf sensor 12" and have the entire flow table now be garbage.
Not exactly precision tuning when you have to base everything around the maf sensor which is what I have been doing as of late.
The trans is shifting the same and I can't tell any difference in S.D. mode.
I'm not sure on the air temps. There is the fuel air multiplier table but I don't know if that's coolant or air temp.
You have full 02 feedback though to keep your cruise A-F's constant and I have a few ideas for external tuning without always running for the laptop.

I read all the VE tuning threads. I'm not a math wiz and won't even take a shot at it. I'm using the formula from those threads of 4096*.708*VE% to figure a baseline number for the cells. Then just go out and log the wideband and make changes until my commanded AF is close to the wideband output.

Pro Stock John

iTrader: (34)

Would a 2 bar map sensor be useful to you Steve, I have one somewhere in the garage.

S_J_H

sure, I would like to try it out John.

Pro Stock John

iTrader: (34)

I gotta find it.

Stealth is selling that TSI (Australian box) in the classifieds... FWIW.

deezel

Pretty sure the fuel air multiplier table is based on coolant temps in degrees C.

Thanks for posting your progress. Please keep us updated.

Bink

Your O2s are updating the LTFTs through the PCM???
joel

S_J_H

They sure are. I have an in car wideband as well and I log it via efilive through the AC pressure switch hookup. At cruise the motor tracks dead steady at ~14.7-1 in closed loop just like it did with the maf connected. L-trims adjust just like they always did from o2 feedback.
I had very little time tonight to play with it. But right now after only a few tries I will NEVER go back to the maf mode. Throttle repsonse is crisp and clean. Startup into idle is fantastic. I'm running 12psi boost on this thing so I have to be cautious. I'm also using a 9th mechanical injector in the mix so I have 3 areas of tuning involved at WOT. The Main VE table, the PE table, and adjustment of the 9th injector flow which is only used at WOT.
And the 1 bar map sensor is proving to be totally sufficient.
Steve

Magnus

iTrader: (4)

Sounds like a very complicated setup.

I started my SD tune about a month or so ago but haven't had the time to really "tune" it.. Tonight I spent some time adjusting the VE table. I looked through many bins and its amazing how much different the VE tables are for similar vehicles from GM... Big difference in Holden vehicles vs the US vehicles as well for almost the same engines.

Once I nail down my VE table I plan to run open loop permanantly. Thus, when you run open headers you don't have to worry about contamination from having O2's at the edge of the outlets.

Bink

Sounds great!! Thanks for the reply. Keep us updated.

joel

S_J_H

I think where some are going wrong and making it overly tedious to tune, is trying to use the maf flow to figure a good VE table. Unless the maf is calibrated perfectly I think it's a better idea to just disconnect it and begin tuning in pure SD mode.



I don't think you need a load dyno to do this. You just need plenty of seat time with a logged wideband as well as EFILive or similar logging software.

The maf flow output is greatly smoothed out and that's going to create problems when using it's output to tune in SD. My maf output will jump up to 10lbs/min in output from frame to frame at WOT. That's about 100hp worth of fueling variance if the pcm used the output without smoothing! The pcm smoothes it out and won't deliver fuel directly from maf output. If it did, these things would run horrible. But that underlines the point why you should not use maf data points for tuning.



I need to test the IAT function. Once I get everything dialed in I'll get 2 bowls of water. One hot and one ice cold. I'll submerge the iat sensor in each and observe injector PW for changes. Right now I'm not sure if the pcm is using IAT temp in the equation or just using O2 feedback. I'm hoping it does use IAT output in the fueling or else WOT tuning will need to be changed with weather changes. For me that's not a problem as I'll simply make nozzle size changes to my 9th injector. But I need to know for sure if IAT is used in the SD equation as there is no ls1edit table for it.

S_J_H

Well my initial joy with this S.D. mode is begining to turn sour.
I found several lean and rich holes with the Main VE tables on my combo. All below 2000 rpm. If you look at the MAIN VE tables for a 98 you can see under 2000 rpm they vary greatly in their curve shapes. This I can fix.

The trans is now locked into full pressure mode. It took a few key cycles before it did that. That is a problem.

My biggest concern right now is weather variances and adaptability. Since this thing is supercharged, large changes in weather really make a huge difference in the engines output. The difference on a hot day with the blower and plumbing heat soaked vs a cold day with a cold blower is almost 100hp.
This could be a major problem with this mode unless I am willing to reprogram constantly.

I'm now having my doubts this can be done without a 2 bar map sensor to help compensate.

Magnus

iTrader: (4)

Yes, SD will command the transmission to full line pressure. This is not an issue on vacuum moduldated transmissions.

Pro Stock John

iTrader: (34)

I'll look for that 2 bar map it disappeared but I'm sure I still have it.

Bink

Auto Tranny dummy here. Can someone explain Magnus' statement?

joel

HumpinSS

iTrader: (1)

I guess the manual guy would benifit from this.

Joel

driving around in normal conditions the PCM isnt applying full pressure to the clutch packs in the tranny. It uses the MAF to figure this out. Without the MAF it sends full pressure to the clutch packs and could burn out the pump under these conditions. This is my understanding hopefully I didnt spread any misinformation. WOnt be the first and definately wont be the last :yup:. A vacuum modulated tranny would vary the line pressure to the clutch packs using a acum signal like a vaccum advance distributor


HTH

XLR8NSS

iTrader: (4)

Those were my two biggest concerns with SD mode.

kp

iTrader: (34)

I have a 2 bar map sensor here I was using to log boost with efilive but now that I have this fancy recall guage you can have it if PSJ cant find his steve. I even had a new pigtail somewhere for it but I would have to dig for that..