View Single Post
      12-28-2010, 12:29 AM   #28
usc335
First Lieutenant
United_States
62
Rep
380
Posts

Drives: '21 M2C, '08 E92 335
Join Date: May 2008
Location: so cal

iTrader: (1)

Please add the following information to the users guide:
----------------------------------
(1) General Command Center Descripition:
http://www.e90post.com/forums/showthread.php?t=427976
----------------------------------
(2) Boost Response:
http://www.e90post.com/forums/showthread.php?t=435393
----------------------------------
(3) Autotuning Basics:
http://www.e90post.com/forums/showthread.php?t=419045
----------------------------------
(3a) Autotuning Logic Overview (Shiv wrote this in some post somewhere):
AUTOTUNING BOOST & IGNITION CORRECTION EXPLAINED:
Traditional tuning involves finding the right compromise of ignition advance and boost. Basic engine theory suggests that you want to have maximum air/fuel charge expansion (during combustion stroke) at approx 14deg after TDC. So that involves initiating the spark at the right time (correct advance value). However, in the case of a turbo engine running on pump gas, the ideal spark advance isn't possible due to octane limitations (ie, running the "correct" advance will cause detonation).

Another way to look at it is recognizing that turbo engine tuning is all about managing the trade off between thermal efficiency (timing) and volumetric efficiency (boost). You can have one very high and the other very low and get poor results/low power. The key is to get them both in their respective sweet spots. In doing so, you get the most power with the greatest safety margin.

In the case of Procede AutoTuning, ignition timing changes are first applied to see how efficienty the engine is burning. The Procede calculates this by seeing what timing values are needed to achieve a knock-free combustion. If the max-load timing value is deemed too low (inefficient combustion), boost is gradually lowered. Conversely, if the engine is running a high knock-free ignition advance value, the Procede attempts to bump up boost (and raise volumetric efficiency) and make more power. While under boost, the Procede is always monitoring the ignition trims that the DME is applying. And it bases its decision on their dynamics. Same stuff I would do when custom tuning a car for any given condition. But the Procede is always doing it.

The AutoTune learning conditions have changed. There is no lower RPM limit however. It will activate at any time, regardless of RPM. But there are several over conditions that need to get satisfied. And there are a bunch of "dynamic" conditions implemented as well. Which means that it will not learn to run higher boost (at higher engine speeds) until the engine actually sees and "tests" those higher engine speeds. So simply doing at 3000-4000 pulls on, say race gas, will not get it to ramp up boost to high boost levels. Instead, the car will actually have to drive through the engine speeds that have the higher boost targets. It's exceptionally cool and way more advanced than the ignition correction-only AutoTune logic we had introduced earlier. It needed to be since boost was not in the mix.
----------------------------------
(4) AFR Display Info:
This value displays on the oil temp gauge (when this display mode is selected). The minimum reading indicates 20:1. The maximum is 10:1. Which means the middle reading is 15:1 or approx stoich. Very simple and self-explanatory. This gauge displays the leaner of either bank1 or bank2 AFR readings. So it's very helpful in diagnosing a bad or flaky o2 sensor. Or a failing fuel pump. Or, in the case of upgraded turbos operating at high boost, running beyond the limits of the stock fuel system.

Question: Why does AFR appear to go lean sometimes when transitioning from cruise to full throttle?
Answer: Those lean spikes are caused by the voltage bias the procede applies to the wideband sensors. It is, in effect, telling the DME that the engine is operating leaner than it really is. Which makes the DME add the desired fuel enrichment. The period between this perceived condition and the DME response can take a second in some cars/software with certain DME software versions. You'll see this effect most obvious in the 3500-4500 rpm range where we apply the biggest bias/fuel correction. If you put an actual wideband on the car, you wouldn't see the same lean spike at boost onset. Hope that helps clear things up.

----------------------------------
Thanks!

Last edited by usc335; 12-28-2010 at 12:45 AM..
Appreciate 0