tici

iTrader: (1)

PLS dont consider my sig: right now the supercharger is disconnected and the car virtually stock.

Just to make a test I disconnected the MAF and copied the high octane timing values in the low octane table, so now I'm in SD.
The fuel trims are all +/- 3% and the AFR at WOT is around 12.5 (as commanded).

The car feels more "alive", stronger and the engine revs faster.
A real fun to drive.

It's not a rocket but I can feel the difference.

Why? What is the advantage to run SD with a stock engine?

Phil99vette

iTrader: (11)

I don't know the answer but I think the VE lookup gets you the correct amount of fuel faster than a MAF setup. I think there is some delay in a MAF due to how it works.
Phil

jimmyblue

iTrader: (11)

The MAF responds in "thermal time" because it has to bring
the hot wires back to hot, after the air steps up. The MAP
sensor can be quicker. Further I think there is more filtering
applied to the MAF to settle down frequency-jitter and make
the Dynamic Airflow.

It would be interesting to log Dynamic Airflow, MAP and MAF
frequency & airflow, across rolling nail-the-throttle events,
and see how the Dynamic Airflow response (risetime & delay)
differs between the two modes. For that matter log the RPM
and injector PW as well. That would be some neat data to
crunch on.

RedHardSupra

iTrader: (1)

If your SD works better than MAF, this just means your MAF isn't calibrated properly.
I believe in recalibrating MAF for any changes, especially in the intake tract. This is also where I disagreed with jimmyblue in the past. He believes that MAF is a calibrated device and screwing with that is more harm than help. While for other sensors i'd completely agree with this logic, when it comes to MAF, i think it's a different case.
I've talked about it with some people that do this stuff for OEM applications, and they told me that the MAF is calibrated for the stock intake system. Any changes there, will throw it off. Since the first thing any LSx owner does is put a lid/filter on it, you already need a MAF recalibration. Also, if you think about how MAF works, it makes me think that the MAF calibration is a good thing--realize you're measuring a flow of a pipe with a tiny sensor in one spot of it. Flow in pipes isn't uniform, so if the flow changes even a bit (think different lid), the airflow at this one spot where MAF has the hot element will probably change. At which moment you're measuring with a miscalibrated sensor.

tici, what you want to do is log dynamic airflow and MAF airflow. i'll bet you they dont track exactly. if they did, your car would perform exactly the same in both MAF and SD modes.

tici

iTrader: (1)

Already done: MAF and MAP react together, no delay (however, that was an AutoTap scan, maybe too slow...)

Here the reason I started to look into the SD stuff:
my fuel tuning works fine as long I drive normally and I'm not stuck in traffic in a summer day. But as soon I get nervous on some mountain roads os overheat the engine in traffic I notice how the fuel trims get allover the place (like -8% at idle and so on).
At that point I can turn off the engine, turn it on again and the fuel trims get back to the old values ( = +/- 2%).
When I switch to SD I see exactly the "allovertheplace" values
With some fantasy I can assume that the PCM switches from MAF to MAP in certain particular conditions, maybe when the MAF values are too far away from what is stored in the VE table

Right now I'm collecting data to redisign the VE table. My goal: reach +/- 2% fuel correction in SD and then recalibrate the MAF.
So far it looks like the MAF is out by 4%, is that possible? The car is a 98 and has 70'000 miles on it... time for a new MAF?

Here the Stock 98 VE table and the result from the scan I made yesterday

turbolx

This is precisely right. Changes in the flow into/out of the MAF sensor will skew its reading. Correcting the transfer function to indicate the actual (new) mass flow rate at a given output frequency often smooths out a lot of the errors that result from cold air kits, modified tubes, ported sensor housings, etc...

patSS/00

iTrader: (6)

I think jimmyblue has it exactly right, changing current flow thru heated wires is not a super-fast way to measure changes in airflow. It would be interesting to do a simple experiment: place a MAF by itself in front of a fan, hook it up with +5/+12 (don't remember the exact wiring), hook up a multimeter to measure the output, then look at how fast the output voltage changes when you block the airflow with a piece of cardboard.

jimmyblue

iTrader: (11)

Your throttle transients (unsteady MAP) pop you out of the
blended airflow into straight speed density so you might just
be stumbling over the discrepancy in fuel and/or spark from
the airflow difference. Since you see big differences in the
trims between modes, this may be the root of it.

RedHardSupra

iTrader: (1)

yup, this is exactly the reason why i dont believe in tuning MAF without VE first--because of our dual/hybrid mode of operation, if the airflow from MAF and SD modes dont agree, computer is very eager to go into panic mode. there's even tables that compare the two as a reality check if something's out of whack.

turbolx, i had all these ideas long time ago, but couldn't find the right words, the MAF thread on CC really made it all clear for me, and that's all your work--thank you, and possibly see ya in May in TX

patSS: it's a good experiement to do, but I think i've read somewhere that the hot element is so tiny and so sensitive that MAF readings can very quickly adjust to the real temperature, we're talking milliseconds here. i'd really like to see the experiment done tho, just stick the measuring setup into a fridge for a moment and see how quickly it takes for it to adjust.

12secSS

iTrader: (30)

I agree with Greg and Marcin on this, based on my experience. If you have any fueling imperfections, like spiking lean upon slamming the throttle open, you need to look at fuel transient tables to correct it. Otherwise SD vs MAF equals no difference in driveability nor power ... if both are tune correctly and within (say) +/-1% of your target AFR. I have logged Dynamic Airflow and MAF Airflow, when layed on top of each other they become one.

ChrisS@sdpc

I have a question regarding this test, were the spark tables different between SD and MAF? If I understood correctly, you copied the High table to the Low table for your SD test? If so, one thing that would be interesting to see is log of timing vs RPM and airflow particularly in the areas that you feel the car is running stronger.

Many of you have probably alread seen this, but here is good technical paper discussing air metering capabilities of a MAF.

SSpdDmon

iTrader: (10)

The idea here was to actually keep the same timing. In SD, the car runs off of the low octane timing table, right? Assuming he had a knock learn factor/multiplier of 1 while the MAF was still enabled, he would have been running off of the high octane timing table. So, copying the high timing to the low maintains the same spark when switching to SD excluding any timing adders/subtracters for IAT and what not.

ChrisS@sdpc

I understand the theory, just really looking for confirmation. I have seen way too many posts indicating SD had much better response, power, etc. only to find out that during the test people were modifying the spark advance. So, I would just hope that the test could be done under reasonably similar conditions, and that spark values for RPM and MAP could be compared. Just my $.02 (which may only be worth a penny).

SSpdDmon

iTrader: (10)

Gotcha... Seems like a worthwhile argument, so I'll throw in the other penny.

tici

iTrader: (1)

Agree with jimmyblue about "jumping" in SD mode while accelerating, but why does it happen only while I'm speeding or I'm stuck in traffic? I understand while speeding, but in traffic things happen pretty slow...

While in SD I checked the real timing and I saw the values stored in the high octane timing table. Once you disconnect the MAF or set the failure frequency to zero the PCM will switch to the low octane timing table (I tried), so the only way to go is to copy the high octane values in the low octane table.

OK, I checked the new VE table and it works fine (fuel trims very close to zero allover the place, AFR @ WOT pretty close to the commanded values). Tomorrow morning I'll reconnect the MAF and I'll see what happens...

tici

iTrader: (1)

I understand and agree with this argument. But look at the pics I posted: the 98 stock VE table and what comes out of a scan (smoothed and so on). They ARE a LOT different
If it was that important why would GM load such a crappy VE table?

joecar

Because it's GM.

tici

iTrader: (1)

Don't tell me they don't have a test engine and a calibrated MAF

They also load different VE tables for F-Bodies and Y-bosies even if they are the same (compare for example the 97 - 98 models: the cams are a little different, but not THAT different )

OK, I reconnected the MAF and things look not too bad at low and middle load.
I just don't know how to determinate the right MAF curve while in PE... any idea?

ringram

Its a best guess attempt designed to work under the widest and loosest conditions. Backed up by closed loop o2 sensors and SD mode. Its not designed as a perfect fueling system. Its designed to be reliable, robust and emissions freindly.

If you want something better then tuning tools are your friends.

SSpdDmon

iTrader: (10)

As long as you're not maxing out the MAF, you log commanded AFR, WBO2 AFR, and MAF Hz. (WBO2/Comm.=%Correction) for each MAF Hz. Notice the trend and smooth out the curve accordingly. That's how I do it at least.