That's the problem I always had with using the graphs. It would build the curve and show the equation, but how can you extract the X & Y values from the curve? I tried doing that a long time ago to no avail.
That's the problem I always had with using the graphs. It would build the curve and show the equation, but how can you extract the X & Y values from the curve? I tried doing that a long time ago to no avail.
OK, got it. Looks like the raw .xls file is the old version, and the archive has the new version.Originally Posted by redhardsupra
Go back to the site, 2.0.1 awaits you
oops, my bad. good catch Jim, it's all fixed now, thanks.
Or y'all could try this approach: Since we most likely will NEVER reach frequencies below 2K and above 10K, you can calculate out the percentage difference b/w your new derived MAF values and whatever ones you were using before at the extremeties where data is missing and basically extend those trends.That's the one area of this whole process that really doesn't fit the Excel way of doing things. But on the other hand, my thinking is that if I never stumbled upon certain values when testing (aka pushing it in all directions) then I should never encounter it during normal operation.
But for these who absolutely _must_ have the complete calibration (I know I do) here are the directions, which are identical to how I do it the 'old way', as described it in my writeup:
1. Plot a XY graph of freq vs dynair
2. Plot a 3rd degree polynomial best fit through it.
3. Display the equation for that best fit line, with maxium precision.
4. Use that to determine the points for which you have not been able to obtain calibration empirically.
This method have been tested twice already, seems to work fine, but don't go crazy believeing the machines
Hope that clears it up, I should probably put that in my writeup now. Ugh, more typing...
IE: let's say as you go closer to 10K you find that your new curve is on average flowing 25% more air. Well then for the missing data above 10K, scale your old values up by 25%. Do this likewise on the low end
This is about the only scaling that should be done since it is difficult to ever sample these ranges. I did 4 30-145 mph runs and could only muster a small handful of 9xxx Hz samples. I'm sure 10K+ can be done but realistically how often do we drive around over 145 mph???
2002 Quicksilver Corvette Z06
And then...
friggin texasOr y'all could try
another super-quick approach to the end values, would be to just simply linearize it; think about it, both ends are quite straight, why not just use few calibration points before/after to throw in a quick linear approximation, right?
i used to have few values in there that i did by hand/eye, just to make it look like it's a part of a bigger whole.
food for thought: if you can't get that much airflow at 145mph, does it mean you're not flowing as well as people that are getting the rare 10kHz+ values?
that may be a possibility but that would only be a comparison at 145 mph... I guess the only way to tell is to line up curves and compare. The thing is though, some of us measure air in 70 degree weather, some in 30 degree, and definitely all of us at different altitudes... That would most certainly play a factor in all this...friggin texas
another super-quick approach to the end values, would be to just simply linearize it; think about it, both ends are quite straight, why not just use few calibration points before/after to throw in a quick linear approximation, right?
i used to have few values in there that i did by hand/eye, just to make it look like it's a part of a bigger whole.
food for thought: if you can't get that much airflow at 145mph, does it mean you're not flowing as well as people that are getting the rare 10kHz+ values?And there's size of MAF housing, then inconsistencies in the port job, descreened or not, ported or not, ram air or non ram air, who's running crappy motor vaccuum... the list could go on!!! Speaking of, send me your maf values! For the inital comparison!!!! Haha we could start a new repository of MAF curves to look at
2002 Quicksilver Corvette Z06
And then...
Aha. I was able to calculate my new MAF table tonight. Of course there is good news and bad news. I am running an SLP lid, and billet MAF ends, w/excess plastic removed from center support. It took me awhile, because my calculations were coming up slightly off. Here is what I did:
1) set MAF fail to 0 hz.
2) logged MAF frequency and dynamic airflow
3) plotted everything into a scatter chart
4) added trendline, poly 3 (perfect, straight through the center of my logged data)
5) calculated new table using the trendline formula, substituting the correct MAF frequency in hz for X.
The result? Poopie, worthless table. That's what stumped me. So I looked over the numbers and realized that the formula was correct. The problem is that by default Excel doesn't factor in that you could be dealing with numbers that require such excessive granularity. So, I reformatted the formula with 6 decimal places. Voila! A perfect MAF table that matched the data I logged.
And that leads me to my problem. I can't fit the values in above something like 10125 hz. They are too big. My table maxes out with values of 511.99, and my last value is as high as 609. So what is the likelyhood that I will exceed 10k hz at WOT at 6400 rpms?
Another note: It is no wonder that people run too lean with MAF ends. With a lid and MAF ends your MAF table is 30+ percent off.
I have attached links to my spreadsheets for your viewing pleasure. Please let me know what you think. DON'T POOP IF your computer takes a long time opening the large spreadsheet, Excel doesn't seem to like poly trendlines with that many values. It will take awhile. The smaller one shows a MAF flow comparison. I don't have trim results yet...too late for today.
Now I need to know if the value limit on my MAF table is a PCM limitation or something that can be corrected in the HP Tuner sofware. Anyone?
http://users.adelphia.net/~someoneelse/MAFSheet.zip
1998 Camaro Z28 M6: SLP Lid, Adj. Panhard, SFCs, CAGS!, Hotcam, Hooker Shorties, Katech Tensioner, Catback with Dynomax 17221 3\" Muffler, Mobil-1, Bosch +4s, 160 Stat, !EGR, !MAFless, !FRA, S3 TB
do the formula with 30 (max) decimal places, sometimes real numbers don't start until like 8th decimal point and go on till 20-25th.
try the binning method in my spreadsheet, at least you'll get to see how few (relatively speaking) values you got for the upper range values.
AND ALWAYS MAKE SURE YOU'RE CONSISTENT ON UNITS! <--that kicked my ass more than once.
You're not kidding about Excel being slow. My dataset has 35000 rows in it. It's vomiting all over itself trying to perform a scatter chart of od 32000 rows - because that's Excel's limit on 2D graphs. So it's not graphing anything above 6935HzWill that be high enough to trendline all the way to 12000?
Can someone assist with the formula - my algebra is a bit foggy. If my formula is
y=(9E-10x^3)-(7E-06x^2)+(0.0279x)-28.216
Do they mean E is e, 2.71828? Or are they using scientific notation to omit the leading zeroes? ???
DO FULL 30 DIGIT PRECISION, you'll see a whole new world!
3E3 is 3000 3E-3 is 0.003
E in scientific location is just n*10^m
Mr MAFless has decided to join the party!I'm on a stock, descreened, CNC ported, F-body MAF, with ram air and lid.
Another answer to another note: This is also why you shouldn't just scale everything up by 5% like alot of people suggest. My MAF was 30% off in many places as well! And 20% is some places.... and 10% in some places...
redhardsupra: So my MAF is all fixed up and I'm happy to report that my trims are perfect. Except when I'm at idle and low speed and rpms... But that's more cam and O2 related than anything... stupid badass cam....
Also, you guys that are wideband tuning the trims, after you get the VE square you will want to re-enable closed loop, and log in SD mode since that's how your car will be operating. The way you can check yourself when your done is go out, reset your trims, and start logging dynamic airflow (lb/min) and mass air flow (lb/min). Drive around and let your trims settle in nice, and when you playback your log, watch the chart display and put your two airflows in the same viewing cell. If you've done everything right, you will see your MAF and dyn airflow charts lying directly on one another. You may need to change the range of units from 0-60 to 0-30 to observe closer. Obviously they're not going to be perfect matches, but you can see now how there will be much less discrepancy b/w SD airmass calcs and MAF calcs. Your MAF will behave like the SD expectation, but be much quicker and more resolute when determining airmass!
2002 Quicksilver Corvette Z06
And then...
OK, I think I got mine to work, in Excel at least. No place to host it, though.
I'm concerned about the formula being correct due to not being able to use rows 32001 through 32632 - which correspond to ~6900-10400Hz. I find it hard to believe I reach 512g/s @ 10375, and 681 @ 11250. I'm running a the FTRA, a paper filter, a whisper lid, and a stock MAF.
Any suggestions on how to selectively remove ~700 rows so I can see the entire trendline formula?
I think I'll wait to see how AnotherUser's results turn out on this test ;D
I cut out enough rows to get under the 32000 row barrier. I dropped 100 at a time from the largest ranges. It changes my curve drastically
1500-2250 ~10g/s, whereas stock and the previous curve were not flat.
I'm done fiddling with this for a few days. I think I'll put the stock MAF curve back in until I see others' results.
Jim,
this is exacly why me and Mr TXhornes281 switched to the 'bins' approach. Trendlines might be easy, but it's hard to gather a lot of data for all ranges in 32k samples.
With my spreadsheet, it's really easy, so there is excuse not to use it anymore.
Give it a whack, you should be pleasantly suprised.
This sucks. Anyways, now the trims are too rich. They are not as stable as before either. It is mostly about 6% rich, but I get weird spike where it wants more or less fuel. I don't know how reliable this is going to be. I am going to pull that new MAF table and log some more data.
1998 Camaro Z28 M6: SLP Lid, Adj. Panhard, SFCs, CAGS!, Hotcam, Hooker Shorties, Katech Tensioner, Catback with Dynomax 17221 3\" Muffler, Mobil-1, Bosch +4s, 160 Stat, !EGR, !MAFless, !FRA, S3 TB
Mine went rich also by about the same amount... and the throttle response has become surgey.... I am going back thru the whole thing to see if I missed something....
2007 CTS-V LS2
Can someone explain to me again why, with a stock MAF, one needs to alter the MAF table? If more air is coming through from pre-MAF mods, the MAF just ends up reporting a higher frequency, no?
The dynamic airflow PID... how does it get calculated?
Well, that should hold true, but most guys modify the induction too; this changes flow characteristics and consequently the maf calibration as well.
98 M6 TA, mods: tsp torquer(233/233 .589 112), Patriot golds, ls6 oil pump, ls6 intake, tsp lid, pacesetter LT\'s and ORY, magnaflow, fra, HPTUNER!
if you are using a stock screened MAF, chances are you probably don't need to recalibrate your MAF, but one easy way to tell, is to log SD dynamic airflow vs MAF airflow. Even the smallest mods can change the airflow requirements for combustion. Just set the MAF fail frequency to zero and log those two things (lb/min). When you play you log back, look at the chart display and set it up to have dyn airflow and MAF airflow in the same box. You will see how far off or not the MAF's calibration is to the SD motor calc of combustable airmass. The MAF was designed by GM for stock LS1 motor. (Duh... lol)
The minute we start modding here and there (especially descreening and porting the MAF), changing the actual airflow required by the engine to make clean combustion at a desired AFR, then the MAF no longer tells the PCM the correct airflow to stay inline with what the motor really wants/needs. You probably are right though about the extra airflow being compensated by higher frequency registering, but when someone like me who's ported their MAF out, you don't really change the air velocity, but more so just air volume. The air velocity is what will determine the frequency and the air velocity all depends on what your motor's vaccuum wants. With that said, a change in air velocity is picked up by the MAF, but air volume is not.
For example, my motor uses approximately 20-30% more air than what my stock MAF calibration tells it. This is a problem b/c the PCM has to decide who to believe and if it's biased more toward the MAF, well you can see where this is going. That's why people run lean after doing MAF mods b/c the MAF calibration does not pick up the extra volume of flow. It tells the PCM, "Hey dude you're gettin X amount of air!" So your PCM delivers the correct fuel calculation based on that air measurement, but you're actually gettin X+? airflow and there's no additional fuel being added for that.
The dynamic airflow, I think is a calculation from the main Volumetric Efficiency calculation. Dynamic Airflow is the airflow that is contributed towards combustion. There's a sticky on LS1Tech: "VE table cracked" that goes into detail about the VE calculation and what the components of it are. One of those components is SD airflow which makes total sense since we tune the VE table in SD mode. If the PCM had no way to calculate aiflow if the MAF failed we'd be in some deep stuff!
2002 Quicksilver Corvette Z06
And then...
JimMueller: You're asking all good questions, if you keep reading into TXHorns' and my own MAF investigations on ls1tech, you'll find all the answers. JimmyBlue has a really nice explanation how MAF gets the values.
MAF is a bitchy, tricky device, so we need to be clever to get it to cooperate with our perpetual need for modding.
My new discovery is how sensitive MAF is to airflow changes in general, not just in the pre-MAF section. I put in LS6 intake on last week, and recalibrated my MAF, and it looks like there's more airflow through the whole range. That's kinda cool, it demonstrates that if something's holding up the airflow even past the airflow measuring device, the whole flow is still affected, not just a section of it. I wonder if heads or cam would change MAF too...hmm...anyone wanna do a camswap really quick?
Reply With Quote