New ATR is out Jan 25 2012 build: 1.9.1.0-3289

[rfinkle2]

reading up on speed density asap.

[Tokay444]

We have a map sensor on the manifold don't we?

[Dano]

Originally Posted by djuosnteisn Yup. While i totally see the application and potential usefulness for the VE table..... i'm still not 100% convinced on reason for it's existence. The whole purpose of a MAF sensor is to account for airflow. If VE falls off for some reason or another (like higher rpms, or inefficient turbo, etc etc etc)... airflow drops, and the maf completely accounts for it. Lex had a decent idea suggesting it may have been used to account for air lost during over lap..... but i think even that's a hell of a stretch, hahaha. But i'm not going to complain. Hopefully it sets the stage for potentially better uses down the road... like speed density. I heard Cobb had a similar release on the gtr shortly before they started supporting SD on that platform. So who knows... maybe they're just starting to get their ducks in row.

lol already posed the question to David as to when we would get SD like the scoobie guys...for those who don't know IIRC it took cobb over a year after the announcement of its coming to make it to fruition over there. NASIOC has a thread covering it.


anywho...he gave me an answer... something along the lines of "there are a few things we need to get into place before we can get to SD".....hum...two weeks after my question we get a VE table....coincidence? You be the judge.

[f-castrillo]

Originally Posted by djuosnteisn Make sure your "use boost based" toggle is enabled (checked), and all your load tables are set up accordingly. Then set your "max boost throttle close" table to like 24 or something for good merit. Should fix your issue.

I've tried enabling and disabling the "Use Boost Based Tuning" toggle, and I had my boost limit throttle close table set to 23 or 24 at the time. I even went back and forth with Braden @ COBB via PM, and we never got to figure it out; that's when he suggested I play with boost targeting instead

Originally Posted by atvfreek 2.24 sounds like an unachievable load at 5k on stock turbo. Every stock turbo car Ive played with are at 2ish for load. Was your wgdc maxed or close to maxed (above 80%)? Tappin

Not sure if I screwed up along the way (I don't think I did), but at the time of my load tuning attempt, I saw calc loads of up to 2.36, and it would hold well above 2.00 up to 5500 IIRC. Nowadays I barely crest 2.05, I think. My WGDC was nowhere near maxxed either, it was still well within my WGDC ramp limiter range, but at some point, the logs indicated that the ECU stopped adding WGDC completely, and there was no throttle close either.

Originally Posted by djuosnteisn Agreed fully. But... The load target greatly depends on the maf cal. If his maf cal is just wildly off, and he's compensating by targeting overly lean afrs... may still be feasible. But yeah, f-cast.... make sure your load targets are even feasible first, haha.

My MAF cal is still within +/- 3%, even as far back as my load tuning attempt. As for my load targets, my ECU was able to achieve high calc. loads, so I targeted slightly lower than what I was seeing, just to be conservative.

I'm kinda happy with where my tune is at ATM anyways, so I won't bother with the load tuning for now. I'll be watching the VE discussions though

[Tomas]

Originally Posted by Tokay444 We have a map sensor on the manifold don't we?

Yes. And it also includes BAT sensor. Its all in 1.

[phate]

So, VE tables didn't do anything we hoped for. I'm too many beers in at this point to post logs, but I'll start a thread tomorrow.

[Tokay444]

Originally Posted by Tomas Yes. And it also includes BAT sensor. Its all in 1.

So what would be the purpose of putting one on in the maf housing like Dj eluded to?
With speed density, would we not be able to eliminate the intake all together and just run a filter or a screen on the turbo for a better pressure ratio?

[atvfreek]

Originally Posted by Tokay444 So what would be the purpose of putting one on in the maf housing like Dj eluded to? With speed density, would we not be able to eliminate the intake all together and just run a filter or a screen on the turbo for a better pressure ratio?

Having a map/baro sensor before and after the throttle plate is better for load/boost calculation. It measures the difference on each side of the throttle plate. This is how the Ford sho ecoboost does it without a maf sensor.

Tappin

[cld12pk2go]

Originally Posted by Tokay444 So what would be the purpose of putting one on in the maf housing like Dj eluded to? With speed density, would we not be able to eliminate the intake all together and just run a filter or a screen on the turbo for a better pressure ratio?

Traditional speed density only has a MAP sensor in the intake manifold. It HAS to be after the throttle body or else fueling will be totally screwed.

This is how my Mustang is setup...

I have no knowledge of how the Ecoboost is using the second one.

[superskaterxes]

still not sure why you would want to go SD over MAF.

if the ECU is so heavily based on the MAF why would u want to try and work around that? no one has reached a state were there is no more resolution for the MAF (unlike the subi guys who have some 300g/s limit).

its like the whole boost tuning logic. the ECU is load based and half the tables are scaled using a load axis so why would u want to brutalize your tune just to hit a certain PSI level?

hopefully someone can show me the way of SD!

[rfinkle2]

I read some about speed density last night, and was hoping to get some opinions on it from you guys, even if we may not use it in the future.

[Ziggo]

Originally Posted by phate So, VE tables didn't do anything we hoped for. I'm too many beers in at this point to post logs, but I'll start a thread tomorrow.

Tag me please, I am interested in the results.


Zigatapatalka

[Ziggo]

Originally Posted by rfinkle2 I read some about speed density last night, and was hoping to get some opinions on it from you guys, even if we may not use it in the future.

The lazy\ricer reasons:
Boost\vacuum leaks don't affect fueling
BOVs

Better reasons:
MAF sensor is an intake restriction
MAF sensor requires a straight section of piping(better flexibility with intake piping)
Better part throttle response, particularly when compared to BIG maf cars. Allows massive intakes with little to no sacrifice in drivability, think 4" all the way (remember when the flow in the intake goes sub-turbulent, the MAF sensor readings become erratic)

The bad:
Think MAF cals are a pain in the ass? Try calibrating a 2d VE table, which will have to be touched for ANY hardware change after the throttle.
Our valve gunk will make VE table change over time.
VE table is not particularly stable over temperature, as turbo becomes more efficient at lower temperatures, lower exhaust mani pressure makes VE improve.

The Megasquirt has a function where you can run MAF below a certain threshold, and speed density above it, or vice versa. If you go this route, you could say, run a standard size MAF so drivability doesn't suffer, and before the MAF maxes out, switch to speed density. You gain headroom in that fashion, but IMO if you are maxing the stock MAF, you probably would be better off with a bigger intake anyway.

If you go the opposite direction it's racecar status and you use speed density for low load situations, but use MAF once there is enough flow to make the sensor accurate, so you get accurate, stable fueling at high throttle/WOT operation. You can really skimp on the VE table this way because the speed density area of operation is all vacuum, so pressure changes won't change VE very much, and who cares if it idles /off idles like shit, it's a racecar.

*All above information is solely the opinions of the author and does not reflect the ideas or opinions of his employer or the producers of this website*



Zigatapatalka

[rfinkle2]

Thanks Ziggo.

It seems like there is pretty heated debate and it was difficult to find factual info. on the pros and cons of one vs. another, but your explanation explained it exactly as I had hoped.

/secretly hoping not to have to delve into speed density @ this point

[Ziggo]

Originally Posted by rfinkle2 Thanks Ziggo. It seems like there is pretty heated debate and it was difficult to find factual info. on the pros and cons of one vs. another, but your explanation explained it exactly as I had hoped. /secretly hoping not to have to delve into speed density @ this point

Personally I wouldn't touch it with a 10ft pole while wearing level 4 mopps


Zigatapatalka

[Nataphen]

@phate
Please tag me when you start that thread as well. I'm barely on here anymore except for subbed threads, so I probably won't remember to look.
@Ziggo
Don't ever change, LOL.

[Dano]

Originally Posted by superskaterxes still not sure why you would want to go SD over MAF. if the ECU is so heavily based on the MAF why would u want to try and work around that? no one has reached a state were there is no more resolution for the MAF (unlike the subi guys who have some 300g/s limit). its like the whole boost tuning logic. the ECU is load based and half the tables are scaled using a load axis so why would u want to brutalize your tune just to hit a certain PSI level? hopefully someone can show me the way of SD!

afaik the 300g/s cap is gone in the scoobie world...so IDK why they are so enamored with the thought of running SD....maybe its only the ricecars that want it so they can VTA and not worry about it.

from my discussions with Lex and what Ziggo has said here it sounds like shit to me.

[rfinkle2]

Originally Posted by Dano afaik the 300g/s cap is gone in the scoobie world...so IDK why they are so enamored with the thought of running SD....maybe its only the ricecars that want it so they can VTA and not worry about it. from my discussions with Lex and what Ziggo has said here it sounds like shit to me.

SD = Shit Density? Had to do it.

[djuosnteisn]

Originally Posted by Tokay444 So what would be the purpose of putting one on in the maf housing like Dj eluded to? With speed density, would we not be able to eliminate the intake all together and just run a filter or a screen on the turbo for a better pressure ratio?

It was a joke.


The VE table is load dependent, and the load source is the maf sensor.

For SD to work, you need 3 main ingredients, RPM (obv), MAP and manifold temp.

If our newly exposed VE table was MAP dependent, then no worries, it's ready for SD prime time right away (or at least playing around with it)...... BUT because it's load dependent (read MAF dependent).... it's not so easy.


So i was kidding around about using a MAP sensor, and wiring it into the MAF wire harness.... so that the MAF input is now MAP. And then the VE table would be MAP referenced.







But it was 99.9% a joke. Unless someone really wants to calculate out which loads each MAF voltage / rpm point corresponds to.... and then use that to create a VE table... and re-configure all their load dependent tables......

so yeah..... 99.9% a joke, lol.

[Tokay444]

I hais teh stupids!

[rfinkle2]

Originally Posted by djuosnteisn It was a joke. The VE table is load dependent, and the load source is the maf sensor. For SD to work, you need 3 main ingredients, RPM (obv), MAP and manifold temp. If our newly exposed VE table was MAP dependent, then no worries, it's ready for SD prime time right away (or at least playing around with it)...... BUT because it's load dependent (read MAF dependent).... it's not so easy. So i was kidding around about using a MAP sensor, and wiring it into the MAF wire harness.... so that the MAF input is now MAP. And then the VE table would be MAP referenced. But it was 99.9% a joke. Unless someone really wants to calculate out which loads each MAF voltage / rpm point corresponds to.... and then use that to create a VE table... and re-configure all their load dependent tables...... so yeah..... 99.9% a joke, lol.

I think I'll join in and laugh after I read that about 10x's.

[Dano]

hard enough to get the ECU to target boost consistently....SD would have to be next to impossible

does anyone here know why the scoobie guys are so excited about it?

[phate]

Volumetric Efficiency table discussion:

Volumetric Efficiency Table and You

[Ziggo]

Originally Posted by rfinkle2 I think I'll join in and laugh after I read that about 10x's.

Since the VE table is load referenced: you can't use MAP to calculate load to lookup VE to calculate load. Circular references mean nobody gets anywhere.

Supposing I were hacking the ECU and wanted to implement speed density, my function would halt the MAF lookup from the analog IO and instead inject a MAF value into the register that lookup was writing to. The injected MAF value would be calculated from the MAP, RPM and BATs. That way any function that was reading that register would continue to operate normally with a synthetic, speed density based MAF value.

The question is, which do you think is more accurate and faster::

MAF sensor voltage reading analog IO read
1D maf table lookup based on voltage to obtain MAF table values at high and low side of reading
1D linear interpolation between high and low values from table
Write interpolated value to register for "currernt" MAF

OR:

Read MAP value
Read RPM value
2D VE table lookup for high/low side values on either side of MAP and RPM
1D interpolation for VE based on MAP at low RPM
1D interpolation for VE based on MAP at high RPM
1D interpolation of two previous values for VE based on RPM
Calculate MAF for 2.3L @ RPM based on 100%VE @ MAP and BAT
Correct airflow based on inerpolated VE
Write airflow to MAF

[rfinkle2]

Originally Posted by Ziggo Since the VE table is load referenced: you can't use MAP to calculate load to lookup VE to calculate load. Circular references mean nobody gets anywhere. Supposing I were hacking the ECU and wanted to implement speed density, my function would halt the MAF lookup from the analog IO and instead inject a MAF value into the register that lookup was writing to. The injected MAF value would be calculated from the MAP, RPM and BATs. That way any function that was reading that register would continue to operate normally with a synthetic, speed density based MAF value. The question is, which do you think is more accurate and faster:: MAF sensor voltage reading analog IO read 1D maf table lookup based on voltage to obtain MAF table values at high and low side of reading 1D linear interpolation between high and low values from table Write interpolated value to register for "currernt" MAF OR: Read MAP value Read RPM value 2D VE table lookup for high/low side values on either side of MAP and RPM 1D interpolation for VE based on MAP at low RPM 1D interpolation for VE based on MAP at high RPM 1D interpolation of two previous values for VE based on RPM Calculate MAF for 2.3L @ RPM based on 100%VE @ MAP and BAT Correct airflow based on inerpolated VE Write airflow to MAF

I'm a good student of Ziggology, and would choose the 1st of the 2 methods (.5 because of the tone in which the question was written fml, and half out of wanting the ecu to do as little interpolation as possible, although I'm certainly no engineer).

Thanks for explaining the majority of the differences in SD vs. Maf based calibration.

I was reading many opinions last night, and came to the conclusion that I wonder why an oem like Ford would use SD as a factory calibration.

I don't want you to spend your time dumming things down for me Ziggo (even though I need it lol), although I appreciated the time you've already spent.

Thanks for bothering to type out what would have taken me hours of reading through opinions on which is better and trying to pull out what are the facts.

[Tokay444]

The current quickest ms6 on the planet is SD tuned.

[Dano]

doesn't make it the easiest to tune or drive...

jussayin

[rfinkle2]

Originally Posted by Tokay444 The current quickest ms6 on the planet is SD tuned.

Paveltol's? Did you and Dano see his injector seal thread yet?

[Tokay444]

Mazdagaragepat.
Yes, I saw that.
I can't get a picture of the part though.
Every page I find that part number on has a different picture.

[rfinkle2]

Originally Posted by Tokay444 Mazdagaragepat. Yes, I saw that. I can't get a picture of the part though. Every page I find that part number on has a different picture.

I chose the wrong country. LOL

[Dano]

Originally Posted by rfinkle2 Paveltol's? Did you and Dano see his injector seal thread yet?

no PM me the link pls

[FreeFlyFreak]

Originally Posted by djuosnteisn i'm still not 100% convinced on reason for it's existence. The whole purpose of a MAF sensor is to account for airflow. If VE falls off for some reason or another (like higher rpms, or inefficient turbo, etc etc etc)... airflow drops, and the maf completely accounts for it.

How about this as a reason for existence.......... as a back up.
In case the MAF sensor fails?
Or maybe to detect a failing MAF sensor and throw a trouble code?

[Fatguy729]

or maybe there is a point that the SD/VE over rules our standard ECU logic? or at least a factor at some point for some reason? like AFR corrections for different fuel mixtures? Just throwing shit out there...

[Celestspeed3]

Originally Posted by rfinkle2 I was reading many opinions last night, and came to the conclusion that I wonder why an oem like Ford would use SD as a factory calibration.

It's probably because its cheaper to manufacture. Cost of a MAF vs MAP sensor over xx vehicles is a pretty good savings.