Pelican Parts
Parts Catalog Accessories Catalog How To Articles Tech Forums
Call Pelican Parts at 888-280-7799
Shopping Cart Cart | Project List | Order Status | Help



Go Back   Pelican Parts Forums > Porsche Forums > 911 / 930 Turbo & Super Charging Forum


Reply
 
LinkBack Thread Tools Rating: Thread Rating: 1 votes, 5.00 average.
Author
Thread Post New Thread    Reply
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Quote:
Originally Posted by Speedy Squirrel View Post
There are very few new gasoline vehicles running with MAF only. The range of airflow is too difficult to measure with just one sensor. This is especially true of a turbo motor, which acts like a small engine at idle, and a large one at WOT. It is common to use a MAF to control idle and part throttle only, and switch to speed/density at large throttle/boost. The speed/density measurement has become quite accurate when throttle adaption, and VE learning from the WRO2 are factored in.
No disrespect here just my thoughts on the topic of boost.
An engine does not know if it's at boost it only knows how much mass of air it has ingested. I've helped do a SuperCharged and a Turbo 3.2 running just MAF in a blow through setup with a lab grade MAF designed for this. No rising rate FPRs and no fancy MAP based ignition retard. Simply stock Motronic with MAF and a very custom chip I developed. The math for calculating injector pulse width is very simple if you have a calibrated MAF you can trust. Math looks like this:

GivenAirFlow / RPM = AirFlowPerRev (we now know how much air is ingested per rev)

A 3.2L can at a VE of 1 can ingest about 1.9grams of air in 1 rev
Here's this math approximated. 1Liter of air at 68F sea level is equal to about 1.2 grams and a 3.2L motor takes in 3.2L every 2 revs at WOT assuming a VE=1 so the 1.9grams works out like this:
3.2L / 2revs * 1.2 = 1.9 grams.

We now know that at a VE=1 this motor can ingest 1.9grams of air.
Anything below 1.9g is below VE=1 and typically the only way to get above 1.9g is with some sort of forced induction. But the motor does not know it's being forced to take in more air it simply sees boost as elevated ambient pressure. You need to think about this for a bit and then things become much simpler.

So now for fueling calculations:
Let's take an example where we assume a max flow per rev under boost to be 50% above VE=1 or a VE=1.5, under this max condition air flow per rev would be:
1.9g * 1.5 = 2.85g per rev
so we would be ingesting 2.85g/rev at a VE=1.5 now we need to match grams of fuel to that air mass like this:
Most EFI systems assume an AFR of 14.7 for this first calculation and then apply the target AFR correction later. So if we want to know how much fuel in grams is needed for 2.85g of air it's very simple calculation:
2.85g / 14.7 = 0.1939g of fuel needed to match that air mass.

Then the EFI models the injectors so it knows the relationship between injector size and grams of flow per millisecond. I could go on to explain how this is done but all you need to know is that we want to calculate a mx inj pulse width and pick an injector that achieves a pulse width of about 8ms for batch injection (per rev fueling) or 16ms (sequential). No magic in these numbers, a motor spinning at 7000RPMs is turning at about 8.6milliseconds per rev, so you MUST get all fuel injected within the 8.6ms time.

So for batch injection the inj pulse width can be from 0-8ms and if 8ms is at VE-1.5 then at VE=1.0 the PW would be 5.33ms (5.33 * 1.5 = 8ms).

So now we know that a PW > 5.33 is VE>1.0
and PW<5.33 is VE<1.0

From here we apply Fuel correction for richer than 14.7AFR and we then apply Ignition values for the given Pulse Width at a given RPM.

The main point I'm trying to make is that you can do boosted applications without
any real consideration to the fact that the engine is boosted (The MAF based EFI does not care) if you have a good air metering device you can trust then it does not matter that it's boosted, the formulas for calculating fuel and Injector PW are unchanged. Furthermore VE is directly tied to Pulse Width and it has a direct relationship for the given setup. I showed this by illustrating that the PW=5.33ms has a direct relationship to a VE=1.0

Many OEM EFI system do employ other devices such as MAP or TPS but this is mainly as a backup for a failed MAF and also for Accel Enrichment. But our Horizontally apposed motors need very little Accel-Enrichment and practically none at all above 4000RPMs. Accel Enrichment was mostly needed and very important in wet intake systems and not so much in Injected systems that inject fuel at the valve. Above 4000RPMs air flow through the runner is sufficiently high enough that not much fuel accumulates on the walls of the runner so we don't need Accel-Enrichment to recover that wall fuel.

One last point: a proper MAF can accurately measure air flow even under boost, assuming it's properly sized. The problem is that folks are trying to boost 3.2L engines and retain the stock AFM, that won't work! The stock AFM is only good for about 12,000 g/min and any boost above 4PSI will easily exceed this limit. I use a MAF with a max flow rate of 22,000 g/min it has plenty of head room for even VE as high as 1.8!

__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible

Last edited by scarceller; 11-19-2014 at 04:55 AM..
Old 11-19-2014, 04:39 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #181 (permalink)
Chain fence eating turbo
 
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,844
I would love mass air mine, but like SS said, I'm ok with slight AFR fluctuations and slightly lesser MPG's.

At highway speeds, mine gets great MPG's thanks to you stating you run around 50* of timing at those speeds and light loads.

Really made the car enjoyable on longer drives.
Old 11-19-2014, 04:47 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #182 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Quote:
Originally Posted by Tippy View Post
I would love mass air mine, but like SS said, I'm ok with slight AFR fluctuations and slightly lesser MPG's.

At highway speeds, mine gets great MPG's thanks to you stating you run around 50* of timing at those speeds and light loads.

Really made the car enjoyable on longer drives.
Tippy,

MAP and/or TPS can work well so long as you also include IAT and Altitude compensation and it's done correctly. Folks who live in Colorado where both IAT and Altitude can change over the course of a short drive really need this done correctly. With a MAF it's all baked in since the MAF measures density of air the EFI does not need to perform calculations to estimate this very import variable. The more I read and learned the quicker I realized that for most street cars MAF is the very best solution but not the only one. And the new more modern MAFs are not as prone to intake pulse reversions as the older hot wire MAFs. Here's a pic of my MAF I use, this is one of the very best on the market today:
See next post for pic
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible

Last edited by scarceller; 11-19-2014 at 05:04 AM..
Old 11-19-2014, 05:01 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #183 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage


This is a Lab grade MAF each one is flow benched and a ISO 9000 report is created for tracking. You get Flow Data for the given MAF and tolerances are kept very tight (less than 1%) from unit to unit. And very affordable, the days of hot wire devices are gone.
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible

Last edited by scarceller; 11-19-2014 at 05:07 AM..
Old 11-19-2014, 05:05 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #184 (permalink)
Stranger on the Internet
 
patkeefe's Avatar
 
Join Date: Apr 2004
Location: Bradenton, FL
Posts: 3,191
Sal, is it correct to say you have written an algorithm and coded it into the Motronic chip specifically for the MAF you reference above?
__________________
Patrick E. Keefe
78 SC
Old 11-19-2014, 05:28 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #185 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Quote:
Originally Posted by patkeefe View Post
Sal, is it correct to say you have written an algorithm and coded it into the Motronic chip specifically for the MAF you reference above?
Correct! I rewrote the entire air Fuel portion of the code to correctly handle the MAF. My fuel tables are no longer trim maps they are target Lambda maps done just like modern day EFI systems do. Also the code handles the super fast response times of the MAF. The MAF has 10ms or better response times while the AFM is at above 100ms, the door in the AFM is heavy and takes time to react to a sudden air flow change. You can't just make a MAF pretend to be a AFM they are not the same devices.

Two very knowledgeable tuners taught me the basics for tuning and how the formulas work I then combined this with what I learned about the Motronic and I then re-wrote the parts of the code that needed to be changed. Here's the setup in the car, next post
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible

Last edited by scarceller; 11-19-2014 at 05:42 AM..
Old 11-19-2014, 05:40 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #186 (permalink)
 
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage


This is the 2nd generation MAF I did, the first was a Ford Racing Hot Wire MAF, it worked well but this MAF is far better.
No hokie MAF to AFM signal boxes, the MAF signal goes directly into the DME as is and is processed by the DME properly.
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible
Old 11-19-2014, 05:43 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #187 (permalink)
Chain fence eating turbo
 
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,844
Sal, you have my gears working.

Where do I buy this MAF cartridge?

Ultimately, I'd like to use ITB's, so not sure if a MAF sensor would be worth the effort?

Maybe it would make tuning easier? I haven't throught it yet. Just thinking out loud.

MAP signals change so rapidly with ITB's, you have to use TPS movements to calculate what the the engine is injesting.

Maybe the MAF would simplify ITB's?

Pondering here....
Old 11-19-2014, 09:18 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #188 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Quote:
Originally Posted by Tippy View Post
Sal, you have my gears working.

Where do I buy this MAF cartridge?

Ultimately, I'd like to use ITB's, so not sure if a MAF sensor would be worth the effort?

Maybe it would make tuning easier? I haven't throught it yet. Just thinking out loud.

MAP signals change so rapidly with ITB's, you have to use TPS movements to calculate what the the engine is injesting.

Maybe the MAF would simplify ITB's?

Pondering here....
This ITB setup can easily be MAFed
Itb Efi Mfi?

And another example that's easy to MAF
'72 911T Convert Varioram Intake to PMO ITB??

But I think you'll give up some mid range torque with ITBs in favor of moving torque up the RPM range. ITBs are best on motors that can rev 7000-8000RPMs
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible

Last edited by scarceller; 11-19-2014 at 10:01 AM..
Old 11-19-2014, 09:48 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #189 (permalink)
I would rather be driving
 
jpnovak's Avatar
 
Join Date: Apr 2000
Location: Austin, TX
Posts: 8,116
Sal, Thank you for a Most Excellent post. As usual it was very informative.

I think this has thread has transitioned into a comparison of types of EFI as compared with the original question of EFI versus CIS.

The MAF can be found here.

http://www.pmas-maf.com/mass-air-sensors.php

Cory, ITBs should not be a problem for MAF. As long as you have an intake plenum in front of the ITBs where the MAF can get a calibrated signal. IN the case of a turbo you just need one of the connection pipes to be the correct diameter. Each MAF is calibrated to a specific tubing size such that its mass/volume/velocity curves are consistent with the calibration conditions.

I have not had a chance to try a MAF but it is very intriguing.
__________________
Jamie - I can explain it to you. But I can not understand it for you.
71 911T SWT - Sun and Fun Mobile
72 911T project car. "Minne" - A tangy version of tangerine
classicautowerks.com - EFI conversion parts

Last edited by jpnovak; 11-19-2014 at 11:45 AM..
Old 11-19-2014, 11:41 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #190 (permalink)
Chain fence eating turbo
 
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,844
With little reversion using ITB's and Sal's ISO compliant MAF cartridge that directs airflow in a way to prevent intake pulse sampling errors, the MAF might be the hot ticket for ITB tuning.

Hmmmm 😍
Old 11-19-2014, 11:53 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #191 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Jamie,

The MAF comes installed in a housing, I order the 80mm housing the inlet side is 80mm and the outlet tapers down to 76mm. I then sent him my actual filter setup on the inlet side and he flow benched the entire setup and delivered back the setup with the full flow sheet. Getting this correct is the 2nd most important part of EFI, the first is having the injector specs and correctly modeling the injectors. You can NOT bypass these two steps! Most common EFI setup errors are these 2 areas.

Quote:
Originally Posted by jpnovak View Post
Sal, Thank you for a Most Excellent post. As usual it was very informative.

I think this has thread has transitioned into a comparison of types of EFI as compared with the original question of EFI versus CIS.

The MAF can be found here.

PMAS

Cory, ITBs should not be a problem for MAF. As long as you have an intake plenum in front of the ITBs where the MAF can get a calibrated signal. IN the case of a turbo you just need one of the connection pipes to be the correct diameter. Each MAF is calibrated to a specific tubing size such that its mass/volume/velocity curves are consistent with the calibration conditions.

I have not had a chance to try a MAF but it is very intriguing.
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible
Old 11-19-2014, 12:01 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #192 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
On the CIS topic. I owned a 75 2.7L with CIS and if I had that car today I'd keep the CIS but I'd rip out the factory CDI and install a programmable stand alone ignition that sampled RPM and MAP and then allowed you to build a custom Load vs RPM ignition table. If you did that you'd have fully programmable ignition across all engine loads and rpm.

In summary:
- keep the CIS it works well!
- ditch the ignition and install fully programmable system
- and I personally would use old school inductive coil if possible.

I have not researched programmable ign systems much but several are available, find a good one that has a 2 axis table with load on one axis and RPM on the other, and Load is deduced from MAP or TPS.

The table would look something like this but the rows would be load 0-100%
http://forums.pelicanparts.com/uploads24/3+2LIgnition1416337729.jpg

Hint: this table would be a great starting point for a 3.0L CIS car! Also try finding a product that offers hit-trace feature, this allows you to see on the laptop exactly what cell your in during a live tune session while on the dyno. With hit trace you can quickly optimize the ignition for peak torque for the given load.

This system at a glance looks decent
http://www.autosportlabs.com/product/megajolte/
But it's all Ford based components, and no more distributor at all! Useds Ford EDIS, decent ignition system, more than enough for these motors.
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible

Last edited by scarceller; 11-19-2014 at 12:17 PM..
Old 11-19-2014, 12:07 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #193 (permalink)
Registered User
 
9Thirty's Avatar
 
Join Date: Oct 2011
Posts: 217
I installed this, man what a difference. MSD Digital Programmable 6AL-2 - 6530.

And the coil. Blaster 2 Coil Hi-Performance - 8202
Old 11-19-2014, 03:34 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #194 (permalink)
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Quote:
Originally Posted by 9Thirty View Post
I installed this, man what a difference. MSD Digital Programmable 6AL-2 - 6530.

And the coil. Blaster 2 Coil Hi-Performance - 8202
Can you share how you tuned the ignition?
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible
Old 11-19-2014, 04:42 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #195 (permalink)
Registered User
 
9Thirty's Avatar
 
Join Date: Oct 2011
Posts: 217
A master tech did it on a dyno. BEHE Performance I would never attempt that. John has worked with Porsches for over 30 years. It starts right up, idles at 800, smooth and quick throttle response, AFRs are 12 at full redline. He also adjusted the Leask WUR. It put out 350 whp and 353 wtq on a chassis dyno, .8 bar. With a modified fuel head (Rarly8), I could change the spring to 1 bar and it would probably bump up to 380 to 400 whp. Of course that would take another afternoon on the dyno to make sure the fuel was good along with the ignition and timing. Domino effect.........
Old 11-19-2014, 05:17 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #196 (permalink)
Chain fence eating turbo
 
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,844
Quote:
Originally Posted by jpnovak View Post
Sal, Thank you for a Most Excellent post. As usual it was very informative.

I think this has thread has transitioned into a comparison of types of EFI as compared with the original question of EFI versus CIS.

The MAF can be found here.

PMAS

Cory, ITBs should not be a problem for MAF. As long as you have an intake plenum in front of the ITBs where the MAF can get a calibrated signal. IN the case of a turbo you just need one of the connection pipes to be the correct diameter. Each MAF is calibrated to a specific tubing size such that its mass/volume/velocity curves are consistent with the calibration conditions.

I have not had a chance to try a MAF but it is very intriguing.
You know Jamie, I think this is the next step. A MAF is just so accurate vs MAP.

Just not sure if it has too much latency using a turbo? I mean, there's a lot of feet of piping before the "metered" air actually reaches the combustion chamber on a turbo'd car. N/A like Sal would be ideal.

Don't know if I'm wrong or right, just thinking out loud again.
__________________
Cory - turbo'd '87 C3.2 Guards/Blk, 3.4, 7.5:1 CR P & C's, soon to be 993SS cams and GSXR 750 ITB's fed by 964 intake, Borg-Warner S366 turbo @ 1.4 bar, Treadstone full bay IC, TiAL F46 WG, HKS 1 1/2" BOV, twin 044 pumps, MegaSquirt 2 (v3.57 board) w/EDIS, Tramont wheels (285's rr, 225's frt), Big Reds frt, 993 rr., tower brace, MOMO wheel
Old 11-19-2014, 06:33 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #197 (permalink)
Chain fence eating turbo
 
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,844
Quote:
Originally Posted by 9Thirty View Post
I installed this, man what a difference. MSD Digital Programmable 6AL-2 - 6530.

And the coil. Blaster 2 Coil Hi-Performance - 8202
This is exactly what every CIS owner should do if going EFI is not in the cards.

There is so much to be gained with load/RPM based timing.

With fully programmable ignition, a 930 would have better throttle response, more torque at part throttle loads, and better overall MPG's.

A 2D dizzy just cannot replicate this without trade offs.
__________________
Cory - turbo'd '87 C3.2 Guards/Blk, 3.4, 7.5:1 CR P & C's, soon to be 993SS cams and GSXR 750 ITB's fed by 964 intake, Borg-Warner S366 turbo @ 1.4 bar, Treadstone full bay IC, TiAL F46 WG, HKS 1 1/2" BOV, twin 044 pumps, MegaSquirt 2 (v3.57 board) w/EDIS, Tramont wheels (285's rr, 225's frt), Big Reds frt, 993 rr., tower brace, MOMO wheel
Old 11-19-2014, 06:39 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #198 (permalink)
 
Registered User
 
scarceller's Avatar
 
Join Date: Oct 2006
Location: Southern MA
Posts: 3,645
Garage
Quote:
Originally Posted by Tippy View Post
You know Jamie, I think this is the next step. A MAF is just so accurate vs MAP.

Just not sure if it has too much latency using a turbo? I mean, there's a lot of feet of piping before the "metered" air actually reaches the combustion chamber on a turbo'd car. N/A like Sal would be ideal.

Don't know if I'm wrong or right, just thinking out loud again.
Tippy,

The PMAS MAFs are designed for blow through. You would put it right at the Throttle-body to reduce this air flow latency.
__________________
Sal
1984 911 Carrera Cab M491 (Factory Wide Body)
1975 911S Targa (SOLD)
1964 356SC (SOLD)
1987 Ford Mustang LX 5.0 Convertible
Old 11-20-2014, 05:16 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #199 (permalink)
Chain fence eating turbo
 
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,844
Oh no......slippery slope.....

I'm seriously considering this.

Old 11-20-2014, 09:44 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #200 (permalink)
Reply

Thread Tools
Rate This Thread
Rate This Thread:

 


All times are GMT -8. The time now is 04:21 AM.


 
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Search Engine Optimization by vBSEO 3.6.0
Copyright 2020 Pelican Parts, LLC - Posts may be archived for display on the Pelican Parts Website -    DMCA Registered Agent Contact Page
 

DTO Garage Plus vBulletin Plugins by Drive Thru Online, Inc.