Welcome to Rideroom, an online automotive community where you can discuss all things related to cars, trucks, motorcycles or any other vehicle you may have.

To gain full access to Rideroom you must register for a free account. As a registered member you will be able to:
  • Create a Rideroom to showcase your cars, trucks and motorcycles on the main Rideroom site.
  • Give "Props" to rides you like
  • Start a Rideroom Group for your car/truck/bike club
  • Post comments on rides and news stories.
  • Post and reply to topics in the forum
  • Communicate privately with other automotive enthusiasts (PM) on the forum
All this and much more is available to you absolutely free when you register for an account, so sign up today!

If you have any problems with the registration process or your account login, please contact contact support.

Go Back   Rideroom Forums > General > Articles
FAQ Members List Mark Forums Read

Reply
Thread Tools
  #1  
Unread 02-11-2007, 03:12 PM
lemming's Avatar
lemming lemming is offline
STOP EATING MY SESAME CAKE!
 
Join Date: Nov 2005
Location: Duvall, WA
EFI Geekery

(Although some of the topics discussed in this article are specific to Ford EEC-V applications, most of the concepts are still broadly applicable to any modern EFI system. My goal here is to clarify some basic points about how EFI systems work, and explain why some common "fixes" offered by the aftermarket community can be dangerous. If anyone actually reads this, I might do some more on stuff like tuning for power adders, adjusting idle on heavily modified vehicles, etc.)

What does the ECU do with its sensors?

Of all the devices hooked up to an EEC-V, O2 sensors and MAF meters are the most critical. The MAF meter is a very simple device--it is basically a set of heated wires. The device emits a voltage (typically fom zero all the way up to battery voltage) based on the mass of air that is flowing across the wires. The computer runs this signal through a transform called the MAF transfer function, and thus calculates the total amount of air flowing through the intake tract. This value is then used in several ways, including two critical operations--fueling and spark.



MAF transfer function. X-axis is in "counts" (voltage), Y-axis is amount of air in lb/min.

Fueling is relatively straightforward. Since the EEC knows how much air is flowing through the intake tract, and it has stored values that describe the behavior of the fuel injectors, it can calculate how long the injectors need to pulse. At part throttle, the EEC runs in a feedback loop, using the O2 sensors to "steer" how much fuel goes into the engine and correct for small issues, including minor variations in fuel quality, minor sensor contamination, and so on. At full throttle, things change much too quickly for the EEC to apply an adaptive correction, so it switches to "open loop" mode, using only the MAF signal and the preprogrammed injector constants to determine how much fuel to spray.

Of course, the EEC can do much more with the MAF signal than just dump fuel into the engine. Since it knows how many pounds of air are flowing into the engine, how large the engine's displacement is, and how fast the engine is spinning, it can also estimate the volumetric efficiency of the engine--that is, how much air is flowing into the engine compared to how much air the engine displaces. This is often called the load calculation.

The EEC maintains several preprogrammed spark tables. Engine speed is on the X-axis of each table, and load is on the Y-axis. Each cell in each table contains a desired level of spark advance for a specific combination of load and RPM. It's not uncommon for the EEC to have at least three spark tables. The first is the MBT (Max Brake Torque) table, which contains the ideal amount of spark for each speed/load point. The second is the Max Allowed table, which contains the maximum amount of spark that can ever be delivered to the engine at each speed and load point; this table is typically used to restrict the engine's torque output to be somewhat less than MBT, especially at low load and idle. The third, and perhaps most critical, table is the Borderline Knock table, which is typically set up to contain the maximum amount of spark advance the engine can handle at a coolant temperature of 180*F without pinging.


Maximum Allowed Spark table; one of several spark tables that the MAF is critical in controlling.
X-axis is RPM, Y-axis is load, calculated from the MAF signal, spark advance (in degrees) is stored in each cell.


The EEC applies some air and coolant temperature adjustments to the Borderline Knock table and generates a value that represents the maximum amount of spark the engine can tolerate. It then compares this to the values from MBT and Max Allowed and picks the minimum of the three. under normal operation (i.e not idling or triggering traction control), the EEC basically just picks the minimum of the three and delivers that level of spark to the engine.

Why does this matter?

EEC logic matters because it's important to know what modifications are Band-Aids that can easily have undesirable effects. For example, Pro-M made MAFs that were "calibrated" to specific engine/injector combinations. If your stock 5.0 had 19 lb/hr injectors and you needed 24 lb/hr injectors for your modified engine, you could buy a MAF calibrated for 24s. For a given amount of airflow, such a MAF would emit a lower voltage than the stock one. This would normally cause the engine to run lean by some percentage, but since your injectors were oversized by the same percentage, fueling was OK and all was right in the world.

Wrong! Although these MAFs would enable modified engines with oversized injectors to run the correct A/F ratio without reprogramming, the load calculation would be wrong. Specifically, the EEC would underestimate the load level and deliver more spark to the engine than it otherwise would. If the MAF in question was calibrated to 36 lb/hr injectors, the load calculation on our hypothetical EEC would be almost 50% low. Unless the factory tune was absurdly conservative, this would almost undoubtedly cause detonation.

Similarly, if you stick the MAF sensor in a different diameter tube than it originally came in, the transfer function gets thrown off--if the tube is larger, the MAF now samples a smaller fraction of the air entering the engine and therefore tends to underestimate the amount of fuel required. Worse still, the EEC may deliver too much spark to the engine (since an engine at lower load can tolerate more spark), increasing the likelihood of detonation. It's important to have a tuner fix the MAF transfer function when you do anything that throws the MAF's accuracy off.

Thus, when modifying engine electronics, it's important to consider how the EEC deals with part throttle fueling, full throttle fueling, and spark control. If you're replacing a MAF with one that is calibrated to a certain set of injectors, you are rolling the dice.
__________________
[2003 4.2L V6 Mustang] [1999 Subaru Impreza WTF]
On our way back from the beach, we were trying to decide where to eat dinner. At that moment, we were passed by a Rendezvous with the Ohio license plate "ASK GOD." So we did, and we ended up at some shitty Texas roadhouse-style place. Next time, I pick.--AchTTung
Reply With Quote
  #2  
Unread 03-19-2007, 08:47 PM
lemming's Avatar
lemming lemming is offline
STOP EATING MY SESAME CAKE!
 
Join Date: Nov 2005
Location: Duvall, WA
I've had a long ordeal with tuning my own car, and it was actually a result of the exact concepts I have described here.

Ever since the supercharger was installed and the 42 lb/hr injectors went in, I've had myriad tuning problems. The idle would hang at 1500 RPM, and even if I used some trickery with the idle air control valve to bring it down under 1000, it'd have a periodic misfire. Below 1500 RPM (and most noticeably at 1200 RPM), it'd buck and shuffle, which made the car a miserable chore to drive in heavy traffic.

Why was it messed up? Because some of the injector parameters were wrong, and this caused me to make changes to the MAF transfer function which totally borked the load calculation. Since the load calculation even controls spark at idle (which is actually the main factor in idle control, NOT the idle air valve), my incorrect load calculation screwed up everything else.

Fuel injectors have four main characteristics, and all of them vary with fuel rail pressure.


Injector behavior; it's important to note that an injector has nonlinear behavior and that it takes time to open!

The first is the offset. This is how long the fuel injector takes to open when the ECU asks it to do so.

The second is low slope. This is how fast fuel exits the injector when it first opens.

The third is the breakpoint, which is how long the low slope phenomenon lasts.

The final parameter is the high slope, which determines the overall rate of flow of the injector.

There are constants and functions in the ECU that are designed to model these phenomena. I had loaded the correct values for the slopes and breakpoint, but the offset was wrong. The offset value stored in the computer was for my stock 21 lb/hr injectors, and my new injectors opened in about half the time! As a result, I was injecting more fuel than expected because the EEC didn't "understand" how quickly the injectors would open after it sent the signal. I'd leaned out the MAF transfer function to compensate, and the load calculation was off by ~20% at idle.


Offset error. Note that the error is much more significant at short pulse widths.
This matters because shorter pulses are used at idle and light-throttle driving.


Although I couldn't directly control the offset parameter, I did have control over a couple of the modifier functions (which include battery voltage compensation), so I just cut it in half, and now life is much better.

So, a year of tuning and frustration came down to an erroneous calculation that occurred because one parameter wasn't being correctly modeled. As I said in my previous post, it's important to get these fundamental things straightened out first because it saves a lot of work in the long run.

If I hadn't changed so many things at once (supercharger, injectors, and some MAF wiring), I might have realized my mistake earlier and saved myself a lot of hair-pulling.
__________________
[2003 4.2L V6 Mustang] [1999 Subaru Impreza WTF]
On our way back from the beach, we were trying to decide where to eat dinner. At that moment, we were passed by a Rendezvous with the Ohio license plate "ASK GOD." So we did, and we ended up at some shitty Texas roadhouse-style place. Next time, I pick.--AchTTung
Reply With Quote
  #3  
Unread 07-07-2007, 12:54 AM
pontisteve pontisteve is offline
Registered User
 
Join Date: Jul 2007
This is a start to great information, and I hope to hear more of your experiences. Learning from someone else's mistakes can save you a lot of wasted tuning time. Keep writing.
Reply With Quote
  #4  
Unread 08-15-2007, 04:59 PM
94whitesnake 94whitesnake is offline
Registered User
 
Join Date: Nov 2005
yeah man this is good stuff.

Hopefully I can someday play with my own forced induction toy!
Reply With Quote
  #5  
Unread 08-15-2007, 05:14 PM
NitrouStang96 NitrouStang96 is offline
Registered User
 
Join Date: Nov 2005
This is all information I can use and comprehend better once I have a bit more of a grasp of what's going on. I currently understand it all, but I'm not making any logical connections to my application.

So, let me ask you this - how did you start learning the basics? What were the basics you learned at first to start tinkering with it yourself? At what point did you start learning and tweaking this advanced stuff? Did the information come from your shop, a manual, the internet... all 3? Can you upload your XCal2 Pro Racer installer to my FTP?
Reply With Quote
  #6  
Unread 08-15-2007, 05:15 PM
lemming's Avatar
lemming lemming is offline
STOP EATING MY SESAME CAKE!
 
Join Date: Nov 2005
Location: Duvall, WA
I'm going to have to write a follow-up to this someday soon. I learn new stuff every once in a while.
__________________
[2003 4.2L V6 Mustang] [1999 Subaru Impreza WTF]
On our way back from the beach, we were trying to decide where to eat dinner. At that moment, we were passed by a Rendezvous with the Ohio license plate "ASK GOD." So we did, and we ended up at some shitty Texas roadhouse-style place. Next time, I pick.--AchTTung
Reply With Quote
  #7  
Unread 08-15-2007, 05:25 PM
lemming's Avatar
lemming lemming is offline
STOP EATING MY SESAME CAKE!
 
Join Date: Nov 2005
Location: Duvall, WA
NS96: The logical connections are very simple. If you want the EEC to manage the engine correctly, you need to make sure that its most important calculation is correct. For this to happen in your application, you are mainly going to be concerned with loading the correct injector constants (SCT supplies a bunch for different common injector configurations) and then tuning your MAF transfer function until it fuels correctly. I'm thinking of writing another quick article about that; there are a few techniques that work for this. Since you've supercharged a car that was originally NA, the next major step is to re-norm the spark tables and set appropriate advance values for high load levels.

I learned the basics by asking two people who do this stuff for a living. I mainly started out by tinkering with the transfer function and borderline knock table to improve low-RPM drivability when I first got my 4.2. Although I don't have it, SCT sells a manual with tuning basics, and LaSota sells a CDROM tutorial which I've got (it isn't bad).

PRP licenses are controlled by a USB dongle. No dongle, no run.
__________________
[2003 4.2L V6 Mustang] [1999 Subaru Impreza WTF]
On our way back from the beach, we were trying to decide where to eat dinner. At that moment, we were passed by a Rendezvous with the Ohio license plate "ASK GOD." So we did, and we ended up at some shitty Texas roadhouse-style place. Next time, I pick.--AchTTung
Reply With Quote
  #8  
Unread 08-15-2007, 05:38 PM
NitrouStang96 NitrouStang96 is offline
Registered User
 
Join Date: Nov 2005
Frack, how much is that? And how much better is it than their free software that I've toyed with?
Reply With Quote
  #9  
Unread 08-15-2007, 07:48 PM
lemming's Avatar
lemming lemming is offline
STOP EATING MY SESAME CAKE!
 
Join Date: Nov 2005
Location: Duvall, WA
I don't know, because I didn't pay for my copy of 2.9x (it was given to me by a certain mechanic in Florida who owed me money). The upgrade from 2.9x to 3.x was $100.

It's the same software that actual dealers use when they need to make custom tunes, and it's very powerful. The only real difference from the dealer software is that PRP databases are purchased individually, so when you buy the package, you'll only get the database for your particular ECU and calibration code.
__________________
[2003 4.2L V6 Mustang] [1999 Subaru Impreza WTF]
On our way back from the beach, we were trying to decide where to eat dinner. At that moment, we were passed by a Rendezvous with the Ohio license plate "ASK GOD." So we did, and we ended up at some shitty Texas roadhouse-style place. Next time, I pick.--AchTTung
Reply With Quote
  #10  
Unread 08-15-2007, 09:39 PM
pontisteve pontisteve is offline
Registered User
 
Join Date: Jul 2007
I have LaSota's CD (which I didn't find all that helpful). I also have the SCT Pro Racer manual, which is good... but expensive. I upgraded to Advantage 3.0 (which was worth it... much nicer). But the book I found the most helpful was also the cheapest. It's Engine Management: Advanced Tuning, by Greg Banish. A lot of the stuff in it surrounds the Advantage 3.0 software, as Greg obviously favors using it, and tuning Fords.

At first, the Pro Racer software is a bit overwhelming, so you play with some easy stuff. Mainly, datalogging! After a while, you get the jist of datalogging and what some of that stuff means. Once you familiarize yourself with the datalogs, the tuning software becomes much less intimidating. An accurate wideband becomes a standard necessity of course. I now have 2 LC-1/XD-16 combos.

After being very familiar with the software, I found Greg's book to be VERY helpful. He was an OEM tuner by day, and a hot rod tuner by night. So he knows both sides, and I think he really gets it right. I bought my copy on Ebay or Amazon for a song, and was really happy with it.

The key thing I picked up in his book is just how to model airflow easily (tune the MAF). If the MAF is right, and the injector settings are right, the car will do whatever else you tell it to in the tune. The rest of the stuff is pretty much fine tuning it. Greg's book had some really helpful tips in it that I didn't find anywhere else. Buy it. And I would also recommend to anyone to buy the Pro Racer 3.0 software, even if you don't know much about tuning. Like anything, you get your feet wet and learn as you go. You can always start tinkering small, and leave the big stuff until later when you know more.
Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 01:36 PM.


Powered by vBulletin® Version 3.6.10
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.

Page generated in 0.03836 seconds with 9 queries