M42club.com - Home of the BMW E30/E36 318i/iS
DISCUSSION => Engine + Driveline => Topic started by: bmwman91 on April 27, 2007, 08:50:27 PM
-
This is something I have been thinking about for a long time. Aside from needing to put together a better portfolio of my skills and abilities to help me get a job, I wanted to get the idea out there. Check out the link for my latest nerdy endeavour. Milking the M42 in as many ways as I can here!
http://bmw.e30tuner.com/articles_econo1.php
Enjoy. Any and all thoughts are welcome.
-
http://bmw.e30tuner.com/articles_econo1.php
Cliffs Notes for the lazy please. ;)
-
WHAT THE HELL DID I JUST READ !!!!!!!!!!!!!!!!!!!;)
You're ready for a job in Detroit .....................
-
Excuse me, Doc. I didn't understand that diagnosis. Can you just tell me how long I have to live?
-
I've thought about doing this with my LC-1. Sending a constant lambda to the ECU so I can unplug the stock sensor and have complete RPM/TPS authority over my fueling.
Let us know how it works.
-
i was thinking the other day if i could devise an ECU program that can shut off two cylinders on the freeway when desired for cruising. The only problem is that with our 4 cyl. cars it might be a bit rough unless done properly.
-
bmwman91 = one of the damn smartest people I've ever had the pleasure to speak with... Good luck on the project bro, everything looks great.. Its amazing how I can actually understand what you're talking about now after 2 years of auto school.. We're learning about narrow-band O2's and all the running inputs to the ECU ect. right now :)
-
Haha thanks for the kind words people. This particular article is a bit more technical than my others...a lot of that has to do with me using it as part of my portfolio I am sending to would-be employers. Job-finding is a real pain in the rear!
In short, for those who requested it:
I am using a wide-band O2 sensor and some digital trickery to run the car a bit lean under highway and idle conditions. I spend a bunch of time investigating theoretical hang-ups posed by the idea, with the main concern being for the environment. Phew, that was quick!
-
Very nice technical write up.
So you basically just need to get a Unichip, Megasquirt, or some sort of management that will allow you to finely tune your partial load maps. A load dyno will help in that process. I tuned my boosted gsr motor to still get 30mpg on the freeway using hondata, but using throttle % instead of load site. I wish there was a "hondata" for our m42's.
-
A stand-alone would be easier in some respects. I just do not feel like dealing with another MS installation and actually trying to tune it as well as the Conforti chip is. The thing I am working on has its own feedback inputs and should allow for about the same level of control and mileage improvement. Hopefully...
-
I think what you've built is essentially a "Johnson box" as used on no cat, no 02 sensor, no closed loop feedback Euro cars. When such cars were imported to the USA, the federal gov't required closed loop emissions controls. Firms such as Johnson came out with boxes to control fuel delivery based on 02 sensor feedback. Boxes were made by several company's, but the "Johnson" box was the most popular. i have several in my collection. And the Valentine unit is the nicest. Such boxes manage the resistance on the water temp circuit and "fool" the ECU into thinking the motor is colder (or hotter) than it really is, thus richening the mixture, or leaning it out. This technique was also used by Steve Dinan in the 80's to control fuel delivery to his turbo cars.
Here's a schematic from my web site which shows how the Johnson boxes worked.
(http://bmwturbos.scottiesharpe.com/library/20050825105751_Johnson%20Box.jpg)
-
My solution: I bought a 2006 Charger Daytona, which comes with MDS (Multiple Displacement System). It shuts off four of the eight cylinders when they are not needed.
I can hardly tell when it does it, although the exhaust note definitely changes when it goes to MDS mode. Full power return is almost instantaneous when required - 350 HP is a hoot! And over the just-under-10k miles I've had the car, I am getting 19 MPG in mostly city driving.
:D
-
so on this you are basically trying to do a lean burn concept like Honda and other companies have done right?
-
Yes, I am trying to do a controlled lean burn under cruising conditions. The way the Motronic deals with the EGO input is not satisfactory for safely doing what I want by just reprogramming the LC-1 WBO2.
Of course, I am making all of the info public as open-source type projects can be improved from multiple heads thinking. If anyone has suggestions, they are always welcome!
Scottie, that is very interesting. I never heard about these systems. Manipulating the CLT input is an interesting method. I am not sure if the warm-up modes effect ignition timing on the more "modern" ECU's. Of course, I want to run leaner rather than richer, so I suppose this is out. Interesting though.
-
I have two questions:
1. At low AFR, will the power not also be reduced so effectively requiring the throttle to opened a little more negating the benefit?
2. Shutting off cylinders on the M42 will loose power due to need for the engine to still have to push them through their compression strokes.
-
1. I have definitely thought about this. From what I have read, the internal combustion engine is at its most efficient when run a bit lean (equivalence ratio of 0.9, or ~16.3:1). The amount of ignition advance also influences this. I will just have to see what happens here when I do this. So maybe if things turn out to work this way, I will lock the AFR there and still go super lean at idle. I guess if I REALLY wanted to save gas I would turn the car off at lights and in bad traffic.
2. Exactly. This would also severely throw off my EGO sensor reading with 50% air running through the exhaust. A V-motor has a much better setup for this since it is easier to split the exhaust lines and use 2 EGO sensors.
-
1. I have definitely thought about this. From what I have read, the internal combustion engine is at its most efficient when run a bit lean (equivalence ratio of 0.9, or ~16.3:1). The amount of ignition advance also influences this. I will just have to see what happens here when I do this.
This is true but if you want to run at the ragged edge, I would highly recommend an EGT gauge at the least. Most factory programmed ecu's from efficient cars like Honda run very close to 1600 degrees at cruising through maximum advance and very lean afr's. If you go over, say bye bye to your engine.
-
Yeah that's the real issue to balance here. The Motronic is not designed to run lean like that, so it will not advance the ignition timing much at all during cruising. That eliminates the big problem with detonation and blown pistons/valves. At the same time, it might end up requiring me to use more gas to maintain enough power. It will be interesting to see how this goes.
I am actually thinking of changing the thing, assuming it works. I might do a directly scalable AFR deal, dependent on TPS. Like 19:1 t 0% throttle and 16:1 at 40% with adjustable AFR targets based upon the TPS between that range. Above 14:1 it will just run what it is programmed to normally.
-
Yeah that's the real issue to balance here. The Motronic is not designed to run lean like that, so it will not advance the ignition timing much at all during cruising.
I really think for maximum efficiency you do have to advance the timing. Have you looked into any piggyback units like the Apex-I ITC? It will intercept the cam/crank signals and send a different value back to the ECU, hence fooling it into thinking it is ahead or behind.
I think using the Apex-I S-AFC in conjunction with the ITC would be a good alternative to any ems or piggy backs since these units are easy to use and setup. Of course there will be some limitations but since they are piggy backs, you won't be starting from scratch as it's just messing with the existing factory program.