WOW! Sounds like your project was a success! However, If you could manage to get some dyno time along with a WBO2 in the tailpipe. You'd then have some real data to look at. It's not cheap but it would be adventagious to you as numbers would really help sell your kit if you plan on producing them.

yea sounds like it works well! im sorry about all my sceptical questions!
it all worked well in the end, and ur obviously like me with the fact that you made it yourself and it works makes it so much cooler

No apologies needed, I posted here about it so that you guys could come up with some things I didn't consider. Criticism is appreciated, though it appears that the controller does indeed work as designed. I did all my driving today with it on the car but I don't think I got in closed loop much at all. It was so cold that with the heat blowing inside the engine wouldn't warm up enough to get out of open loop. So much for trying to measure an MPG improvement over this tank of gas.:rolleyes:

I did a lot of driving over the holidays. Almost all of which was with the controller pulling fuel out. I managed to increase my average fuel mileage by 3.5 MPG!!! Sure wish I had this back over the summer when gas was at 4$ a gal. That increase in MPG was with various amounts of lean running and not just full lean so there is probably just a little bit more MPG to be had. Granted it's no powerhouse running lean but it does still go down the road just fine. I havn't noticed any pinging at all, even at the full lean setting but the weather is cool and the air is thin up here near Denver. It may be a different story when summer gets here.

Still no negative issues, no C/E light, or problems to report.:D

Duxthe1,

I just found this thread, and am VERY intrigued. I have recently been talking to a fellow enthusiast about this very subject; fooling the ECU via a modified 02 sensor output. (However, the intended application is not a CIS or CIS-E car, but instead LH-SFI M119 applications.)

The CIS-E cars have a certain degree of simplicity which is very much in your favor; I believe getting the ECU to “digest” an altered signal is fairly straightforward and WILL work well, as you have discovered. (It would also work for the older CIS systems with a frequency valve instead of an EHA.) I applaud your inventiveness & diligence in developing the methodology. I particularly like the idea of being able to adjust the mix on the fly.

I was not looking at partial load situations, but strictly full load enrichment. I had envisioned a much cruder system: using a resistor to drop the voltage sensed by the ECU, thereby feeding a false lean condition with the modified signal being switched into & out of the circuit via a double pole relay tied into the kickdown switch. Unfortunately, this is where my ignorance of the LH-SFI system comes into play – if they go open-loop at full throttle, then my idea will not work - it would then take an actual re-mapping of the fuel flows burned into the LH unit. (And I do know that the 1992 LH systems DID have WOT enrichment, but probably not “enough” to achieve full power potential in lieu of emissions & catalyst life expectancy)

Also, being an MB tech and therefore being MUCH more knowledgeable than myself, I was wondering if there is any way to adjust the basic lambda settings on an LH car? (and if so, would that only effect idle settings, or would it shift the entire fuel map commensurately? (the only analogy I can think of is when adjusting an old (non-lambda) CIS car with a hex key near the fuel distributor head, OR adjusting the control pressure via modifying the warmup regulator.)

I would LOVE to hear your feedback on this stuff!

Thanks in advance!


P.S. - If you are planning on producing these controllers as kits, you will find a much wider market if you were to include the LH-SFI cars in your target market. Meanwhile, if you need a guinea-pig to test your prototype on an LH-SFI car…

P.P.S. On your friend’s stroked 500SL, was this simply a long-throw (5.6) M117 crank into a M117 engine, or into a M119?

Works on a KE-Jetronic ...concept or mod to the LH-SFI?

http://www.autotech.com/prod_engine_pwrmod.htm

As built my controller should be able to work on any system that uses a single oxygen sensor. That pretty much excludes most if not all of the LH-SFI stuff, possibly not on some of the early ones, maybe. Interfacing with the mass air flow meter signal is the generally accepted way to tune most EFI. I built the controller to give us CIS cars a way to alter the mixture.

So far I've got around 1k mi on the car with the controller in charge of the O2 signal. So far so good.

My buddys SL is an 84 500 Euro block with the 560 crank rods and pistons.

There is no basic adjustment on a US spec LH-SFI. There are a few ways you could hack it but nothing simple.

The rear o2 sensor of most SFI cars doesn't affect the fuel trims of the car, including Mercedes, so I think your controller would work. It would be helpful to tune the closed loop fueling of the car, making the transition from closed loop to open loop smoother when making custom fuel maps.

I believe all LH-SFI use a single O2 sensor. The dual sensors didn't start until ME injection rolled out in 1996, AFAIK. Interesting idea about modifying the MAF signal on LH cars, though!

:boat:

RBYCC - Thanks for the suggestion, but that unit works by feeding a false signal directly to the EHA, therefore it is not applicable to LH-SFI systems.

Turbo E320 - I believe that the function of any second (post-cat) O2 sensor is strictly to provide feedback on catalyst efficacy.

GSXR is correct, the LH-SFI systems ARE single 02 sensor systems, therefore (in theory) this type of signal "correction" could work.

Duxthe1: Any interest in potentially expanding your market beyond CIS-E?

P.S. I'd love to hear more about setting the baseline lambda mix on LH-SFI cars, even though it is a "difficult" hack. Nothing worthwhile is easy.. (usually) ;)

Exactly, The O2 sensor in "most" fuel injection systems have very little "say so" in mixture adjustment. If you're running LH injection it would be best to look at altering the mass air signal for fuel enrichment as duxthe suggested. In fact, I have a piggyback controller for this application if anyone wants to buy it.

As far as being able to hack an adjustment for the basic fuel delivery, I have two words.... fuel pressure. :eek::cool:

In years past I have taken a bosch style fuel pressure regulator and cut off the dimpled end and replaced it with a threaded adjuster. Not exactly pretty but very effective.:D

Only problem with altering just fuel pressure is that it's very "linear" which engine power curves are not. With this method you'd be running too rich in areas where you should not.

I did say it was hack but is more effective than your argument suggests. Since the map is unaltered in the ecu it will run richer for the same inputs across the entire range of delivery. Whether the engine will run better enrichened across the whole map, and whether the ecu will adapt out the enrichment are not addressed.

The last time I used this trick was on a speed density setup where I was pushing 21psi at 2 liters. :eek: It was making approx 100hp over stock on a stock ecu.

Yes, the ECU would adapt and negate the enrichment, up to the system's inherent adaptation limits. It would do the same with fuel pressure hacks, or anything else where the ECU sees the mix at the 02 sensor as being anything other than stoichiometric. (obviously speaking of closed-loop mode here.) WOT enrichment is one thing, but "on the fly" tweaking during closed-loop operation has some inherent advantages, as Duxthe1 has discovered.

Those are the reasons why I'm so interested in Duxthe1's methodology. Same questions apply:
1.) are you planning to produce this?
2.) would you like to expand into a wider potential market?

I know that there would be PLENTY of interest from certain groups of enthusiasts that have LH cars!