Using the EFI Correction Box

[TomoHawk]

In any case, the question was how to use the corrector box and figure out if your ECU was drifting. How much is just a matter of measuring the resistance you added.

[EuroDat]

I used a 2k pot (made by Alpha, p/n: B2K) to adjust mine. Ended up at around 1250ohms. That was driving by feel with acceleration tests.

After that I cleaned the dust out of the ECU with a vacuum cleaner and now its around 950ohms.

If I had a spare ECU I would try washing it in ultra pure water and drying it for 4 to 5 hours at 50 degC to see if that improves drift any.

The water would need to be deionized or double distilled to have a low microsiemens range preferably under 10us/cm* (microsiemens). Then it has the best absortion rate. We use this process where I work to clean Siemens S5 plc's with about 90% sucess rate. The plc's use less energy, run cooler and perform better after washing, but they have pleny of spares, I on the other hand don't:disappoin.

Edited May 21, 2013 by EuroDat

[TomoHawk]

Please tell us more about how you got your final value using the driving by feel method? I think it would also be important to see what resistance works bets with long-distance driving, which is what I do. Perhaps there is a way to set up a correction box for both situations?

Edited May 21, 2013 by TomoHawk

[siteunseen]

Put longer wires on there and run them through the fire wall leaving the box inside while you drive. Put some fuel in and go play, you'll feel, hear and see the differences. When mine got "right" the temp gauge was cooler the fuel gauge didn't move as fast and the car itself will pull like never before.

[sblake01]

Or perhaps he's an electrical engineer with a long history of working on fuel injection systems (going back to Robert Bosch's ORIGINAL digital EFI based on the Motorola 6805 MCU...which he helped program) and you misread his original post.... really? REALLY!

Apparently your quotation marks around the word "chipping" in the post that illicited that response weren't big enough?? Edited May 22, 2013 by sblake01

[TomoHawk]

Digital computers fast enough to control a FI system, and rugged enough to survive in an automobile didn't exist outside of the military until well into the 1980's.

For sure. My first car as a 1985 1/2 Escort, which was the last one made with a carburetor. With the 1986 model, Ford had EFI installed.I should've waited a few months more, for a totally different headache!

[TomoHawk]

I have an Android tablet PC with GPS and a good G-sensor. Maybe I can use one of those dyno applications with this "POWA" box make a

Come to think ofit, I'll have big POWA decal to put on the box too

[TomoHawk]

Well, I spent the time to connect the extension wire and connect the corrector box, and I did drive the car around the block twice (a few miles each.) I went through a quick pre-check in the driveway, trying the different ranges, and the different resistance extremes. At each stop sign or traffic light, I changed the resistance, then try an acceleration test. It was very hard to tell if the thing was 'pulling hard' compared to the uncorrected system. I think that I would have to say that just a small amount of added resistance gave the bet improvement, though it would be hard to say, since I was not able to get any real, quantifiable data to compare. The acceleration test itself is so informal that you could even say that no added resistance gave the best "feel."

I didn't have time to try a test on the highway.

I did note that too much added resistance caused the engine to stumble a little, though it happened at a long red light, so it wasn't a dangerous thing. A better way to test for the right amount (if any) of added resistance would be a very good thing. I will have to inquire with the person who invented the resistance box to see how he used it.

It also occurred to me that other things like the ignition timing and the AFM setup could have some effect that makes things feel better or worse...

But at least there is this cute little box in the car with a label that has in big letters, "POWA!" which ought to impress some people.

[Wade Nelson]

Didn't Herr Dr. Bosch pass away in 1942?

Yeah, it was a real bear having to work with his ghost. We had spirits, mediums, Tarot readers, you name it trying to discern how Herr Bosch wanted the company that bears his name to implement fuel injection digitally.

For instance, Madam Suarend INSISTED that exhaust gas temperature sensing and knock control form the primary feedback loop for Bosch's D-EFI. Whereas Senora Sweed said that wastegate control with octane sensing and the new, LAMBDA sensor should provide the primary feedback.

Working with dead automotive engineers, it's not for lightweights.

As for "How to improve" the L-Jetronic system, in what way would you like to improve it? Fuel economy? Power? Everything's a trade-off. Elsewhere in this thread there is a description of the resistors in the L-jetronic and what each one affects. If I wanted to "improve" the L-Jetronic system in one particular area, I might begin by altering the values of one of those resistors, OR, as this hack does, replace it with a potentiometer so I could vary it while driving.

Of course I'd also want to ADD an 02 sensor or an Air-Fuel Ratio sensor to my exhaust stream so I could have a guage and get some feedback on what I was doing, short of putting it on a dyno or sticking a gas probe up the tailpipe.

Did you even bother to read the article about Harley Davidson EFI that I posted? If you really want to "improve" the L-Jetronic ANALOG system you first step would be to COMPREHEND how more modern, digital EFI systems work -- and fall short --- and then see which of those lessons you might apply to the L-J via analog circuits rather than digital "hacks."

Oh, and have a great day...

[Captain Obvious]

Working with dead automotive engineers, it's not for lightweights...

I never trusted a thing Sweed said. I'd put my money on the Madam.

Backing up a little... I'm not sure how to interpret your earlier post. Did you help program the digital EFI (which used a 6805 MCU as it's brain), or were you involved in the programming to create the 6805 processor itself? In other words, did you work on the internal architecture of the 6805 or were you using it as your target on your board?

Either way, it's cool. But I just don't know how you really meant it.

[Wade Nelson]

I helped program the 6805 MCU to monitor load (vacuum sensor) and rpm and do a table lookup from a 2d map as to the "correct" pulse width. From there we added acceleration enrichment, crank enrichment (similar to choke), 30 seconds of hi idle on start-up, idle boost when Air Conditioning was commanded, and other "features" in software that mimic'd what the best carburetors & assorted folderal did.

And that was just getting started --- no carburetor EVER did decel fuel cut-off for improved MPG's....

We once had acceleration enrichment and decel fuel cutoff programmed "too close" and the car bucked like a bronco as it threw you (and your foot) on and off the gas, first enriching (added power throws you back in the seat....foot comes off the pedal...back into decel cut-off, you get thrown forward, foot goes back on the pedal... we laughed all the way back to the garage in Tempe...

02 sensors, knock sensors, coolant temp sensors, idle air control, mass airflow meters were well in the future when we began driving the first D-EFI controlled Audi around the block under 6805 control...

"Look Ma! No carburetor!"

[Captain Obvious]

Cool. Was this for actual money, or your senior project?

If it was for $, who were you working for? Bosch? Audi? 3rd party contractor?

When we meet, remind me to tell you a story about control loops and very high gas mileage. I'm not writing anything down because I'm not sure the NDA has expired yet, but I'll tell you in person. :tapemouth