80-84 people.
12-13-2003, 12:54 PM
#34
Rennlist Member
A good idea (though I have never had much trouble with my 928's L-jet). In addition the the WSM and the "Bosch Fuel Injection and Engine Management" book I have a copy of an 18 page supplimentary Porsche publication called "928 Air Flow Controlled Fuel Injection" from 1980. Part number seems to be WKD 459 521. Let me know if you want a copy for your project and I will try to arrange it.
In addition the the standard information it would be nice to have a place for unpublished information that is hanging around this group such as: How does the AFM really work:
-what is its resistance as a function of flap position and air flow and rpm?
-at what flow or rpm does it peg and the computation stop using it?
-how do they behave when the fail?
What are typical injector on times as a function of rpm and load?
It would be nice to have good plots for temp sensors of resistance vs temp.
It would also be nice to have a place where people could post a concise list of problems and solutions as they have problems with their systems and fix them. This forum is o.k. for that but the info gets pretty spread out and sometimes people post questions when they have problems but don't really follow up with the details of what went wrong.
Two other interesting and related possible projects:
1) Take someone's extra L-jet cpu and curve it....Put it on a test jig that allows us to vary mock 'sensors' and plot how it varies injection time vs variations in temp sensors and air flow and 02 sensor etc. These curves would be pretty useful as a starting point to anyone trying to go to an aftermarket injection system.
2) Take apart an extra cpu and reverse engineer it. I'm not sure if there is a software aspect the these early units or not, but if not it should be pretty do-able. Once we had a schematic we could figure out how to modify it. So the guys who are going with bigger intakes and are worried about lean mixture might just be able to put in a tweeked brain rather than chainging to a different injection system. Basically the same idea as chipping a later car.
I have everyting necessary to do these two tasks except an extra cpu and time. If anyone is interested in doing them I would be happy to help.
In addition the the standard information it would be nice to have a place for unpublished information that is hanging around this group such as: How does the AFM really work:
-what is its resistance as a function of flap position and air flow and rpm?
-at what flow or rpm does it peg and the computation stop using it?
-how do they behave when the fail?
What are typical injector on times as a function of rpm and load?
It would be nice to have good plots for temp sensors of resistance vs temp.
It would also be nice to have a place where people could post a concise list of problems and solutions as they have problems with their systems and fix them. This forum is o.k. for that but the info gets pretty spread out and sometimes people post questions when they have problems but don't really follow up with the details of what went wrong.
Two other interesting and related possible projects:
1) Take someone's extra L-jet cpu and curve it....Put it on a test jig that allows us to vary mock 'sensors' and plot how it varies injection time vs variations in temp sensors and air flow and 02 sensor etc. These curves would be pretty useful as a starting point to anyone trying to go to an aftermarket injection system.
2) Take apart an extra cpu and reverse engineer it. I'm not sure if there is a software aspect the these early units or not, but if not it should be pretty do-able. Once we had a schematic we could figure out how to modify it. So the guys who are going with bigger intakes and are worried about lean mixture might just be able to put in a tweeked brain rather than chainging to a different injection system. Basically the same idea as chipping a later car.
I have everyting necessary to do these two tasks except an extra cpu and time. If anyone is interested in doing them I would be happy to help.
12-13-2003, 06:38 PM
#37
Nordschleife Master
Thread Starter
Answers.
I've bord, so I may as well answer your questions how.
1. The ECU is completely analog. There is no pracitcal way to "chip" it. If you want to "chip" an L-jet ECU, then you are basicly going to have to replace it.
2. The AFM's voltage is on a log curve toair flow. It does not varry with rpm. Airflow will varry with rpm, so it makes a convient measurement, if you run it at WOT. The AFM returns up into 7 something voltes around 3000 to 4000 rpm at WOT depending, and almost apperes to hold. However, there is a slight change in voltage up till around 5,500 rpm, and the AFM just about maxes out at around 5,500 rpm, WOT on a nearly stock 83. Which is intersting because thats the max hp point on an 83. The ECU never stops useing it.
3. Assumeing my back of the envelope calculations are correct, anything over 240 hp is going to require modifications to the fuel system. You can't oven the injectors anymore, and there is enough pressure to force enough fuel through the stock injectors, with the stock regulaters.
To increase the car's proformace, the best I can tell, you need to eighter install larger injectors, maybe from a 914, or higher pressure regulaters, maybe off an S4. I think the regulaters are cheaper. From there you'll be required to crank the main spring in the AFM down to lean out the mixture to make it possible to run with the larger injectors. This will increase the air flow rate at which the AFM will max out at. Untill you max the spring out, it should NOT hose up the fuel curve. However, I say should, because I don't know what the max of the spring is.
This should provide better part throtal responce, and mixture control, than an FMU.
I've bord, so I may as well answer your questions how.
1. The ECU is completely analog. There is no pracitcal way to "chip" it. If you want to "chip" an L-jet ECU, then you are basicly going to have to replace it.
2. The AFM's voltage is on a log curve toair flow. It does not varry with rpm. Airflow will varry with rpm, so it makes a convient measurement, if you run it at WOT. The AFM returns up into 7 something voltes around 3000 to 4000 rpm at WOT depending, and almost apperes to hold. However, there is a slight change in voltage up till around 5,500 rpm, and the AFM just about maxes out at around 5,500 rpm, WOT on a nearly stock 83. Which is intersting because thats the max hp point on an 83. The ECU never stops useing it.
3. Assumeing my back of the envelope calculations are correct, anything over 240 hp is going to require modifications to the fuel system. You can't oven the injectors anymore, and there is enough pressure to force enough fuel through the stock injectors, with the stock regulaters.
To increase the car's proformace, the best I can tell, you need to eighter install larger injectors, maybe from a 914, or higher pressure regulaters, maybe off an S4. I think the regulaters are cheaper. From there you'll be required to crank the main spring in the AFM down to lean out the mixture to make it possible to run with the larger injectors. This will increase the air flow rate at which the AFM will max out at. Untill you max the spring out, it should NOT hose up the fuel curve. However, I say should, because I don't know what the max of the spring is.
This should provide better part throtal responce, and mixture control, than an FMU.
12-15-2003, 08:46 AM
#39
1981 Here....... Been all through the FI system so I'd be happy to contribute.
I would also like to see a modifications section. Maybe John to contribute his findings on converting from AFM to MAF with the Bastard...... That would be a great addition to a solid resource on L-Jet.
Scott
I would also like to see a modifications section. Maybe John to contribute his findings on converting from AFM to MAF with the Bastard...... That would be a great addition to a solid resource on L-Jet.
Scott
