Converting FI to dual Webers

MAS sensor with 4 throats

I think that my ignorance is showing. I still do not understand how a MAS system would work with four separate throats on the intake system. There could be a manifold over them with the sensor at the entry of the manifold, or a sensor in each throat, or a sensor in one that has a multiplication factor in the computer. What works here?

Paul D.
 
Why? Case for FI

How difficult of a job is this? Would it be much better to find an earlier, non-FI car if I planned on converting to dual Webers, or is it a simple swap either way?

Conversion is easy enough, but why?

I had the same situation; I bought a charity auction '87 FI car that I later found out, the PO had set up with DCOEs for track duty, and for the auction had quickly slapped the FI back on. After the auction I bought the DCOEs seperately from him, and thought I'd put them back on. 8 years later Carl has my DCOEs on permanent loan, and I'm very happy with the Bosch FI.

I drive the car frequently (when there's no salt on the roads), and love the instant, no-fail start up. Any sounds from carbs would be drowned out by the roarty exhaust I installed (I know, I have a dual IDF Spider.)

My X came with a hot cam (of unknown specs) which causes some poor running during warm-up, but after that it doens't stop. In 8 years I've never had an FI related repair issue. Whereas the IDF Spider went through a period where every drive was plagued by dirt in one of the idle jets. Fixable by the side of the road with the tool kit I bring, but not fun. In the X I don't carry a tool kit.

Not a sermon, just a thought.
 
I think that my ignorance is showing. I still do not understand how a MAS system would work with four separate throats on the intake system. There could be a manifold over them with the sensor at the entry of the manifold, or a sensor in each throat, or a sensor in one that has a multiplication factor in the computer.

Not sure what a MAS is...

I'm referring to Mass Air Flow, or MAF-- also known as hot wire.

It's an extremely effective means of measuring the mass of air consumed by the engine. It would go inline between the air filter and the plenum that feeds the intake runners, whether through a single throttle plate or ITBs.

A MAF doesn't care whether it's measuring the airflow going to one throttle plate or several. Actually, neither does an AFM such as the one fitted to injected X1/9s. As an example, my M3 came from the factory with ITBs and a single AFM.

Newer cars use a similar physical arrangement, whether they have a single throttle plate as fitted to most every transportation appliance, or ITBs like those used on performance cars. They just use a MAF to measure airflow, rather than an AFM.

In order to install a MAF/ITB system on an X you'd just make an airbox that fits on top of the ITBs, similar to an air filter housing. Volume of that box is important, but a topic for another discussion. The MAF would either attach to the end of that airbox with a hose leading to the air filter housing, or be installed inline in an arrangement similar to the stock AFM.

Hope that helps,

///Mike
 
Hi Erwin,

No doubt the FI is much easier to live with and more reliable on a daily basis, if you're using the car regularly. It probably provides plenty of performance for most people under most circumstances, too. It's just that (as you probably know from the IDFs on your spider) dual Webers are a lot more "fun". IMO, having Webers on an X1/9 goes a long way towards making it truly seem like a "baby Ferrari" and for a car that's only taken out for an occasional track outing or an inspiring drive through the canyons, their quirks are tolerable.

Anyway, that's my :2c:

Regards,

Larry
 
Fuel Injection Performance Upgrades

Mike, you have certainly touched on some very interesting points here. It is clear that you have put quite some thought into the subject. I too have been researching how to improve the stock fuel injected engine.

Clearly one of the first things that needs to be done to the engine (before modifying the fuel injection itself) is to raise the compression and change the cam shaft: both of which severely limit the performance of the motor. I believe that this should be done first as any subsequent modifications will depend on the characteristics of the engine. As you mention, adding a higher lift cam doesn't seem to work too well with the standard Bosch system at idle, so this is one thing which needs to be addressed in the FI modifications.

I would also note that an excellent source of information on the L-Jetronic and later Bosch systems is "Bosch Fuel Injection & Engine Management" by Charles o. Probst.

There are a number of good choices in aftermarket EMSs, of which Megasquirt is certainly the most popular, presumably primarily because of its price. When Megasquirt entered the market there were very few EMSs offered, and those that were available were rather pricey. Megasquirt literally brought programmable EMSs to the masses. There are quite a few more options now, so it's much easier to find something that meets your specific goals. FWIW, the Autronic units look to be very interesting, although they are significantly more expensive than Megasquirt.

Motronic

I would like to investigate the conversion of an X1/9 from L-Jetronic to Motronic. It seems that this modification has been done by several BMW owners (although they are lucky in that later model BMWs used Motronic). In the case of the X we would need to add a RPM and TDC sensor on the front of the crank shaft and swap the distributor with one without any form of mechanical or vacuum advance. The advantage of this swap is that it is relatively easy to set up, parts are cheap and you end up with control of both injection and timing. As the Motronics are digital, the timing and injection maps can be modified.

Modified L-Jetronic

I'd also note that there are a couple of mods that you can perform on the standard L-Jetonic system to provide some extra performance (within limits). The most popular is to increase the resistance from the coolant sensor which tricks the "computer" (it isn't really a computer in L Jetronic, but let's call it that for now) into thinking the car is cooler and therefore supplying more fuel. A similar mod can be done on the air temp sensor (mounted in the AFM). I don't believe that either mod offers better performance in all circumstances and therefore I have made a console which allows me to "dial in" increased resistance to both inputs (as well as being able to switch in the full throttle switch before full throttle). Again, these mods can't really be considered a viable long term solution, but they do provide an interesting lesson in how the L-Jetronic responds to modified inputs. Note that significant mixture enrichment can be detrimental after a certain point as the Lamda sensor will not be able to compensate and the catalytic converter can go past its operating temperature range. There is also an AFM mod that you can do, as documented here:

http://www.hiperformancestore.com/Ljetronic.htm

I'm also running a fully programmable ignition system with no mechanical or vacuum advance in the distributor (timing curve depends on RPM and manifold air pressure, measured by a MAP sensor).

Just to clarify, the X1/9 EFI (Bosch L-Jetronics) uses an airflow meter (AFM), not a mass airflow meter (MAF). AFM is old technology, but MAF is used on virtually every injected car made today, including most of the top line exotics. IMO it is an excellent means of measuring airflow, although most DIYers seem to use MAP (Manifold Absolute Pressure).

Some Xwebers are running 40/80 or wilder cams with the stock AFM, although most of them will admit it's not ideal. AFM uses a spring loaded plate to measure airflow. That plate pulsates with the intake reversion caused by a high-overlap cam, which presents a number of problems at lower RPM.

Are you sure that it is the AFM which is the problem here and not other characteristics of the standard L-Jetronic configuration? I'm planning to install a Euro 1500 cam in my FI X and will be researching how to negate any negative characteristics of this configuration.

On top of that, the AFM literally presents a "barn door" to the column of air moving through the intake tract. This interruption of the intake charge inertia has a large detrimental effect on power output, particularly torque-- the engine is not only being strangled, but much of the "inertial supercharging" effect of intake resonances is negated. IMO, that is the single biggest drawback of the stock injection system.

This is a common criticism of the AFM system (and clearly a hot wire system is superior in every way) although I have also read that the effect is not as great as many are lead to believe.

However, you could either replace the AFM with a MAF (my preferred solution) or remove it and use a MAP or other means of measuring airflow, but still retain the rest of the original components (other than perhaps injectors), and integrate that with the aftermarket EMS of your choice.

If you replaced the AFM, EMS and injectors, you wouldn't have much of the "original components" left to retain! :wink2: BTW, using a MAP sensor is even worse than an AFM (according to Bosch anyway who moved away from this technique after the D Jetronic system used in the early 70's). Manifold air pressure is dependant on a number of corrections for temperature, altitude, etc.

This would remove the intake obstruction presented by the AFM, which would result in instant power gains all by itself. It would also solve many of the problems presented by a more aggressive camshaft, at least with regard to the EMS.

Agreed, but I think that the whole problem needs to be addressed together: air density measurement, injector timing and ignition timing. I'm thinking that starting with a pre-configured Motronic might be an easier starting point than Megasquirt where you basically start from scratch.

My own plan is to do just that initially, in order to create a good, simple base mechanical installation so we can test out and become familiar with the particular EMS we'll be using. Once I'm comfortable with how it all works I hope to move to an individual throttle body (ITB) arrangement, both for the extra power and for the sound. I intend to use a MAF for both arrangements because I prefer the advantages it offers.

I have no data to support the following statement, so take it with a rather large grain of salt, but I suspect that a well tuned aftermarket EMS using MAF and running through a nicely ported, port matched, and polished stock plenum/runners would probably yield roughly as much power as the average DCNF setup.

Good luck, I'm looking at doing something similar (i.e. a well tuned fuel injected car which still complies with the relatively strict rules we have here in Germany concerning modifications and emissions). I think that we differ slightly in the implementation approach, but the goals certainly seem similar. It would be interesting to swap notes when the projects are complete.

Cheers,
Dom.
 
Motronic

I would like to investigate the conversion of an X1/9 from L-Jetronic to Motronic. It seems that this modification has been done by several BMW owners (although they are lucky in that later model BMWs used Motronic). In the case of the X we would need to add a RPM and TDC sensor on the front of the crank shaft and swap the distributor with one without any form of mechanical or vacuum advance. The advantage of this swap is that it is relatively easy to set up, parts are cheap and you end up with control of both injection and timing. As the Motronics are digital, the timing and injection maps can be modified.

For those who may not know the history, the Bosch Motronic system came after the Larrytronic but before the Curlytronic.

:laugh::laugh:

Pete
 
Wise guy!

:wink2:[ame="http://www.youtube.com/watch?v=AKtwlHV1-O8&feature=related"]YouTube- three stooges nuk nuk nuk[/ame]
 
No doubt the FI is much easier to live with and more reliable on a daily basis, if you're using the car regularly.

Interesting you should say that; I fully expected to have to clean and tune the entire IDF assembly every spring, but to my surprise the car starts right up and the carbs haven't needed a synching in many years. They do foul the plugs at times, have a hard time starting in the cold, and get their idle jets plugged quickly, and have a tough to cure flat-spot just off idle. I still don't think the carb sound is as good as a roarty exhaust sound.

I was hoping Carl would chime in too, since he used to swap out his IDFs to FI and back for the bi-annual smog check.

Good luck with your project, any which way it goes.
 
Motronic on a 128

you do know that the FI Yugos came with a Motronic EFI setup, right? It would be the 'easiest' way yo get full Motronic control.

I would argue the point that because Bosch went away from MAP sensors, the AFM is superior. The AFM is superior to the D-Jet MAP that was designed in the '60s. Modern digital MAPs are very accurate and can be made to perform very well (even with cams).

MegaSquirt is NOT that difficult to set up. The current codes offer self learning algorithms that make it a snap. I, for one, have built almost a dozen for FIAT based cars and have had no problems with them. I did have an issue with my DCNF throttle bodies (getting a good vacuum signal), but the later codes address that nicely. Also, with MS you can set up a 'blended' Alpha-N/Speed Density strategy that uses the TPS for low rpm/idle conditions where cams can hurt the vacuum signal, then changes over to MAP based fueling at higher rpms which is more accurate than TPS (especially if you want to turbo it).

Swapping a Motronic into a differnet model of BMW is EASY because they offered the same basic engine with and without Motronic. Adapting one of there ECUs to a different engine (say an X) would REQUIRE the ECU to be reburned to a new code and not many people have access to the software to do that. It would also have to be tuned on a dyno ($$s), where the MS can be tuned on the road while driving (either maually via a laptop or via self-learning).

And yes, the X is very easy to add MS to because it has a manifold/injector available to use. The 128 engine needs almost no acceleration mixture increases for rapid throttle openings (unlike my 2.0L twincam), which really helps in tuning. You really only need to focus on the main fuel map/table and then fine tune the accell enrichments at the last phase.
 
Motronic Question

Is the Bosch Motronic system used on the late Yugos different from the first generation Bosch L-Jetronic system? If so, does it use a Mass Air Flow (MAF) sensor in place of the Air Flow Meter (AFM)? I just never heard of a Bosch L-Jetronic system that used a MAF. If this system is out there, or if there is a way to convert a Bosch L-Jetronic system with a AFM to instead use a MAF sensor, I want to know how to do it, as the MAF system is far superior and much more tune-able. If the system doesn't yet exist, there have to be some really smart guys here who could figure out how to come up with a direct MAF sensor for AFM replacement and any corresponding changes needed to the Electronic Control Unit.
Dave
 
Yes, the Yugo is different than the L-jet. It is a digital Motronic like the BMW and uses an air flow meter BUT the resistance of the Motronic AFMs is OPPOSITE of the earlier L-jet. I haven't confirmed this for myself, but that was the difference between the earlier BMWs and the 535is I used to own. Many of the BMW manuals I have refer to the AFM as a MAF.

IIRC, the L-jet's AFM resistance climbs as the door is opened, the Motronic falls. The motronic's AFM also does not have the fuel pump switch.

The later Motronics have a hotwire MAF (LH-jet) and are common to Volvos, SAABs and BMWs body styles E34 on up.
 
Now this is something I've wondered...

The later Motronics have a hotwire MAF (LH-jet) and are common to Volvos, SAABs and BMWs body styles E34 on up.

In order to eliminate the "doggie door" restriction in the intake tract, how hard would it be to use a hot-wire MAF, and then create a small circuit to convert the MAF signal to the corresponding output from the AFM? I don't know that much about electronics, but a signal conversion does not seem all that hard to me... if it's difficult to do in "analog", then use a AD converter, play with the signal in digital, and then a DA converter at the other end.

It must be much more complicated than I am stating here (or complicated enough that the gain is not worth it), or some clever electronics guy would have done it by now.

Who's an electron whiz around here?

Pete
 
This place is well known in the BMW community for making quality AFM to MAF conversions on older Bimmer engines. I'm thinking about getting one for my '85 635CSi. Maybe if enough FI Fiat guys approached them about a conversion for their cars, they'd look into the feasibility of bringing one to market.

http://www.millerperformancecars.co...ypage=flypage.tpl&product_id=58&category_id=6

Note: On BMW conversions the factory chip has to be replaced with the Miller chip when converting to MAF. Do FI X1/9s even HAVE a replaceable chip in their computer?

I'm no electronics wiz either, just offering this idea up for consideration. :whistle:
 
Pretty sure they don't.

Note: On BMW conversions the factory chip has to be replaced with the Miller chip when converting to MAF. Do FI X1/9s even HAVE a replaceable chip in their computer?

I'm no electronics wiz either, just offering this idea up for consideration. :whistle:

But I'm not talking about hacking way down into the guts of the computer. I'm only talking about adapting a MAF, using some custom circuitry, to output a signal that "looks" like a signal from an AFM. Strictly a sensor change, not a change to the main computer.

Pete
 
Exactly what I was wondering

Surely someone on this forum is electrically-inclined enough to figure out how to do this MAF in place of AFM. Come on all you electrical engineers, let's see what you can come up with.
Dave
 
Yes, it has been done. The issue with hot-wires is that they are calibrated to airflow and determining what a given airflow would relate to AFM door opening may be difficult without the proper tools. I'm not sure if the output of hot-wires are logrithmic, but the AFM is.

Also, keep in mind that the AFM 'door' maxes out at max torque (opens no further) while a hot wire would continue to send feedback on increased air mass. I'm not entirely sure the simple ECU in the X would be able to calculate that correctly.

It would be much easier to use a MAP sensor and calibrate it's output to simulate an AFM. I have a circuit at home to do this (found in a Miata hop up book).

If you really wanted a MAF instead of a MAP using MegaSquirt, some of the codes do allow the use of one in place of the MAP sensor.

In the end, you may spend just as much to develop a 'fix' for the early Bosch ECU that still isn't adaptable to changing conditions/tuning than to just buy a modern digital system (MS isn't that much $s and will do a lot more than you will ever want to do with an X).
 
Agreed, the right "fix" is an aftermarket controller....

In the end, you may spend just as much to develop a 'fix' for the early Bosch ECU that still isn't adaptable to changing conditions/tuning than to just buy a modern digital system (MS isn't that much $s and will do a lot more than you will ever want to do with an X).

And I suspected it would be difficult, or the results not worth the hassle. Otherwise someone would have offered a kit by now. I'm not even really that sure that the removal of the doggie door will offer that big a performance gain, I think someone once posted some information that suggested it was not the parasite most people see it as.

So I assume the MAP conversion you refer to would remove the doggie door and supply a signal to the ECU to simulate it? Not sure if I am reading your post right. If so, do you have any more info on that conversion?

The MS route is a plunge I will take before too long here, but by many accounts the learning curve is steep. I know you (and others) have done it, but the sheer volume of information (or mis-information, in some cases) out there is frightening.

Pete
 
I'm not sure if the output of hot-wires are logrithmic, but the AFM is.

Yes and no; According to the Bosch L-Jetronic Technical Instruction document the relationship between air flow and flap angle is logarithmic, but the resulting output voltage is inversely proportional to the air flow.

As for converting the signal from a MAP sensor to give a AFM compatible output, you'd need one more signal to get there. Either a throttle position input or simply an RPM signal would get you close, perhaps even close enough.
 
Back
Top