Hi,

I am thinking hard about fitting aftermarket engine management to my 190E 2.5-16.

I have heard one foreign engine tuner claim they saw a gain of something like 35bhp on a 2.5 with no changes other than a modern management system. That may be farfetched (?) but I want to go to EFI anyway. I don't plan any other modifications.

I understand the KE fuel injection is mechanical, so I will need to fit electronic injectors. How would I go about this? Do I need to fabricate a fuel rail, and how do I mount the injectors?

I have no catalyst and no lambda sensor. Will I need a lambda sensor? Does it need to be wideband?
I understand I'll need to fit a new crank position sensor (even though I already have one)?

What other things does the ECU on a M102 or M103 control or read, that I need to think about? What about the engine fan? My car does not have a/c so that's one less thing to worry about.

__________________
190E's:
2.5-16v 1990 90,000m Astral Silver
2.0E 8v 1986 107,000m Black 2nd owner
http://www.maylane.demon.co.uk/190esmall.jpghttp://www.maylane.demon.co.uk/190esmall2.jpg

Silvercosworth over at 190revolution.net sells a complete EFI conversion kit using a Megasquirt ECU. His email is cosworth2316@hotmail.com Tell him Ortolan referred you and get a discount.

Ah yes, I am aware of this.

However I understand megasquirt is rather "DIY" and I'm not experienced enough around engines to be confident with that?

Also, hasn't he been missing for 6 months??


P.S. When I say lambda sensor I mean O2 sensor!!

__________________
190E's:
2.5-16v 1990 90,000m Astral Silver
2.0E 8v 1986 107,000m Black 2nd owner
http://www.maylane.demon.co.uk/190esmall.jpghttp://www.maylane.demon.co.uk/190esmall2.jpg

Yes, that sounds about right. Open up that throttle body and you'll get even more!


This is what I did for mine. It's as simple as installing injector bungs and brackets for the fuel rail.
http://www.peachparts.com/shopforum/attachments/mercedes-benz-performance-paddock/32830d1145671661-advice-m117-560-cis-efi-conversion-attn-supersec-efi-m103.jpg

You don't need a cat, you do need an O2 sensor. Wideband is prefferable but narrowband will work. It is recommended to at least tune your mixture maps with wideband O2 and on a dyno. Some aftermarket ECU's will work with your stock crank sensor such as Autronic, some will not.
www.autronic.com


Lots! You need to add a TPS to your Throttle body. You need a MAP sensor, Temp sensor to replace the stock sensor. You can use your existing Idle air valve since it's PWM. You'll need to scrap your EZL for a GM one or just go distributorless. The fan is easy, just a wire from the ECU to the fan relay.

Anyway the list doesn't stop here. If your serious about this project, get this book! It will answer all your questions and will save you lots of time and money.
http://www.amazon.com/Building-Tuning-High-Performance-Electronic-Injection/dp/1884089798

__________________
Current Stable:
01 ML55 AMG
92 500E (a few mods)
87 300E (lots of mods)
00 Chevy 3500HD Diesel Box Truck
68 18' Donzi Marine
06 GT i-Drive7 1.0 Mountain Bike (with GPS!)

PREVIOUSLY OWNED:83 300SD, 87 420SEL, 88 420SEL, 90 420SEL, 86 560SEL, 86 190E 2.3-16V AMG, 94 E320

If you want some real NA power ask Dan16V if he'll build you one of these babies!
FINALLY!! 16V Custom Throttle Bodies---DONE!!

__________________
Current Stable:
01 ML55 AMG
92 500E (a few mods)
87 300E (lots of mods)
00 Chevy 3500HD Diesel Box Truck
68 18' Donzi Marine
06 GT i-Drive7 1.0 Mountain Bike (with GPS!)

PREVIOUSLY OWNED:83 300SD, 87 420SEL, 88 420SEL, 90 420SEL, 86 560SEL, 86 190E 2.3-16V AMG, 94 E320

Thanks 300EVIL, nice to know some of the great mercedesshop knowledge is still here! I thought I was going to get no replies for a bit..

I was going to get the engine management system to handle the igniton as well. But there's no need to go distributorless is there??

__________________
190E's:
2.5-16v 1990 90,000m Astral Silver
2.0E 8v 1986 107,000m Black 2nd owner
http://www.maylane.demon.co.uk/190esmall.jpghttp://www.maylane.demon.co.uk/190esmall2.jpg

No problem Pentoman, Glad to help!
There's no "need" to go distributorless however, there are advatages.

No timing adjustments. (Don't remember if the 16V has vacuum advance..)
No distributor cap and rotor.
No moving parts to wear out.
No distributor to accumulate moisture and cause starting problems.
No distributor to drive thus providing less engine drag.
And a hotter spark thus more rapid burning combustion.

Unfortunately going distributorless comes with an extra cost. About $600...

Not necessary but better than stock.

__________________
Current Stable:
01 ML55 AMG
92 500E (a few mods)
87 300E (lots of mods)
00 Chevy 3500HD Diesel Box Truck
68 18' Donzi Marine
06 GT i-Drive7 1.0 Mountain Bike (with GPS!)

PREVIOUSLY OWNED:83 300SD, 87 420SEL, 88 420SEL, 90 420SEL, 86 560SEL, 86 190E 2.3-16V AMG, 94 E320

Hmm some points there I hadn't thought about.

Others have said to go distributorless too. One guy with an E30 M3 said he chose not to to begin with, but thinks now he should've done it from the start as it would work out better in the long run.

But I'm trying to keep things as simple as possible, especially as I'm very new to this stuff. And the 16v spark wires look just right going to the existing distributor!!

Also, can't the ignition computer (which would be part of the new system) perform ignition advance duties even if it's running through the distributor?

I haven't had any suggestions on what needs to be done with injectors/fuel rail yet. This is my biggest stumbling block.

__________________
190E's:
2.5-16v 1990 90,000m Astral Silver
2.0E 8v 1986 107,000m Black 2nd owner
http://www.maylane.demon.co.uk/190esmall.jpghttp://www.maylane.demon.co.uk/190esmall2.jpg

Well, the ignition module is not part of the aftermarket ECU but the outputs for them are. Yes, the ECU will take over any advance or retard of ignition timing.

Did you see the pictures in my first and second post??
http://www.peachparts.com/shopforum/attachment.php?attachmentid=32830&d=1145671661

__________________
Current Stable:
01 ML55 AMG
92 500E (a few mods)
87 300E (lots of mods)
00 Chevy 3500HD Diesel Box Truck
68 18' Donzi Marine
06 GT i-Drive7 1.0 Mountain Bike (with GPS!)

PREVIOUSLY OWNED:83 300SD, 87 420SEL, 88 420SEL, 90 420SEL, 86 560SEL, 86 190E 2.3-16V AMG, 94 E320

The correct answer to what must be your real concern; how much more power will your engine make? is NONE. Read this answer please in reply #16, I think to the DAN16 thread about the home-made throttles. This fellow actually did this conversion, so now you can have the correct answer. Now that you know the car not going to accellerate any faster or top out any better, do you still "want efi".? why? You see the engine doesn't know where the fuel came from; it doesn't care if the little gas sprayers are mechanically controlled or electrically controlled. The original bosch system you have is capable of good fueling accuracy; of course, or it wouldn't be on the car in the first place. For you or anyone who goes aftermarket: I don't like to see you have a computer controlled timing curve; for a very good reason, the directions that come with your controller are probably not perfect and having complete controll over the ign. timing is too dangerous for an amateur enthusiast. I laugh when I write this; last year the previous year's winner of the USA Engine Masters, engine building competition destroyed his engine in the final round of dyno tests, by adding on 3DEGREES of timing before the final pull for all the marbles. FULL control of ign. timing should probably be reserved to the OEM's. Enthusiasts who do their own efi should obtain some kind of mechanical distributor with a vacuum can, even if they have to find a machinist to put the top half of different make distributor on their stub and drive. I admit this is only an opinion; but to say that more engine damage is caused by incorrect ignition timing than any other single factor, is not an opinion. People have been looking at scuffed pistons and broken rings for years and nodding wisely as they proclaim, yes it must of gone lean". For every engine damaged y running lean there are ten damaged by advanced timiing. An msd #6 spark box and a square frame coil from practically any ford product from 1988 to 1996 or something; will make sparks that'll light wet newspaper;(well enyway, real good sparks!) If the coil is square and black and came from a junked ford you just saved 60 to 100 dollars and no compromise. Distributorless ignition has certain advantages for the OEM, but won't do a thing for you. The way engineers spell better is; "cost/benefit", the cost will be there all right, but there's no benefit. ps. I know there are real professional tuners in england, and your factory injection IS TUNABLE. (many people think it isn't). Cheers, Jack. and have a courage pale for me.

But isn't the fuel distributor plate a restriction in a K-jet system? Besides that, while it is possible to tune a K-jet system, it's not very easy.

You can run fuel injection without a lambda sensor, but it will be harder to tune. D-jet systems ran without a lambda sensor, and while they weren't tunable (you had to keep the intake and exhaust nearly stock, and don't think about changing the camshaft), the system worked very well for its time.

Megasquirt is similar to D-jet, using pressure, temperature and engine speed to measure the air intake. I'm sure that there are commercial designs which are similar. These are harder to get right than MAF systems, but would be fine on an older car with less-strict emissions requirements.

The first thing people think when they look at the Bosch mechanical air mass device; is oh dear it must be an air restriction. It offers less air restriction than an oversize four barrel carb. Look at it carefully, at high flow levels it "gets out of the way". It was designed by very dedicated professionals. It's a clever piece of equipment. Can the home enthusiast produce a better fuelling curve by fiddling around with his after-market box? Very un-likely. I've had two cars delivered to my shop on car trailers in the last year; they were both supposed to be improved by the addition of their after-market controllers. One was a corvette, and the other one was a Dodge. If they had said Mercedes on the front the results would have been exactly the same; an engine is an engine, and knowledge is knowledge and neither one of these cars was drivable due to months of owner and friends and volunteer experts, "tuning" them. Ibuild fuel controllers and convert cars that don't have 02 sensors or Throttle position sensors and all is well; or even better then well. It's certainly true that an o2 sensor would generate usefull information for the amateur tuner; for a professional its not necessary. In no case is it necessary or desirable to use the 02 sensor during normal operations. It exists purely for the sake of the smog control/CAT package; nothing else. I've seen two cars in the past year or so that had failures in the oem engine controller ignition coltroll module. One chevrolet sedan and one MB. In both cases the erratically retarded timing effected the performance. Since I have to make these cars perform properly and I don't have forever, or an infinite budget; I converted both of them to mechanical distributors. The owners are very happy. This type of problem probably exists in other cases, and it's very difficult to diagnose; If you're pretty sure your early computer mandated timing car doesn't perform properly; and particularly if it is intermittent, and seems to have bad hair days; you might want to change to a mechanical distributor, with vacuum can. Part of the advertising, and the popular opinion about the modern engine controllers is the wonders of the precision timing control; that adjusts continually to match this that and the other thing. This is basically non-sense. Asimple, conservative mechanical advance curve, with a reasonable roll-on rpm, for your application, and a conservative, (usually a mechanically restricted vis a vis stock), vacuum can will offer excellent performance.

The mega squirt controller doesn't function similarly to the Djet system at all. All systems are tunable. If you have knowledge you can make any of these systems do whatever you want it to. I can put any kind of cams you want in your D-Jet car and make it perform perfectly. Without complete knowledge of the subject of internal combustion engines and tuning; we fall victim to the oft-repeated cry, "it isn't tunable" . To clarify what the real issue here is; I have taken recently to answering people who tell me, "darn this EFI, it's all weird and electronical, so I can't fix it", the following, " you can't fix a curburetor eaither". I look 'em right in the eye and just tell em. You know I spent a huge amount of time studying textbooks and following older tuners around; being the student,k jdoing the scut work. I paid retail for my knowledge. And this statement above is literally true'; and I know it's true, to them, "A CARBURETOR IS UN-TUNABLE". This is the actual issue; knowledge. There is a top notch performance tuner in every geographic area; and if you really feel the cars not delivering the way it should; seek out professional diagnosis first; before you start re-inventing the thing. An interresting point is the claim quoted in the original note that some engine made 35hp more merely by switching to a modern control system. Of course this is easily possible provided that the engine was badly de-tuned to start with; but we are meant to think that a properly tuned engine developed 35hp owing to its change of controls. Here's the way an engineer will examine this claim. What does 35hp mean? It means you're burning an additional 17lbs. of gas. an hour; and in order to do that you have convinced this N.A. engine to inhale and process 3.64lbs, per minute, of additional air; over and above what it did previously. Secondly, it might mean that you had caused an enormous inscrease in thermodynamic effeciency; not possible. Neither one these outcomes is possible; hence it's just more conversation and we can safely ignore it. I';m just putting some observations from real life up here; it's not for me to tell any one what to do; some people succeed easily in their efi conversions and some people end up with un-drivable cars. The after-market vendors are not going to tell you this. Some people can read the tuning instructions that some with the thing and go directly to bingo; some of them have instructions that no-one could read or use. I would not like people to think that converting a non-airstarved, performance engine to modern efi is going to cause some wonderful increase in pereformance. It will not. If it does; it means you fixed a $200 tune-up problem for $2500. Cheers, jack

Hi John,
Man you are long winded.... However, your obviously an experienced old school tuner and I appreciate your knowledge and opinions.

Bottom Line: Can you get more power by switching CIS to EFI....

Yes you can!

How? Okay,,,, as Matt L stated the metering plate does cause negative pressure, not a lot, but none the less it's there. It's been proven that factory intake housings and tubing can even cause negative pressure. You don't think they flute the ends of a Hilborn manifold because it looks cool.

Okay so what? Maybe 2-5HP removing this restriction if your very lucky....

Yep! Now lets move on.

Timing, Yeah you can grenade your engine if you don't know what your doing or your using inferior circuitry to do so. However, The latest aftermarket EFI systems out there do a very good job, way better than any relay in a distributor. Some even use knock sensors to sniff out a problem before it becomes destructive and will automatically retard the timing. Thus the novice can easily tune their timing without catastrophic failure. Given you have everything installed correctly and are tuning on an eddy current dyno it's as easy as adjusting the timing on the particular site on the map to maximum HP then back off a few degrees. Piece of cake! So with accurate and safe ignition timing figured in, we now have maybe 15-20 extra HP!

Next issue, The fuel injection method. If CIS and carbs were so good, why aren't we using them anymore???? Well, first off emissions restrictions but second is economy and performance. As we all travel foward in time newer and better things come around. I'm sure in 20 or so years we'll be discussing how we can make more voltage/current from our hydrogen fuel cells to make our electric cars go faster!

The key to EFI being superior is the INPUTS....
The more the computer knows about what's going on the better it can meter fuel and adjust timing. A carb is a mechanical computer, It monitors airflow and density and even has a nifty little accelerator pump to squeeze off a little fuel during sudden acceleration,,, But thats about it. CIS is about the same. While carbs and CIS are great, they were a stepping stone to EFI.

EFI relies on much more accurate numbers from its inputs, and there's a lot more inputs! With all these variables now being seen by the computer it can now way more efficiently meter fuel and adjust timing. As you know, the stock CIS system was built to maintain stoichiometric mixture; except for WOT in which the EHA kicks in and delivers enrichment thus better thermodynamic efficiency and more power. Sure you could tweak the CIS for this efficiency throughout the load levels but not nearly as precise and effective as EFI.

So, With a mixture somewhere between 12.5 to 14.0 throughout the engines RPM range versus trying to maintain 14.7 there is power to be had!
Maybe 5-15 HP plus a better torque curve!

Play with those numbers and you'll end up close to 35HP....

Conclusion: EFI removes these "sweet spots" you find with carbs and CIS and broadens the performance and efficiency to the whole RPM spectrum.

Obviously Pro Stock tuners like Warren Johnson and Greg Anderson want to switch to EFI because there is a huge advantage. They can't change their displacement but they know EFI is far superior to carbs and will let them tune for lower ET's and higher MPH. This change is going to happen very soon and when it does EFI technology will really take off. These guys have tons of money to throw into R&D.

As for tuning without an O2 sensor,,,, the question would be why???

NASA doesn't just point the Saturn V rocket at the moon and fire! They use computers, mathematical calculations and physics to figure out the correct flight plan.

A good wide band O2 sensor and controller would set you back $500. I built my own for $100. For that insignificant of a cost why try to guess what your mixture ratio is. The ability to tune each load site at a particular RPM to the correct mixture is paramount to drivability much less performance. Again, the more inputs, the better the tune. Knowledge is Power!

You talk about EFI as if there is some mystery to it. You say some people will get it to work and some will not.... Well, if you embark on a project like this and have no idea what your doing, you will fail. Not the engines fault, not the computers fault. Aftermarket EFI is so good now "most" computers will idle the engine right out of the box with no data input whatsoever!

Adam

__________________
Current Stable:
01 ML55 AMG
92 500E (a few mods)
87 300E (lots of mods)
00 Chevy 3500HD Diesel Box Truck
68 18' Donzi Marine
06 GT i-Drive7 1.0 Mountain Bike (with GPS!)

PREVIOUSLY OWNED:83 300SD, 87 420SEL, 88 420SEL, 90 420SEL, 86 560SEL, 86 190E 2.3-16V AMG, 94 E320

There are many issues raised here; it takes a long time to reply to all of them; I'm not a dinosaur fondly at the past; I know about the knock sensors; which are not available to the person converting there 16v2.5. I'm making realistic reccomendations for ordinary people. If everyone could or did tune up their systems on a dynamometer, there would probably be less problems; but they don't. Many people can't afford any such thing. Also the dynamometer without knowledge is useless. Your plan to adjust for maximum horsepower and back off a few degrees is not good. The first 7degrees of backing off results in a 2% loss of power which needs to be accepted in the interests of engine life. 9deg. would be better for a street car, cause you don't drive it on a dynamometer. The combustion chamber temperature will increase as you keep the pedal down longer than you did on the dyno, requiring less advance, and the engine controller will not monitor this. This is way OEM engine controllers are not set to a few degrees less than peak horsepower. I refer you to Internal Combustion Engine theory and design by Charles Taylor, Vol 1&2, MIT PRESS. The life of the engine can be cut in half by the doubling of peak pressures and peak temperatures, (roughly speaking), that occurs within these magic 7degrees; without detonation ever occuring. To make a long story short, the engine under discussion will not inhale 3.64 lbs. mass of air per minute more than it did originally, no significant increases in ni, (thermodynamic effeciency), are possible; as you see a variance of 7degrees in timing is only worth 3.4hp or so in this engine; the 35hp. doesn't exist and won't appear. Notce that #18 reply, or thereabouts is from someone who did this and says of the power difference, no real difference. This is what a person needs to know when considering this kind of modification. The progression from cis to electronic engine controllers was driven by economics and emissions regulations. Considering only a non-smog car, such as pentomans car, considering only performance aspects of the situation; the car can be put in top performing tune with out any major modifications. There's an unfortunate planet-wide assumption that anything that doesn't have two micro-processors in it must not be very good. This is not always true. It's very important to understand that advertising=lying, and its everywhere. The performance car magazines "review" a new, expensive fuel controller and rave on about the wonders of "full sequential injection", for instance; These are info-mercials. The writter gets a free controller. The reality; sequential fuel controll is utterly useless to any of these people on this forum. There's no engineering reason to expect to improve performance by sequential injection; and careful A-B DYNO tests reveal exactly what we expect; nothing. I'm not really a normal person. I have a degree in Physics and have studied all the engineering materials relevant to this subject. I'll answer any question you want me to; I'm sorry if the answer doesn't seem agreeable. To me its just a factual, objective, engineering subject; I mean there is one right answer to each correctly formed hard science question; I don't mean to bother anyone by supplying it. Cheers, Jack