EHA Valve Adjustment
 

Anyone tried this?

https://www.youtube.com/watch?v=rdQUP0VAgHk

Curious about results.

Without watching the vid.. yes the EHA is adjustable. It can be used as a band aid for poorly performing injectors and / or vacuum leaks. A little goes a long way. I don't adjust more than 1/8th turn at a time.

I can't believe I never knew about this trick. I'm almost certain I have vacuum leaks, but haven't had the time to interrogate.

A follow-up questoin. Won't this mess-up my Lambda setting?

I did this to mine and it made a huge improvement. Great drivability.

It wil change the lambda by a significant degree. Hence the very small adjustments at one time

Actually, no. Provided that the adjustment that is made does not take the A/F ratio outside the operating window of the feedback system. While the mixture remains within that window the O2 sensor will drive the ratio to stochiometric.

Fascinating. So, the computer will compensate? Very cool.

I've gotta give this a try.

PS. Wow, I just realized that I've been a member here for almost 16years. Time flies.

I would urge against any EHA adjustments until vacuum leaks are eliminated.

RayH

Ray, can you please explain why? Thx.

The fuel metering depends on the MAF (Mass Air Flow Sensor - the mechanical flap in this case) being able to measure the mass of air entering the engine. In fact, all such systems depend on being able to meter the mass of air entering the engine.

So, all air must pass by that flap if the system is to "know" the mass of air entering the engine. If any air gets into the engine by not passing the flap, it is not metered by the system and the amount of fuel entering the engine is unlikely to be appropriate to prevailing conditions.

Vacuum leaks can allow un-metered air to enter the engine, so any attempt to set fuel delivery is thwarted from the outset.

Does that make sense?

I found that certain driving conditions led me to vacuum problems. High speed driving was perfect. Idling, particularly when hot, was the most testing time for my vacuum-leaky system.

RayH
PS: I use the term MAF in this context because that is what it does. Some say that the term MAF should only be applied to the hot-wire/hot-film devices in later systems.

I'd like to add a comment that hopefully won't be rubbing peoples noses in the dirt.

The term vacuum leaks is in some ways a bit misleading. The important thing to ensure is that there are no leaks in the rubber piping that goes to and from the air idle control valve, the crankcase and the air flow / mixture sensor (plate) assembly.

As explained above the air flow plate is just a mechanical balance that is connected to a plunger that allows more fuel into the fuel distributor. The amount of air drawn past the air flow plate is meant to be proportional to amount of air drawn into the cylinders...

...now the way Mercedes (well Bosch actually) have set up their air balance is that it is designed so that the engine will struggle to idle! To allow the engine to idle air is diverted behind the air sensing plate via the (diversionary) pipes that go to the air idle control valve...

...the important message concerning "vacuum leaks" is that if this pipework (that by-passes the air flow sensing plate) leaks then you get what is called "false air" in the system. This buggers everything up! Well it doesn't bugger up absolutely everything but it certainly messes up the fuel air mixture and confuses the ECU.

Yes, I agree entirely. I have recently replaced all those hoses. However, they are not the only source of un-metered (or false) air. There are many others including all the various tubes that hang off the manifold and also the seals on the fuel injectors.

Loads of potential, in fact, for this rogue air to screw things up.

Another thing I'd like to add. These cars benefit hugely from attending to the basic, low cost items such as HT, vacuum leaks, clean electrical contacts, smooth linkages etc. So many people with bad idle or misfiring assume immediately that they have a major problem with ECUs, fuel distributors, EHA adjustment etc.

My mantra is, "Start with the basics before getting morbid and thinking the worst".

Mind you, just because you're feeling optimistic doesn't mean your car isn't b@ggered.

RayH

Perfect Sense! Thx.

That's exactly what I've been experiencing - hot idle is problemmatic. Now, I did just replace my voltage regulator, distributor cap & rotor, and a couple of plug wires, so we'll see how she performs once I get my serpentine belt fixed.

Can anyone point me to a schematic, video or instructions for identifying all of this piping. I'd love to replace, but I really want to know what I'm doing before I start.

After being away for a month I had a no-start problem. Now, I had been having rough idle for the last several months before being out of town for a month. Wow, you should have seen the inside of the distributor cap. I'm not sure how piston #3 even fired given the state of those connectors and the #3 plug wire. So, yes, I totally agree that good electrical connections are paramount and perhaps the easist fix.

All of the diagrams and the information you need is in the FSM

This is free to download on the startekinfo site

Parts diagrams are available on the MB Atlanta site - alternatively for an annual fee you can pay for real EPC also via startek

To comment on both your points:

Assuming you've got best quality HT components now, all should be well there. Did you replace the shield behind the rotor arm? To recap, the distributor cap, rotor and shield are rather expensive but I bear in mind that my originals lasted 19 years and were still working so I can't really complain.

For cap and rotor, MB ONLY unless cash is very tight in which case Bosch/Beru MAY be OK - but ONLY Bosch/Beru.

I now have all MB aside from HT leads that are Beru - after 5 years all seems well.

Also, non-resistor plugs are the best choice. Most aftermarket websites suggest resistor types by default. I use Bosch (very rare now), Beru or NGK. These are the cheap, copper-core plugs.

Continued hot idle problems may be the injectors leaking, especially if any misfire clears after a minute. In my case, the hot start would lead to misfiring until the leaked fuel had burnt off; thereafter OK. So, new injectors. My M104 engine demanded injectors of about double the price of the earlier M103 injectors. So, I parted with about €360 for the injectors and, with the seals and guides, €500 in total. Not cheap but the end result was most satisfying.

Also, make sure your Idle Control Valve is operating correctly. If it doesn't close properly, it will introduce un-metered air and that's bad as we've already discussed. Some people report that you can flush out the ICV with WD40, petrol or other solvents and all is well. Worth doing.

Once you have eliminated all the problems we've now discussed, make sure your temperature sensors are good by measuring the resistance at cold, medium and hot and comparing those values against a chart that you should be able to find somewhere on the Interweb. You have a coolant temperature sensor that sends signals to the engine ECU. You may have an air intake temperature sensor. Both these influence fuel mixture so make sure these are good before attempting to adjust duty cycle.

Buy a multimeter with Duty Cycle as this makes things simpler.

Adjust Duty Cycle to around 50%, as recommended. Read up on this because the procedure will depend on your car: Kat/non-Kat etc.

In my case, I set the Duty Cycle a bit higher than 50% as this made everything more responsive. In theory, the Duty Cycle should change by no more than 10% between idle and 2000 rpm. However, in my case, it rises from about 55% to over 65% - more like 20%. Having said that, the car goes really well so who cares.

As has been said already, these cars are old and certain adjustments are sometimes fiddled with to compensate for out-of-spec components elsewhere. I'm sure that's the case with mine. A good example of things being a bit awry on mine is that I get about 35 mpg on a motorway dash of 500 km. Pretty good. Around Paris, with endless stop/start, about 14 mpg. Not good. Maybe if all associated components were new that would change.

Bonne chance.

RayH