Mercedes-Benz Forum banner

1 - 3 of 3 Posts

W124 400E, W126 500SE
971 Posts
Discussion Starter #1 (Edited)
Hi, could someone point me in the right direction? I already wrote about this a while ago but never got clear pointers

The car is 1986 500SE with KE-Jet (EURO)


1) car starts fine, idles fine
2) during warmup, it wants to die when I step on the gas more than a little.
- In idle, there's a hole @ 1500 RPM (when in neutral and full throttle it reaches 1500 RPM, stays there for a while and then shoots up)
- If I step on it too hard in Drive, it dies (very funny at intersections :/ )
3) when hot (and I mean really hot - >90 deg.C) it gets better but still not perfect
4) it drives better when I adjust the mixture to run lean, but then it misses in idle. When I adjust it a little to the rich side, it idles great but doesn't really have power and wants to die in drive
5) oscilloscope diagnostics
- everything looks good but it runs very lean with open throttle. It should get rich immediately but it doesn't
- also the throttle plate potenciometer is shot (gonna replace that, but it shouldn't cause it to die)
6) I adjusted the mixture with X11 pin 2-3 to get 50% duty cycle in idle and about 60% @ 2500 RPM, it sometimes jumped to 80% and stayed there (O2 sensor shot now?) - that was with old wires that caused a lot of misfire so I guess it's solved now. EDIT: it's probably not 50% now, will recheck in the afternoon.

I messed up with the mixture adjustment when I got the car because it did this occassionaly and I used to fix it this way on my 420SE, but now it does it consistently. I changed the air and fuel filters recently, spark plugs and cables were shot so I put in new ones (also had to make new threads for 2 spark plugs, some idiot ruined them), and it made the car run a alot better but the problem is still there.

New potenciometer is on it's way, but I'm also thinking about adjusting EHA.
Also got throttle plate cleaner and I'm gonna clean the gunk out of idle valve, throttle plate and the whole intake manifold.

Has somebody seen this behaviour? Did I miss anything? And does KE-Jet have some sort of adaptation that requires reset or a long drive to settle after all this work? Could the be an ignition advance problem? I didn't touch that (and distributor looks good), but I'm pretty sure it should be set without EZL to some basic position...

Also, how important is the mixture adjustment? (I'm not talking about legality and emissions, but for now I just want it to run - I'll run it through CO check when it's solved). I guess it should run just fine even when misadjusted? And yes, I blame myself for even touching that :/ it's not a shortcut, I've learned...

Thank you

'73 450SL, '83 300CD, '01 E320 4matic
2,446 Posts
You have to understand how the mixture adjustment works with CIS-E.

The % duty cycle is NOT critical. There are probably some people who will argue that point, but it's just a fact.

Once the engine is at operating temperature and the controller goes closed loop, then the engine is controlling the mixture through the EHA valve. It doesn't 'care' what the duty cycle is set to (within reason).

The O2 sensor 'reads' the fuel mixture. The controller interprets that information (and some other information, such as coolant temp, throttle position, etc) and modifies the mixture with the EHA valve. Set at 40% or set at 70%, there is enough available 'swing' left for the EHA to control the mixture. In theory, with a 50% duty cycle set, then basically you allow the controller the ability to compensate equally in both direction. I hope that makes sense.

The controller's goal is to adjust the mixture to the ideal parameters stored in it's program.

Therefore, the best way to adjust the mixture is for the best drivability. That may not necessarily be at 50%, and there's nothing wrong with that. If you have to set it outside of a range, maybe 35% to 65%, then perhaps you need to do some digging.

The number one cause of drivability problems on CIS systems is false air, ie vacuum leaks. You have to find them all, and fix them all. Any time air can get in where it doesn't belong, and pass through the combustion process, it's going to end up modifying the mixture and the O2 sensor is going to pick it up. Air generally makes the O2 sensor interpret the mixture as lean. If the controller sees the mixture as lean, it's going to keep trying to richen up the mixture. If you have a bad injector that is not fueling a cylinder properly, and that cylinder is pumping mostly air, that air is going to cause the computer to just keep trying to richen the mixture.

All of that needs to be repaired first. The ignition and fuel systems have to be 100% for the system to work properly.

Once those things are in good order, then you might fine it drives fine.

If you still have a dead spot off idle, and stalling or stumbling, there are a couple of things to check. The air sensor plate basically 'jumps' when you hit the gas. It will cause a quick burst of fuel which acts similarly to the way that an accelerator pump works in a carburetor under the same conditions. The system works mechanically for any quick input you give it. The controller can't react all that quickly.

So first check to make sure the sensor plate isn't binding. With the engine off, gently press on the plate and make sure it's properly centered. It should not be touching anywhere around its circumfrence. If it is, loosen the screws and adjust it. That may be all that's required.

If that's ok, then lean the mixture out (counterclockwise) about 1/8 of a turn. Bogging and dead spots are often caused by too rich a mixture.

Once you've finished all this, replace the O2 sensor. If you cannot make adjustments, replace it ahead of time. The EHA should default to a 50% duty cycle with the O2 sensor disconnected. On particularly difficult to tune engines, disconnecting the O2 sensor and adjusting the system for the best drivability can often help. Then when you plug it back in, it will typically run a little better and return better economy.

W124 400E, W126 500SE
971 Posts
Discussion Starter #3
Thank you very much.

I understand how duty cycle works (it's not a lambda readout, but a rich/lean ratio readout).

I had it emission tested yesterday, it was lean so we set it to achieve optimal CO readout (I wanted to fiddle more but the technician called it perfect), duty cycle after emission testing was 50% @ idle and 60% @ 2500 RPM.

So I guess it now runs as good as it can - optimal range for EHA adjustment with proper emissions should mean injection and ignition are OK. I'm still not ruling false air out, but since emissions are good even under throttle I'm not sure that can be it.

I had it tested for vacuum during the oscilloscope test - we did not have the proper datasheet to know how much should be there according to MB, but the technician called it good and I trust him. I did inspect vacuum lines and did not find a leak.

Throttle plate is not binding, I cleaned it with throttle plate cleaner (slightly) and inspected - all looks good.

It does this in both closed loop and open loop modes. In open loop it is better (it doesn't die, just stumbles).

So what other possibilities are there? I'm going to take it to oscilloscope test again after I replace the faulty potenciometer and see what happens when it stumbles, but I'm not sure we can test every component.

Can you tell me what exactly acts as an acceleration pump there? The injection head itself? Maybe my fuel pressure is terribly off? I checked both pumps and replaced the filter already...
1 - 3 of 3 Posts