What can cause a low Lambda

[mespe]

1988 190e Key On engine off Lambda pulled from X11 3 &2 is 10.6 volts. or about 15%. starting the car and left to idle lambda is around 11 volts.

Obviously something is major wrong here, but what?

O2 sensor replaced (output voltage on orig .1 on new .46, so it was bad.)

Spark plug #4 which was cross threaded and about a 80 thousands Gap has been replaced with a properly gapped Plug

So what, besides the O2 sensor, should I check/test to get the lambda within spec for KOEO (key on engine off).

And how can I test it?

[dkveuro]

Looks like a false air leak. Really need to read the O2 at the wire under the passenger side carpet. Should have a 0.000 to 0.990 voltage reading on a DVM .

Key on, engine off should read 0.500 or there abouts.....this is ecu reference voltage.

With engine running and warmed up, you need to see a swing of around 0.025vdc to 0.890vdc with a new sensor.

In reality the reading will be 0.250vdc to 0.890 vdc at idle.
There is an idle valve on the 190e that has a hose from the intake to the IAV and it is underneath the intake runners. These split or get blown off.

Put your hand under the intake from the distributor end and feel for a fat hose to the IAC....often you'll find it displaced or split.

Part # is A102 094 48 83.



.

[ozawa]

i may be able to help but am confused why you're testing lambda with the car not running. lambda only really comes into play once the car is warmed up and in closed loop fuel control. with the car fully warmed up the oxygen sensor should be putting out a nice sine wave fluctuating between about 150mV to 850mV. you can get by with a nice DVOM but the ideal tool is a digital graphing multimeter or oscilloscope...

[mespe]

I'm kinda new to troubleshooting Mercedes gas engines. I was checking pins 2&3 on the round test point on the Left inner fender. I understand that Lambda is an active component of the CIE injection system.

Doing a static test on Lambda, I've come to realize that something isn't kosher with my car, not because of the rough idle, but becausae the static Lambda reading is way off.

So I'm looking for some input as to what effects the Lambda reading form the X11 connector.

When the throttle microswitch is opened, the Lambda reading changes to a closer, more better reading (I think, remember this is all new to me)

[dpetryk]

Lambda is not a component or a thing. Lambda is the term given to a condition where the air and fuel ratios are at a defined point. The oxygen sensor is used to measure this condition based on how much oxygen is present in the exhaust gas. The goal is to keep the engine running at the lambda point. The computer uses the oxygen sensor reading to adjust the fuel mixture to keep it at the optimum ratio - "lambda". This is a closed loop regulating system which is constantly adjusting the air fuel mixture under all conditions except for a few like wide open throttle and in some cases idle.

If the engine is not running the readings you see are meaningless - as far as lambda is concerned. The engine has to be running for the oxygen sensor to work.

There is no such thing as a "static test on lambda". The readings you are observing indicate how much deviation the lambda computer is having to use to keep the engine in lambda. The ideal situation is that the fuel injection system is perfect and it always delivers the exact amount of fuel. Well real world says its impossible. SO you observe the lambda signal and adjust the fuel injection system so that lambda is as close to 50% as possible under all conditions. The EHA receives a signal to modify the fuel pressure to keep the engine in lambda. It only has so much authority to operate with. If the fuel adjustments are too far off, the system will never achieve lambda.

Hope I have not confused you. It would be wise to understand the system basics so you know how it fundamentally works and then you will have an easier time working on it. By the way, all cars use this closed loop regulation scheme to achieve lambda. They just implement it in different ways. Some are all electronic, others are part mechanical and part electronic like the CIS system.

[mespe]

Well Thank you dpetryk, that's what I needed to know - The basics. I've read somewhere to do a "static" lambda test and the values without the car running should be 70% Fed, 85% Cal. Well I was at the other end of the spectrum.

Interesting how you mention lambda is not used at idle for some vehicles, well I've come to the conclusion that this is the case for my 1988 190e 2.3l

No matter what I did, at idle, the value (voltage, where lambda = 1-(V,in/ VCC)) would be way too high. This changed whenever the accelerator pedal was pressed (causing the microswitch on the accel cable to open) and I was getting reasonable lambda readings (4-6 volts or roughly 50% duty cycle), but the car would stall at idle, until warmed up then it would idle at 2000RPM.

After much viewing of threads and reading up on links, and looking at parts pictures on fastlane, I figured out how to adjust the fuel mixture. I cranked it up a bit and now the cold idle is fine,,, time to check chain stretch and tensioner,,,

[dpetryk]

Glad you fixed the problem.

I did not say that lambda is not used for some vehicles. In fact I said the opposite. If the car has an EHA and or an oxygen sensor then you can be sure that it has a regulating system for lambda.

Your car does have lambda.

[mespe]

Sorry about the omission, what I meant to say was lambda is not used at idle,,, or so I believe, the car does indeed have the EHA and O2 sensor,,,

I fixed my previous post to reflect what I meant so future readers wouldn't get to confused.

[Rick76]

The CIS-E control unit has diagnostics built into it.

For your year and model:

Ignition on, engine off - 70% = Federal version control unit, 85% = California version control unit.

Engine at idle - oscillating reading indicates no malfunctions in system, otherwise -

0% = CIS-E control unit receives no voltage or is defective.
10% = idle speed contact
20% = full load contact
30% = coolant temperature sensor
40% = Air flow sensor position indicator
50% = O2 sensor
60% = not used
70% = TD signal
80% = altitude correction capsule

[stevebfl]

Closed loop control is maintained at idle on all MBs that have closed loop control. Most go open loop on full throttle.

The reading at the diagnostic conn is a created reading often called integrator. It is a duty cycle and the point of view of a duty cycle is critical. Fifteen percent duty cycle is also eighty-five percent duty cycle. If you look at what is happening it is easy to understand. You might want to look at the scope patterns in this article: http://www.peachparts.com/Wikka/EngineControls

The figure one that shows 30% duty cycle is refering to the amount of time the voltage is at 0.0v versus the amount of time it is at battery voltage. This is 30% off duty cycle. It is also 70% on duty cycle.

[mespe]

Thank you Rick Steve for your posts. It seems that whenever the decel microswitch is closed (throttle cable resting on the switch) which happens at idle, I get a reading of around 12 volts on X11 pins 3 and 2. As soon as the throttle cable is moved just a hair, the switch opens and and I can see Lambda numbers in the 4-6 volt range,,, These numbers stayed constant at 2.6 when I turned the lambda tower CCW a few turns, but returned to teh fluctating values in the 4-6 volt range after readjustment of the tower.

My problem now is a fast idle at 1400 RPM,,, I don't think the car can pass emissions test with that fast of an idle.

On another note, the speedo/odometer weren't working,,, so I pulled the cluster,,, I didn't see the triangular speedo cable inside, now I gotta pull the tranny side to see if someone removed the speedo cable... I'm thinking the PO really pissed off some mechanic and grievances were taken out on the car.

[Rick76]

Sounds like the idle speed contacts on the throttle body are bad. If this switch is bad you will get a high idle and low Lambda at idle (this is not the microswitch that is on the linkage). There is a wire that runs from the throttle body to a connector with 3 terminals. Unplug it and check the resistance with an ohmmeter. There are actually two sets of contacts - one for idle and one for wide open throttle with a common return. The idle switch should read low resistance (0 to a few ohms max.) and go to infinity when the throttle is moved off of idle.

[mespe]

I think I fixed the problem,,, I don't think that the throttle cable should extend all the way to close the decel micro switch at idle,,, was kinda weird giving the car gas, I'd press on the accelerator pedal and hte pedal would move about 1/2 inch then the car responded,,, after i pulled back teh throttle cable, I got a good idle (600-700) in gear and in park (900).

Seems like Lambda is good, as the car doesn;t sputter or cough when it's given gas. Now for the emissions test, that shouldbe the final say-so whether the car is runnign like it should be.

[Rick76]

The throttle cable should be adjusted so that the linkage microswitch is closed at idle. If you adjust it so that it is not closing, you are just covering up a problem. Check the throttle body idle/wot switch. Good news is that it is easy to test, bad news is that it is not available without buying a new throttle body so the wreckers is the best bet.

[mespe]

Originally Posted by Rick76

The throttle cable should be adjusted so that the linkage microswitch is closed at idle.

I was afraid somebody might tell me that,,, but the car (so far) seems to be 100%.

I have read directions for testing the throttle position, it involves several continuity checks, at idle, part throttle , and WOT.

I originally figured it was supposed to be that way, but I couldn't figure out why the car wasn't in closed loop during idle AND why there was play in the gas pedal. Adjsuting the throttle cable to where the cable just barely comes into contact with the decel microswitch (and letting inertia of the cable, and spring to mometarily close the switch). I guess no such luck.