Here's the deal. Couple of years ago I replaced the old style OVP hoping to correct stalling. Didn't help. Well, a while back the car would not start at all. I still had the old OVP so I plugged it back in. Car started but the old stalling problem that never went away is still there. Check engine light on/off/on/off and eventually engine dies.
Fast forward to today. Gas mileage is poor so I'm gonna take one more try at reading Lambda. I also have my DMM in series with the EHA. Now, I have my analog meter in pin #3 of the X11 connector. Key on ignition off EHA reading 19.95 ma. Start car. Analog meter at pin#3 reading at 12 volts. EHA ma bouncing all over the place!! Adjust the mixture screw and eventually get the Lambda on 7 volts. Old M103 is purring like a kitten but raw gas is dripping out of the tailpipe!! EHA current has never "settled" on any value. Then suddenly the Lambda voltage drops to zero and the car dies. What the heck am I seeing?? Did the OVP have anything to do with that?? Or did the EHA have a hand in it?

Assume that you've checked or replaced your O2 sensor and checked the connection?

__________________
Chuck Taylor
Falls Church VA
'66 200, '66 230SL, '96 SL500. Sold: '81 380SL, '86 300E, '72 250C, '95 C220, 3 '84 280SL's '90 420SEL, '72 280SE, '73 280C, '78 280SE, '70 280SL, '77 450SL, '85 380SL, '87 560SL, '85 380SL, '72 350SL, '96 S500 Coupe

O2 sensor was replaced and is working.

You must resolve the eHA current issue. 20ma is appropriate for key on, ENGINE off readings. The next step is to disconnect the O2 sensor. The ma reading should be zero with the engine running!!. If it is different this issue must be resolved.

If you have 0.0ma then try grounding the computer side of the O2 connection. This should cause the ma to slowly correct to the maximum which is about 12ma. Next place one hand on twelve volts and hold the computer side of the O2 connection with the other hand. (this applies a small >1v signal to the controller) You should slowly go to -12ma, full correction lean.

If the system has the ability to do this, the next step isn't so easy. The next step for testing would be to watch the differential pressure and to verify that it is .4bar and that it moves .1bar greater at 12ma and goes .1 bar less at -12ma.

Control is then working. See what you get with the O2 disconnected.

__________________
Steve Brotherton
Continental Imports
Gainesville FL
Bosch Master, ASE Master, L1
33 years MB technician

Her's what happened,
Ignition on, engine off = 20ma
Disconnect O2 = 0.00 ma eventually (10 seconds)
12 volts thru Cap'n = 9.9ma (engine 'coughed' and went to zero then slowly climbed back to 9.9ma)
Controller O2 wire to ground = 13.8ma

So it looks like electronic control works. You should have gotten negative current flow in the 9.9ma direction.

This range of control will usually make the engine run poorly in one direction or the other. Remember that when you ground the O2 input the system thinks the engine is lean and it richens up the mix by the 13ma (in your case).

The next step is to set the basic (uncorrected) mixture to .5 to 1.0% CO. This is hard to do on MBs because there is no precat test port.

Since the system seems to respond to created O2 signals, one might suspect the real signal, but lets test it. Hook your volt meter to the O2 sensor lead (leave it disconnected from the controller). If you get 1v the system is rich (or the sensor is lieing). Use the 12v thru Cap'n method to lean it out. Does the method make the engine run poorer? Does the voltage drop below 1v?

__________________
Steve Brotherton
Continental Imports
Gainesville FL
Bosch Master, ASE Master, L1
33 years MB technician

Lost me here. How can it make the engine run poorer if it's disconnected? I am testing a disconnected o2 sensor on the sensor side for both tests, right? Would the engine even need to be running for this test?

This is a great way to test the controls, I'm dying to try it on my 190E already.

Thanx for the great posts.

xp

In the last test I mentioned, the concept is to run the car and evaluate the mixture by seeing how it affects the O2 sensor voltage. The sensor is a 0 to 1v voltage generator. It creats voltage that is related to engine mixture within a small range.

Its very likely that the sensor is fouled, but the way to tell is to run the motor (the sensor has to get hot which requires time and engine speed - don't be in a hurry), while monitoring the voltage (sensor on its own not connected). As I stated its likely to be either 1v or zero volts. If its one volt and its working then the mixture is greater than 1% CO (if measured before cat with an exhaust gas analyser - which you really can't do anyway). Here is why I suggested to do the 12v thru Cap'n mixture change. This will lean the car out and unless your are very rich it will bring the mixture through the range 0-0v - 1.0v. if not you can lean it out till it does.

Remember a proper working O2 sensor measures mixture from 0% CO to 1% CO. This is a very small range. The car will start running poorly above 6% CO and below 0.2% CO (maybe higher). The car will be held to mixtures in the range of .3% CO to .7% CO by a functioning system and it can only correct a rich car running less than 4% CO.

If everything is functioning you will be able to lean out the mixture till the O2 sensor just drops from around 1v. Then you will be able to reconnect the wire and the system will correct mixtures less than .5v by adding fuel (within the 10ma correction range we have already tested) till the mixture becomes richer than .5v at which time the current will reverse and the system lean out till it switches the other way. Once the O2 sensor is hot and everything is adjusted then the .3v to .7v O2 sensor swings will be kept in check by about a 4ma total mixture correction.

If this were a training course this would really be a short story. If you don't understand ask me again.

__________________
Steve Brotherton
Continental Imports
Gainesville FL
Bosch Master, ASE Master, L1
33 years MB technician

I re read your post about the O2 sensor test and understood it that time. I re did the EHA test and got the same results. ( and it was -9.9 through the Cap'n. My mistake) I left that hookup and used my handheld DDM for the sensor test. 0.00 volts any way I tested. Even through the Cap'n 12v test. So, sensor is fouled right? Doesn't surprize me, there is more soot under the tailpipe tip than in my fireplace!! But there are two more oddities. Engine is surging wildly and shutting of regularly while warming and would not restart without depressing the accelerator. Also, there is no voltage to the O2 heater connector.

Can the O2 sensor be cleaned? Also, are these tests gonna work if I have a leaking fuel distributor?

Yes. The best way to clean it is to get the engine running properly and then run it slightly lean. Run the poop out of it. Then either test it again or hook it up and verify closed loop control probably after some driving and resetting of mixture.

You eventually must understand closed loop control to understand the system and the various ways to adjust it electronically. Closed loop refers to the fact that the mixture is kept to a precise range by feedback control. The system is made to go rich when it senses lean; it doesn't go far till it has effected it enough that it is now sensed rich and the control heads lean. This back and forth control keeps the average mixture to a very tight band at the exact relative amounts of fuel and air for complete combustion.

If you can get the closed loop to react doing a couple of cycles a second at 2000rpms you have cleaned the sensor.

__________________
Steve Brotherton
Continental Imports
Gainesville FL
Bosch Master, ASE Master, L1
33 years MB technician

I have taken the Lambda adjustment screw to full operable lean when hot and it still drips raw fuel from the tail pipe. I have taken it so far lean that it won't even start cold! Again, my question is, If the O ring in the fuel distributor is leaking, are any of these test valid??

The o-ring at the base of the fuel distributer seals vacuum.

Where do you have a fuel leak? Is it at the EHA? I would always advise fixing fuel leaks first. If the leak is at the EHA I don't think it would keep you from doing these tests. If its at the EHA I would worry that the EHA wasn't doing its job, But I wouldn't worry long because fuel pressures tell the whole story.

Mixture correction at the tower is usually done within a few degrees of rotation. basic calibration after a fuel distributor exchange might require a full turn or two. I have turned that screw many, many turns and have never reached an END. How much are you turning the screw to go from poor running rich through good running to poor running lean?? Does this even happen?? Forget about cold till we get warm right!! First warm, then closed loop and then maybe cold.

__________________
Steve Brotherton
Continental Imports
Gainesville FL
Bosch Master, ASE Master, L1
33 years MB technician

There is a puddle of fuel in the manifold under the throttle body, visible when the airflow plate is pushed down and the throttle is opened. The lower housing of the fuel distributor is damp also.
On the adjustment issue, the engine will run only in a very limited range of adjustment, with the screw approx. 30 degrees either lean or rich. When adjusted hot toward the lean side it runs well, but will not start cold in that setting.