Mercedes-Benz Forum banner

21 - 30 of 30 Posts

·
Registered
1986 560SEL
Joined
·
93 Posts
Discussion Starter #21 (Edited)
H.D.,
I had a chance to investigate the O2 sensor info. I referred to several of your other posts to piece together what I think is the way to test. These threads helped me:
https://www.benzworld.org/threads/99-9-duty-cycle.3004583/page-6 (particularly post #106 and up).
https://www.benzworld.org/threads/m103-rough-idle-and-vacuum-leak.2452785/ (particularly post #17 and all of page 2)

Findings:
1. black to battery ground, red to connected O2 sensor wires
  • Engine off, key on, - 453mV
  • Engine on, cold - 439mV. As the engine warmed up, readings slowly rose to about 898-904mV.
  • At idle, engine slightly running rough, around 700rpm. Idle then went to ~800rpm, then started fluctuating (slowly) 700-800rpm; Still measuring around 900mV.
Now At normal engine temp (80C),
- Idle (~1000rpm) at first steady around 830mV, then fluctuating 070-700, then steady again 830mV; this seemed to reoccur.
[email protected] 2500rpm, measurements were as follows: ~ 190-500mV fluctuating
- Back to idle (stayed at ~1000rpm now), fluctuating 090-700, then steady around 830mV.
Second go round:
  • @ 2500rpm, fluctuating, 150-680mV
  • Idle (1000rpm), 090-700mV
If I did quick throttle blips about 1/2 throttle or more, I got the same "stumble on initial power, and the readings were -002 to 175mV.
When I disconnected the O2 sensor wires, I had no stumble on full power blips, but the idle stayed around 1000rpm.

Measured with O2 sensor wires disconnected, measured at the female end from the O2 sensor:
First time:
  • @ idle - 900 static
  • @2500 - 880 static
  • @ idle - 900 static
Second time:
  • @ idle 896 static
  • @2500 884 static
  • @ idle - 899 static
Now O2 sensor wires still disconnected, , engine off , and measuring resistance, black still using battery ground:
-From the O2 sensor female wire connected to Voltmeter red wire - Resistance = 1.
- From the O2 wire opposite (I assume leading to the ECU) Resistance = 117.4 (I don't know if this is a correct measurement??)
From what I read, these were both supposed to be "0", or infinite??

Next I did warm engine on, idle still staying around 1000rpm, and measured AFM movement:
pushing plate down (almost stalling) - 950mV
pulling plate up with a magnet (really not much upward movement beyond where it was) - 885mV

I have not measured resistance at the sensor body yet, it is a bit difficult for me to get under the car at the moment. If necessary, I will find a way to do that, let me know.
Does this point to a faulty O2 sensor?? Novice that I am :unsure:, It seems that with the O2 sensor disconnected from the ECU it isn't reading voltage like it should?
Thank you.
 

·
Outstanding Contributor
1988 300CE
Joined
·
1,414 Posts
I consider it the highest of all issues in maintaining these engines in good running order - and most ill informed.
So do I. … There‘s no gasoline injection system that you can find more misinformation & misconceptions about on the internet (not only in forums) than the KE-Jetronic.

Most DIY or garage mechanics over the years have gone straight for the Tower adjustment screw (lambda adjustment screw) always assuming it is an idle mixture adjustment screw that will sort out a problem with idle etc. They still have the perception this mechanical injection system is much akin to a carby set-up.
Yep … and it‘s very difficult to convince them not to fiddle around with that adjustment screw, especially since often you can improve an engine‘s subjectively perceived running behavior that way. When they experience that they feel confirmed in their dealing with the system. They‘re not aware that they just mask the cause of the running problem that way … often with negative consequences for the engine. … And when you tell them that by fiddling around with that adjustment screw they wipe out the most valuable diagnostic information that the system provides, they look at you like they would if you‘d tell them that political offices are more and more held by descent, reliable and honest people. … LOL

I had a chance to investigate the O2 sensor info. ...
These are thoroughly performed o2 sensor voltage tests, Rand. (y)

However, the closed loop o2 sensor voltage tests (o2 sensor wires connected) are a lot more informative if accompanied at least by simultaneous EHA current tests, better also by simultaneous LCP tests. Such simultaneous closed loop tests give insight in the complete loop. They allow you to monitor what the o2 sensor tells the CIS-ECU, what the CIS-ECU tells the EHA, what that does to the LCP and whether the o2 sensor detects & reports the expected corresponding air/fuel mixture change to the CIS-ECU again.

Your separate closed loop o2 sensor voltage tests suggest that there‘s either a problem with the o2 sensor or with the CIS-ECU. A problem with the o2 sensor‘s connection to ground via exhaust pipe they don‘t excluded either. They do exclude a short between o2 sensor signal wire & ground though. Of the three remaining suspects mentioned in post #11 the o2 sensor is still the most likely culprit.

Your open loop o2 sensor voltage test (o2 sensor wires disconnected) with the AFM plate pulled up corroborate my suspicion that the Lambda adjustment screw has been messed with too far in clockwise direction. … How much play, if any, do you feel when you gently push the AFM plate down when the engine is not running ? (See what I said about AFM play in This Thread)

Just to be on the safe side, the adjustment tower was touched after the hesitation/stumble started, right ?

I‘ll be quite busy on the weekend and next week so that replies may take a little patience.

Thanks for your appreciation, Gentlemen. :)
 

·
Registered
1986 560SEL
Joined
·
93 Posts
Discussion Starter #23
Update: In a PM, H.D. suggested I go ahead and order/install a new O2 sensor, based so far on my findings, and the fact the O2 sensor to my knowledge, is the original one. He suggested, and I do agree, doing the simultaneous tests on the EHA, O2 and LCP is a fairly difficult process.

I am going to replace the O2 , do the duty cycle measurements again, and then see about correcting the lambda adjustment properly... I won't touch it until I get the go ahead from H.D.!! Yes, it was adjusted after the hesitation/stumble started. I can see that checking the AFM plate seemed to indicate the car is running rich, and I don't know if that lambda screw was messed with before I got the car. Time will tell. I will post here again after the next phase.

Thanks all for reading and reflecting. Special thanks of course to H.D., who continues to further our education.
 

·
Outstanding Contributor
1988 300CE
Joined
·
1,414 Posts
H.D. suggested I go ahead and order/install a new O2 sensor, based so far on my findings, and the fact the O2 sensor to my knowledge, is the original one. He suggested, and I do agree, doing the simultaneous tests on the EHA, O2 and LCP is a fairly difficult process.
Right. … Doing these simultaneous tests is not too difficult either, but a little more difficult than doing the tests separately ... and you need more meters, of course. You put a voltmeter for the o2 sensor voltage, an amperemeter for the EHA current and a fuel pressure gauge and best also a duty cycle meter, best all analog, next to each other so that you can monitor them simultaneously while you do or let someone else do several tests. A video of the meters during the testing fascilitates the procedure significantly. You can rewatch it and pause it at certain moments so that you can monitor the interactions of the involved components in more detail. … Such simultaneous tests can be very revealing.

I am going to replace the O2 , do the duty cycle measurements again, and then see about correcting the lambda adjustment properly...
With the new o2 sensor installed I suggest, for a start, to check the play of the AFM plate at its zero position and to recheck the duty cycle. Proper duty cycle “adjustment“ should be done last at a later time.
 

·
Registered
1986 560SEL
Joined
·
93 Posts
Discussion Starter #25
Is it correct that there is only one O2 sensor in this W126? A mechanic friend said he thought there were 4 (2 upstream, 2 downstream each side)? I have found no evidence of more than one.
 

·
Registered
1986 560SEL
Joined
·
93 Posts
Discussion Starter #27
OK, I got my O2 sensor today, and installed it with no issues. I took the car for a test drive, and although symptoms were somewhat different, there are still issues.
  • On slow take off (normal traffic), there is an almost imperceptible hesitation for a second or two, then it goes away (improved from before).
  • 1/2 throttle from full stop, there is a bit more hesitation, and a very slight stumble (not always), then it goes well (improved from before).
  • 3/4 throttle from full stop, there is a few seconds of stumble, then it takes off.
  • full throttle from full stop, a big stumble, very quick ( and once or twice during testing I had a bit of a 'carb' backfire), then it really takes off.
  • giving it gas while already moving is fine now; no symptoms with a slowly increased pedal, or quick full pedal.
My duty cycle measurements are now as follows :
Key off: ......................................................................Duty cycle:
  • Vp6 (#2 & #6) - 12.76
  • Engine off, key on: 3.58.........................71.94%
  • plate deflection: 10.68............................16.30%
  • full throttle: 9.50......................................25.54%
Engine running:....................................................................Duty Cycle:
  • idle (still around 900rpm): 13.58-13.62.................3.13%
  • 2500rpm: 9.07-10.14..........................................31.55%
  • idle (~900rpm): 9.76-11.15 , (BUT: this returned after a minute or so to 13.57-13.58 and remained stable so I used this as my measurement)..............................3.27%
  • Vp6 now measured again, engine idle: 14.04
I did not yet measure new O2 sensor voltage or resistance, waiting to see if that's necessary.
I did check movement of the CFM plate, and I found it moved freely the first 1-2 mm, then it moved fully but with some resistance. Once fully down I let go and it came up smoothly and quickly. I retested (after priming fuel pumps), and it moved down with some resistance and came up nicely as I allowed a steady upward movement.

synopsis: symptoms are different, still present, though improved. Should my next step be to adjust the Lambda screw? Is it possible that because it is running somewhat rich
(found last time with the measurements made previously), these symptoms are related to that?

Thank you!
 

·
Outstanding Contributor
1988 300CE
Joined
·
1,414 Posts
I did not yet measure new O2 sensor voltage or resistance, waiting to see if that's necessary.
I did check movement of the CFM plate, and I found it moved freely the first 1-2 mm, then it moved fully but with some resistance. Once fully down I let go and it came up smoothly and quickly. I retested (after priming fuel pumps), and it moved down with some resistance and came up nicely as I allowed a steady upward movement.
Does the AFM plate (with the engine off !) sit at its correct “zero position“ ... with its top edge at the same hight as the top edge of the short cylindrical part of the air funnel (measured below the stop bracket), or maximum 0.2 mm higher ? … And does the AFM plate sit concentrically to the air funnel in that position ?

Furthermore, as ‘step 1‘ in post 3, I suggested to check the “movability of the AFM & CP (Control Piston)“. Did you also feel the CP following the AFM plate when you let it slowly move back to its zero position ? … You can check/feel that with your finger if you let it tremble a little during the slow upward movement of the AFM plate. You should feel the CP “dance“ on the roller at the other end of the AFM lever.

I suggest to check ^these things independently of the still existing symptoms. They, along with other things, should be checked before the final duty cycle readjustment via adjustment screw anyway.

Should my next step be to adjust the Lambda screw?
Yes … in the manner suggested in post 3 (‘step 2‘), namely by turning the Lambda adjustment screw about ¼ quarter turn (!) counterclockwise. As mentioned in that post, “mind what I said about turning the adjustment screw under “Adjustment procedures“ in post #2 of my Lambda Control Thread !, referring to the pushing down of the adjustment pin and its latching into the adjustment screw. ... Do this provisional adjustment with air filter installed & warm engine running at idle & voltmeter connected and see how much it lowers the voltage, respectively how much it raises the duty cycle at idle and at 2500 rpm … and what it does to the symptoms during driving.

Is it possible that because it is running somewhat rich (found last time with the measurements made previously), these symptoms are related to that?
Yes … as I said in post #16, with your Lambda adjustment screw having been messed with, an o2 sensor related problem is not the only possible problem. ;)
 

·
Registered
1986 560SEL
Joined
·
93 Posts
Discussion Starter #29
After reading through post #28, I was pretty confused. Then I dug in and reread all the previous material I had already read (but didn't comprehend well). Eureka! I started to grasp the whole picture! Now I get it. By adjusting the pin as per your instructions (in the Lambda Control Thread), the voltage should lower, thus bringing the duty cycle measurements in conditions #4, #5 and #6 closer to the 50% duty cycle target. Small adjustments ccw to lean it out a bit, read the voltages, get as close as I can to 50% then road test.
I can see now, since reading over the initial measurements I made, then the new measurements after having installed a new O2 sensor, where the voltage figures and duty cycle %'s were off, and what I should get to by adjusting the pin.

I will recheck the AFM but it looked like it was seated properly and at the right height. My one question on this procedure is how do I "feel" the CP moving? I get that the CP works in conjunction with the AFM, adjusting the so called "lambda adjustment screw" changes the AFM position and thus the air/fuel mixture. But I am still a little 'foggy' as to where the CP physically resides... is it in the control tower, and is that what I am pushing on when I adjust the pin?? If so, how do I "feel" it move when moving the AFM plate?

Thanks again for all of this. It has taken time for me to start to understand the system, which for me is important. You did say that doing the testing is easy once you do it hands on, and thats what I found, as you said I would! But understanding what and why, thats a different story, and it is all making more sense to me now.
 

·
Outstanding Contributor
1988 300CE
Joined
·
1,414 Posts
By adjusting the pin as per your instructions (in the Lambda Control Thread), the voltage should lower, thus bringing the duty cycle measurements in conditions #4, #5 and #6 closer to the 50% duty cycle target. Small adjustments ccw to lean it out a bit, read the voltages, get as close as I can to 50% then road test.
I can see now, since reading over the initial measurements I made, then the new measurements after having installed a new O2 sensor, where the voltage figures and duty cycle %'s were off, and what I should get to by adjusting the pin.
If you want the duty cycle to be a provider of reliable full-value diagnostic information again, a number of other things have to be done before the final adjustment of the duty cycle close to 50% … all of which would not be necessary, if the adjustment would not have been improperly messed about with.

I had a little time today to make a simple schematic drawing for a better understanding.

Steuerkolben-Einstellung.JPG


For now (after the adjustment has been messed about with) turn the adjustment screw (2) via the spring loaded adjustment pin (3) in the adjustment tower (4) only about ¼ turn ccw and check what that does to the duty cycle in conditions #4, #5 and #6 and to the car‘s running behavior.

I get that the CP works in conjunction with the AFM, adjusting the so called "lambda adjustment screw" changes the AFM position and thus the air/fuel mixture.
It is not the AFM‘s position that is changed that way. It is the CP‘s position in relation to the AFM plate‘s position that is changed that way, as I tried to illustrate in the above drawing. … But the actual challenge (also for the vast majority of professional garage mechanics ;)) lies in the understanding that on the KE-Jetronic the air/fuel mixture is not changed by changing that position, unless the adjustment screw is turned cw beyond the point where the duty cycle reaches ~0% or ccw beyond the point where the duty cycle reaches ~100%, or Lambda control is not active ! … I explained that in detail in post #1 in This Thread.

But I am still a little 'foggy' as to where the CP physically resides... is it in the control tower, and is that what I am pushing on when I adjust the pin?? If so, how do I "feel" it move when moving the AFM plate?
The CP is part (1) in the drawing. It moves up & down in a precise sliding fit in the center of the fuel distributor. A compression spring & fuel pressure pushes it down; intake air pushes it up (via the AFM lever). … (Mind that the illustration of the FD‘s inside around the CP in the above drawing is a very simplified one).

In order to “feel“ whether the CP is following the roller at the other end of the AFM lever, first prime the fuel pumps via ignition key at least once. Then push the plate swiftly all the way down with your finger and let it slowly move back. Let your finger tremble a little up & down during the plate‘s complete upward movement. Then you should feel the CP “dance“ on the AFM roller.

Thanks again for all of this. It has taken time for me to start to understand the system, which for me is important.
You‘re welcome! … And take it one step at a time. Its design is considerably more sophisticated than commonly assumed. Not only the frequent suggestions to fix problems by turning the above adjustment screw are evidence of that. … ;)

H.D.
 
21 - 30 of 30 Posts
Top