Mercedes-Benz Forum banner

21 - 40 of 176 Posts

·
Outstanding Contributor
Joined
·
1,389 Posts
I don‘t want to impose, but
@dolucasi:
... I am skeptical that your cold issues are OVP related. AS the OVP matters the least during cold as the lambda control is actually in open loop mode so some of the sensors are ignored during cold (like O2 sensor, etc). OVP problems are amplified during Hot (not cold) ..
Sorry dolucasi, but that‘s only partly correct. … You‘re right when you say that Lambda control is not active until the engine heats up (with only the o2 sensor being ignored). … But up from cold start all other CIS related sensors deliver input to the CIS-ECU … and in order to process that input the ECU depends on supply voltage from the OVP … in order to control, for instance, warm-up enrichment and idle speed (M103) !

... I would first look into things that effect operation during cold (like fuel mixture screw, ...
Be careful mentioning that screw … rather mention a duty cycle test. That adjustment screw is, unfortunately, touched way too rashly way too often.

... It sounds like your fuel/air mixture is not rich enough during the warm up period and once the engine warms up ...
… which a duty cycle test would clarify.
@codycool55:
Nothing you reported so far indicates a problem with your OVP.
@Strassenkreuzer:
I suggest to check voltage with “ignition on“ (engine not running) at:
- terminal 1 of the disconnected CIS-ECU connector
- terminal 2 of the disconnected ICVplug

Attention: always disconnect / reconnect the ECU connector with ignition switched off !

H.D.
 

·
Premium Member
Joined
·
453 Posts
Yes OVP stands for Over Voltage Protection relay. I now recommend replacing OVP's now older than 10 years old (newly converted believer) no matter what to avoid future headaches.

I am skeptical that your cold issues are OVP related. AS the OVP matters the least during cold as the lambda control is actually in open loop mode so some of the sensors are ignored during cold (like O2 sensor, etc). OVP problems are amplified during Hot (not cold).

I would first look into things that effect operation during cold (like fuel mixture screw, EHA, coolant temp sensor, etc)
It sounds like your fuel/air mixture is not rich enough during the warm up period and once the engine warms up the ECU and actuators compensate for the problem.
Thanks for the insights. I'll probably have my mechanic look at it; engine problems aren't my forte, and I don't know that I have the tools or the know-how to play the "process of elimination" game. It's only begun doing this recently, so I wonder if it might even be a fouled spark plug or injector(?). Plus, I don't know engines well enough to know if the rough idle is one cylinder misfiring—or if it's multiple.

But now that you've mentioned "lambda control" and EHA, I'm gonna let my Googles do the talking and find out what those are. :)
 

·
W201 Moderator
Joined
·
2,224 Posts
Street cruiser, if you want to know more about Fuel injection details than your mechanic does (even if he is doing the work), I would highly recommend reading H.D. 's write-ups on the subject on BenzWorld. They are simple and concise enough and educational. He has them in W124 as well as W201 and W126 (I believe).

Jetronic FI is better explained in these write-ups than the Bosch manual (in my humble opinion) as I have read both (cause I had to, believe me)

Although I comment on FI related issues in various post, I always refer people to read the Bible (H.D. write-ups) for the whole story.

My 2 cents worth...
 

·
Registered
Joined
·
8,916 Posts
The OVP relay is a strange critter and I'm not sure exactly how it's wired into the big picture but when one goes bad it does some really weird stuff to the car :eek warning lights flash or come on and off for no reason, car runs differently than "normal". Under a suggestion years ago I replaced the one in the SE (picked a new looking one from a wrecked car) and suddenly, no more flashy lights, good and proper idle, and overall ran better :dunno: :bowdown:
 

·
Outstanding Contributor
Joined
·
1,389 Posts
The OVP relay is a strange critter and I'm not sure exactly how it's wired into the big picture but when one goes bad it does some really weird stuff to the car :eek warning lights flash or come on and off for no reason, car runs differently than "normal" ...
That lies in the nature of its function ... it provides (limited) supply voltage (battery voltage) for ECU‘s … on the M103 also for the ICV, which is actually also an ECU supply voltage (via ICV) that‘s converted into an adjustable square wave voltage by the CIS-ECU for ICV control.

Fortunately, the OVP is not difficult to check. … I can monitor it (and several other things) during driving with the device shown in the picture in post 22 of http://www.benzworld.org/forums/w126-s-se-sec-sel-sd/2720049-ke-jetronic-lambda-control-duty-cycle.html ... something I recommend enthusiasts of KE-Jetronic equipped cars to think about, BTW. Facilitates diagnosis immensely, especially in case of intermittent problems. … :wink_2:

H.D.
 

·
W201 Moderator
Joined
·
2,224 Posts
CC55, back to your concern since this is your post originally. Looking at your vacuum and comparing to mine, it is not that bad. Your leak may even be on the vacuum gauge side of things (like the vacuum actuators with your HVAC) Mine was like yours a year ago, did not effect the idle or the engine performance in any way. I have fixed the leaks now but I had to get into the dash to do it and I replaced two actuators. The vacuum is adversely effected by your low RPM. At 700RPM you are closer to all the way pegged. You still have one minor leak but again comparing to my engine it should not effect your performance. If you had a MightyVac you could find the instrumentation related leaks very easily. The cost online is about $25 I think.

I'm curious, what month was your M103 born in 1989? Since it has >200K miles I assume after March. Am I correct?
 

·
Outstanding Contributor
Joined
·
1,389 Posts
... You still have one minor leak but again comparing to my engine it should not effect your performance ...
A false air leak causes leaner mixture, which is compensated (enriched) by higher EHA current, which is indicated by a higher duty cycle. … The OP, on the contrary, read out a duty cycle of 0% (post 14), which indicates extremely rich mixture (see post 15 & 17) !

Turning the Lambda adjustment screw ccw might lean out the mixture far enough to get Lambda control back to working again, which would be indicated by a fluctuating duty cycle. The duty cycle might even be adjustable back to its proper range if the adjustment screw is turned far enough (should not be turned out of its thread though !) … but that might just compensate for a problem that would continue to exist, which is not recommendable. … The tests mentioned in post 17 are the much better option. … :wink_2:

H.D.
 

·
Registered
Joined
·
68 Posts
Discussion Starter #29
If you checked the Lambda control adjustment (at X11 across pins 2 & 3) with a voltmeter and read out “a constant roughly 14.1 to 14.2 volts“ (instead of measuring a duty cycle of 0% with a duty cycle meter), then:
- either the engine is running very rich
- or there‘s a problem with the ECU
- or there‘s a short circuit on the wire to pin 3 of X11


Very good question … much better than rashly changing the Lambda control adjustment when the meter shows a too low/high duty cycle. … And I hope that the mechanic made sure that everything else was in order before he set it.

I suggest to check a few things that might cause the 14 volts (0% duty cycle) reading … starting with measuring voltage, with “igniton on“ (engine not running), between X11 pin 3 & ground … and between X11 pin 6 & ground. The quotient of the pin 3 voltage and the pin 6 voltage should be about 0.3. … If that's the case, the third of the above-mentioned possible causes can be excluded (and the second one is less probable).

The next thing I‘d suggest, is to measure the o2 sensor voltage at the (connected) o2 sensor connector under the passenger side‘s floor panel (green cable). ... 0.8 volts or higher indicate “rich mixture“.
If that‘s the case, I suggest to check the CSV (cold start valve) for tightness.
If the CSV is tight, I suggest to check the fuel pressure.
With your engine running at idle, SP (system pressure) should be about 5.3 - 5.5 bar (77 - 80 psi) and LCP (lower chamber pressure) should be about 0.4 bar (5.8 psi) below SP ! … SP higher than 5.5 bar (80 psi), or LCP more than 0.4 bar (5.8 psi) below SP are possible causes for too rich air/fuel mixture.

Let‘s see the results of these test prior to further suggestions.

H.D.
I decided to go ahead and replace the OVP... to very little if any improvement

In the ignition on position, pin 3 to pin 2 read a voltage of 1.84. Pin 6 to pin 2 measured a voltage of 12.14. The quotient is .1515 :(

The green wire on the 02 sensor fluctuated between roughly .46 and .49 volt. It should be at about .45 correct? So the engine is running slightly rich?

To the best of my knowledge the cold start valve was as it should be

I attempted to measure fuel pressures, but my friend with a fuel pressure gauge didn't have the correct fittings for my car so I'll have to try again at a later time.

In addition, while doing these tests with the key the ignition on position I noticed an intermittent quiet clicking /fluttering sound coming from the vicinity of the fuel distributor. I initially thought it was the EHA, but after unplugging it the sound persisted. Any idea on this?
 

·
Outstanding Contributor
Joined
·
1,389 Posts
Okay codycool55 ... I have a little time today for a little more thorough advice / explanations. … :)

I decided to go ahead and replace the OVP... to very little if any improvement ...
I have no doubt that you‘d better say “to no improvement“ instead of “to very little if any improvement“ … :wink_2:

As I said in post 21: “Nothing you reported so far indicates a problem with your OVP“.

Call me old-fashioned … but I don‘t like replacing parts that are still good … and trust me, the OVP is one of the most often unnecessarily replaced parts on cars equipped with KE-Jetronic. … :wink_2:

... In the ignition on position, pin 3 to pin 2 read a voltage of 1.84. Pin 6 to pin 2 measured a voltage of 12.14. The quotient is .1515 :( ...
When I saw your location (Eastern Washington state) I didn‘t bother about asking you if your car was originally registered in California. Maybe I should have, because in that case the quotient of your two voltage measurements would be spot on … which is a good sign for the ECU (although that alone would not prove there‘s nothing wrong with it).

... The green wire on the 02 sensor fluctuated between roughly .46 and .49 volt. It should be at about .45 correct? So the engine is running slightly rich? ...
o2 sensor voltage of 450 mV stands for the optimal (better: optimal compromise regarding) air/fuel mixture (Lambda = 1, air/fuel mass ratio ~ 14.7:1 in case of non-ethanol fuel). A healthy o2 sensor (at its operating temperature) should fluctuate with at least +/- 350 mV around 450 mV, hence between 100 mV and 800 mV, at a frequency of about 0.5 Hz with the engine running at idle.
I explained that in more detail in the link mentioned in post 15.

With the non-fluctuating X11 pin 3 & 2 voltage of 14+ volts you reported in post 14 ... you should see a non-fluctuating o2 sensor voltage of at least 800 mV !
(on the green wire with the o2 sensor connector connected, the hot engine idling and ECU and EHA connected)

The fact that you see less than 500 mV, indicates a problem with the o2 sensor, the ECU, or the EHA.

There‘s a number of very informative tests, which can not be found in any service manual (at least not as far as I've seen) and which to describe all would by far exceed the maximum number of characters in one post. That‘s why I only suggest one simple test for now (might even be enough in this case already):
Check the o2 sensor voltage on the female part of the disconnected o2 sensor connector while the hot engine is running at idle, with the ECU connected and:
- EHA connected
- EHA connected and male part of the o2 sensor connected to ground (with a wire)
- EHA disconnected
- EHA disconnected and engine speed risen to 2000 - 3000 RPM

In post 7 you said that lately you've been needing to replace spark plugs excessively frequently as they are becoming fouled to the point where the car will not start. … As I said above, your o2 sensor should provide at least .8 volt with the 0% duty cycle you reported in post 14. Maybe it's not providing voltage higher than .49 volt, because it can't, because it is fouled too ! … Can happen by (already mentioned) extremely rich mixture, or by burning too much oil !

So, let‘s first check your o2 sensor with the above tests, before we come back to your “extremely rich mixture“ indicating voltage measurements at X11 pin 3 & 2 ... and before we touch the adjustment screw!

... In addition, while doing these tests with the key the ignition on position I noticed an intermittent quiet clicking /fluttering sound coming from the vicinity of the fuel distributor. I initially thought it was the EHA, but after unplugging it the sound persisted. Any idea on this?
The first thing I suggest to check in this case is the carpet in the footwell (if there is any). Your gas pedal might not be completely released and the throttle linkage moved to the point where the throttle linkage‘s micro switch switches. ... :wink_2:
With the gas pedal completely released, it should be switched to continuity. When the gas pedal is pressed it should switch open, before the throttle plate starts to move. … So, it‘s quite a delicate adjustment, which I strongly suggest to check, because if it‘s not correct, it could be the cause for some of the symptoms you described.

H.D.
 

·
Registered
Joined
·
68 Posts
Discussion Starter #31
Yes, I should have believed you on the OVP H.D. But I figured it was work a shot..

I assume a VIN beginning in CA indicates the car was originally from California?

When performing the o2 sensor test the car may not have been at operating temperature. I'll have to double check the voltage the next time I drive. But I will go ahead with these tests the next time I can! Also, should I still be concerned with testing the fuel pressure?
 

·
Registered
Joined
·
1,618 Posts
Cody have you changed the engine temperature sensor yet . A lot of information is sent to the electronics parts on your engine for it to operate right . If the sensor is stuck, then it could tell the computer its cold ,and in doing so the computer will then richen the mixture up . But can work the other way and make it weak when your engine needs rich mixture ..
 

·
Outstanding Contributor
Joined
·
1,389 Posts
Yes, I should have believed you on the OVP H.D. ...
Be careful, believing people … I might be a loudmouth … (many of‘em out there today :wink_2:) … LOL

I ‘believe‘ that it‘s much better to properly test things to find out what‘s causing a problem, instead of changing things to see if it goes away. … :wink_2:

... When performing the o2 sensor test the car may not have been at operating temperature ...
o2 sensor voltage tests only make sense if done with the engine well at operating temperature !

... should I still be concerned with testing the fuel pressure?
Not yet … let‘s get reliable o2 sensor voltage data first.

Btw, nothing you reported so far indicates a problem with your CTS (coolant temperature sensor) either. … :wink_2:

H.D.
 

·
Registered
Joined
·
68 Posts
Discussion Starter #34
Check the o2 sensor voltage on the female part of the disconnected o2 sensor connector while the hot engine is running at idle, with the ECU connected and:
- EHA connected
- EHA connected and male part of the o2 sensor connected to ground (with a wire)
- EHA disconnected
- EHA disconnected and engine speed risen to 2000 - 3000 RPM
Upon attempting to complete these tests I realized I needed some clarification on how exactly to complete them.
Does this image show the correct configuration? This referring to the green wire in the passenger footwell.
Assuming I'm correct..
- EHA connected: .18-.21 Volt
- EHA connected and male part of the o2 sensor connected to ground (with a wire): Not yet tested due to confusion
- EHA disconnected: .52-.53 V
- EHA disconnected and engine speed risen to 2000 - 3000 RPM .27 V
 

Attachments

·
Outstanding Contributor
Joined
·
1,389 Posts
... Does this image show the correct configuration? ...
No … the attached sketch shows how the 2nd of these four o2 sensor voltage tests is done. … For the 1st, 3rd & 4th of these tests you leave out the wire between the female part of the connector and ground. For these four tests there must be no connection between both parts of the o2 sensor connector (open loop test)

Since the data you reported so far sounds a little implausible, I‘d also like to show you how the closed loop o2 sensor test that I suggested in post 17 and that you reported results of in post 29 is supposed to be done. … Both parts of the o2 sensor connector are connected in a way that allows you to measure the voltage between the o2 sensor wire and ground. In the attached photo you see an example of how that can be done. It shows a simple adapter that I made for this test. The black lead is attached to good ground somewhere.

For all o2 sensor voltage tests the engine must be well at operating temperature !

H.D.
 

Attachments

·
Registered
Joined
·
68 Posts
Discussion Starter #36
Open Loop: fluctuated between 750-800 mV
Closed Loop: Fluctuated between 0 -800 mV
Open loop without EHA: fluctuated between 400- 700 mV
Open loop, no EHA, 2500RPM: steady 815, minor fluctuation.

I used the bolt heads used to attach the seat as ground. While taking these measurements the electronics were very finicky, and I got some weird values. Let me know if any of these numbers still seem improbable as I may have made an incorrect measurement
 

·
Outstanding Contributor
Joined
·
1,389 Posts
... I used the bolt heads used to attach the seat as ground ...
That's okay, but you can also use the battery minus pole.

... While taking these measurements the electronics were very finicky, and I got some weird values ...
… which might also have to do with your meter ... plus, assuming that you‘re using a digital meter, an analog meter is much more comfortable for reading such fluctuating values.

... Let me know if any of these numbers still seem improbable as I may have made an incorrect measurement
These numbers differ significantly form the implausible ones you reported in post 34. … Could it be that in post 34 you measured between both parts of the connector, and not between the o2 signal wire and ground ? … :wink_2:

The numbers you report now, look more plausible and more like what I expected. But you forgot one measurement. Anyway, since these tests are very easily done and you‘re not completely sure if you measured correctly, I suggest to repeat them, with the hot engine running at idle and the ECU connected, in the following order:

Closed loop test (EHA and o2 sensor connector connected):
1) Both parts of the o2 sensor connector ‘somehow‘ connected (diagram #1)

Open loop tests (o2 sensor connector not connected):
2) EHA connected (diagram #2 without ground wire)
3) EHA connected and male part of the o2 sensor connected to ground (diagram #2)
4) EHA disconnected (diagram #2 without ground wire)
5) EHA disconnected and engine speed risen to ~ 2500 RPM (diagram #2 without ground wire)

H.D.
 

Attachments

·
Registered
Joined
·
68 Posts
Discussion Starter #38
1) 150-850mV, seems like it wanted to settle around 650-750mV
2) Fluctuating around 770mV
3 fluctuating around 810mV
4) -10mV
5) 810mV
 

·
Outstanding Contributor
Joined
·
1,389 Posts
Okay, based on these significantly more plausible o2 sensor voltages, there seems to be a problem
- with the FD‘s (fuel distributor) basic setting via Lambda adjustment screw
- or with the position of the CP‘s (control plunger) lock screw
- or with the AFM‘s (air flow meter) or CP‘s movability
- or with the cleanness of (at least the lower part of) the FD‘s metering slits
- or with the injectors


I suggest not to touch the Lambda adjustment screw, and to do the following test now:

  • First check the AFM plate‘s “zero position“ according to the attached PDF
  • Then let the engine reach operating temperature and switch it off
  • Push the AFM plate slowly & slightly down with your finger and attentively feel the mechanical resistance. On the first 1-2 mm of the plate‘s travel (starting from its correct “zero position“) you should feel almost resistanceless play. The only resistance you should feel along these first 1-2 mm is the one that‘s caused by the AFM‘s counterweight.
  • Then, latest after about 2 mm !, you should feel the other end of the AFM lever reach the CP. Up from that moment the AFM lever pushes the CP (in the center of the FD) upwards, causing you to feel some slight homogenous additional resistance. Additionally to that suddenly starting sensible resistance you should not have to apply any additional break-away force at that point.
  • Up from that moment there should also not be any noticeable resistance jumps along the rest of the AFM plate‘s travel.
  • After having pushed the AFM plate completely down, quickly and completely let go of it ... it must quickly and smoothly jump back into its “zero position“.
  • Then turn the key once or twice to ignition on for as long as you hear the fuel pumps prime.
  • Then push the AFM plate swiftly completely down with your finger again, but this time don‘t let it freely jump completely back. This time let it swiftly jump back in 5-10 mm steps. Hold it at these steps for a moment and feel if the CP is following the AFM lever (with a little delay) … you should feel it reach the AFM lever less than a second later at each step. This way feel if the CP follows the AFM on its incremental way back to its “zero position“ !
If you want to repeat one of these two test steps, first turn the key once or twice to ignition on for as long as you hear the fuel pumps prime.

Furthermore, I have a question: Has the FD ever been removed ?
If so, has the CP’s lock screw (in the center of the FD’s bottom side) ever been touched ?

H.D.
 

Attachments

·
Registered
Joined
·
68 Posts
Discussion Starter #40
I tried all these tests and they've all been passed, except for centering the plate. I attempted to measure the gap using a basic Napa feeler gauge at .483 mm. The guage is approximately 1 CM across, so I'm having trouble fitting it into the gap due to the curvature. Is there a trick to this test, or are circular feeler gauges (at .5 mm) an option? This was performed first thing in the morning before the car had been driven.

I'm unsure whether the FD has ever been dissassembled.

Additionally, since completing the o2 sensor tests my check engine light has come on. I pulled the code and got 5 blinks. If the chart I'm reading is correct this indicates a fault with the o2 sensor.
 
21 - 40 of 176 Posts
Top