560SL "Japan" poor running engine, PROBLEM SOLVED!

Hi Gunter,
the whole thing sounds weird.. let's see what happens with the new oxygen sensor. 
One hint in general: it's possible your SL has 2 problems / defects at the same time causing similar symptoms. Been there. 
Before you take out the sensor let the exhaust warm up, it will come out easier. When it's out, just start the motor to see if it revs up better. Just to verify the converter is not clocked. Don't let it run to long, the opening points toward the undercarriage and melts the undercoating. 
Put some copper grease or similar on the thread of the new sensor's thread, if it doesn't come pregreased. 
good night, cheers, Werner

Dr-DJet wrote:

Hi,

this gets too confusing for me. I never saw you write of 0,48 Bar differential pressure before. It was close to 0 and that is why you decided to swap pressure regulator.

volker, changing out the pressure regulator did have an effect on pressure differential, but dit not solve the problem. Some-one higher up in the chain of KE-jet command is telling the distributor to lean out the mixture, and I want to find out which scoundrel that is 
First suspect now is the O2 sensor.
 

My107 wrote:

Hi Gunter,
the whole thing sounds weird.. let's see what happens with the new oxygen sensor. 
One hint in general: it's possible your SL has 2 problems / defects at the same time causing similar symptoms. Been there. 
Before you take out the sensor let the exhaust warm up, it will come out easier. When it's out, just start the motor to see if it revs up better. Just to verify the converter is not clocked. Don't let it run to long, the opening points toward the undercarriage and melts the undercoating. 
Put some copper grease or similar on the thread of the new sensor's thread, if it doesn't come pregreased. 
good night, cheers, Werner

werner, I was a bit confused in the beginning of this big conversation, because you mentioned the O2 sensor quite early on, but with the suggestion that maybe the 560 was running rich and lean. 
I was always convinced that there was never any rich running, so I focused on other things, instead of the O2 sensor.

I found this evening, that resistance on the sensor was 165 ohms. With an extra ground, resistance dropped to 8 oms, but I could not percieve any better running of the engine.
I know that the Ecu gets an important signal from the O2, sensor, between 0,1 and 0,9 volts. At optimum mixture, it should send 0,45 volts to the ecu.
If the O2s sends a signal of  between 0,1 volt and 0,45 volt to the Ecu, it tells the Ecu that the mixture is below stochiometric, and then the ecu tells the Eha to enrichen the mixture.
conversely, if the O2s sends a signal of between 0,46 volt and 0,9 volt to the Ecu, it tells it that the mixture is above stochiometric, and then the Ecu tells the Eha to lean out the mixture.
at 0,45 volts 14,7 units of air should enter te engine for every unit of gas. Stochiometric balance.

more tomorrow

I found this evening, that resistance on the sensor was 165 ohms. With an extra ground, resistance dropped to 8 oms, but I could not percieve any better running of the engine.I know that the Ecu gets an important signal from the O2, sensor, between 0,1 and 0,9 volts. At optimum mixture, it should send 0,45 volts to the ecu.
If the O2s sends a signal of  between 0,1 volt and 0,45 volt to the Ecu, it tells the Ecu that the mixture is below stochiometric, and then the ecu tells the Eha to enrichen the mixture.
conversely, if the O2s sends a signal of between 0,46 volt and 0,9 volt to the Ecu, it tells it that the mixture is above stochiometric, and then the Ecu tells the Eha to lean out the mixture.
at 0,45 volts 14,7 units of air should enter te engine for every unit of gas. Stochiometric balance.

more tomorrow

So, to continue;

After learning that the resistance of the O2-sensor was too high at 165ohms, but that adding a ground (and then, having a resistance of 8 ohms), did not change anything, I thought that I was again at a dead-end.

I then decided to remove the passenger mats and footwell, and tried to measure the voltage that the O2-sensor sends to the ECU.
So I fabricated a ground-wire from the chassis, to my multi-meter (setting at 2V DC), with the positive side to the signal-cable of the O2-sensor (leaving the "heating-wires" in their socket. )
The single round plug of the O2sensor directly goes into the Ecu, to underline the importance of it!

I removed the signal wire from the plug, (so up to that point, there was no "outside interference" with the communication between O2-sensor voltage/signal cable to Ecu), and then, as quickly as possible measured

I measured 0,91 volts! Car ran almost the same, with or without O2-voltage-signal attached.
When waiting a few seconds, voltage dropped to 0,55 at the absolute minimum, but never under 0.45volts.

So in effect, the O2-sensor is continuously telling the Ecu that the mixture is "rich" (low oxigen in exhaust-fumes), by sending 0,91volts to the Ecu, and the Ecu responds by telling the EHA to lean out the mixture.
But the mixture IS already lean, so the voltage should be BELOW 0.45 volts, so the Ecu can tell the EHA to richen the mixture.

On tuesday the new O2 sensor, ordered through the dealer, arrives. I will install, measure again, and will communicate my findings.

Have a nice weekend!

MfG
Gunter

 

After lying awake this night I was thinking about the TPS. See link

www.youtube.com/watch?v=mFsi_qYGJw8

I will make a measurment....

Hi Gunter,
don't do too many things by guessing. Have you ever heard about checking the duty cycle? There should be a plug somewhere in your engine compartment on the l/h side, and it's possible to check the duty cycle which would point towards the problem. I think 1987 and later do have a real OBD-plug. There are several articles about this in different forums, just dive into this, if you want. Sorry for not mentioning this earlier..
cheers, Werner

Werner,
that's the problem. I AM guessing what the problem could be.
I incorrectly assumed that the O2-sensor influences "suddens throttle changes."
The O2 sensor DOES influence the mixture, but only if it is not "overruled" by another, more importantant signal. Of course, if you press the throttle, every other signal is less important than the signal that allows lots of fuel to pass through the injectors.

I'm laying a puzzle of 25.000 pieces here, and I only can guess what the final picture looks like.

I measured the "idle/full throttle" switch on the throttle-valve.

There is continuity between idle and ground when the throttle is at idle. Once you touch the throttle, continuity stops, (about 12 volt). If you go to full throttle, there is continuity between the full-throttle-pin and ground at about 9 volts.

So that is also ok.

Volker, I have not re-measured upper and lower pressure of the fuel-distributor as of yet.
(last measurments were 6.34-5.86)
Will try that asap.

Hi Gunter,

I find it difficult to follow your results as it seems to be a quickly moving target.

If your pressures without connector on EHA fit now, then leave it off and your car should run as KA-Jetronic car. If it still has problems then, forget all electronic influence. Then you have hydraulic issues. WOuld not be surprising after rust in tank. You can try to CAREFULLY adjust idle-run CO . This screw is very sensitive and can make your engine cease to work at all. But due to catalytic converter that will not tell you too much.

So the question is;

Who tells the ECU; everybody, shut up, throttle is depressed, MORE FUEL NOW!

Because, as stated, (not under load "road-testing", only stationary) when the engine is cold, the Ecu DOES get the signal MORE FUEL NOW, but after that, something or some signal al of a sudden tells the Ecu not the add fuel anymore when pressing the throttle. And when cold, the exhaust DOES smell a bit of fuel, but once warm, there is a lean mixture. (but I'm not sure of anything anymore....)

I hope I'm makin sense, because I'm running in a maze here....

Dr-DJet wrote:

Hi Gunter,

I find it difficult to follow your results as it seems to be a quickly moving target.

If your pressures without connector on EHA fit now, then leave it off and your car should run as KA-Jetronic car. If it still has problems then, forget all electronic influence. Then you have hydraulic issues. WOuld not be surprising after rust in tank. You can try to CAREFULLY adjust idle-run CO . This screw is very sensitive and can make your engine cease to work at all. But due to catalytic converter that will not tell you too much.

 

Hi Volker
I'm sorry for jumping all over the place. This is purely a hobby of mine, and I gues this is the case for most people on this forum, and I only have my limited sense of logic to guide me in this proces.

I already "touched" the CO screw, and it to changes idle a bit. But between reasoble running when I first had the SL, and the moment all problems began, that screw was not touched. Idle is getting a bit worse, what, with all the stuf I already did, unplugging, plugging, new parts, etc, so I now detached de -lead on the battery, to give the ECU a bit of rest, to reset everyting.

I just pulled a spark-plug. They are Bosch WR8DC, so they are wrong (but were present before problems began). I'm going to order NON-resistor NGK's.

Regarding the rust, if there would be a blockage somewhere, then why, at warm-up, there seems to be enough flow and pressure, and all of a sudden, (I don’t know exactly at what watertemperature), the engine seems starved of fuel??? And that happens quite soon, way before operatingtemp is reached.

so, what happens than, and why has it the effect it has?

The sparkplug was certainly NOT fouled up by fuel. It was dry and brownish. See picture.

Dadurch was du machst, führt zu keinem Ergebnis. Du schreibst an allen Stellen gleichzeitig. Du musst systematisch vorgehen. ...so wird das nix...sorry

Hallo Andreas. I'm not quite sure what you mean.

Exactly what is it that I am doing wrong? Do you really mean that I am not finding the solution to my problem, because I am writing in different forum-topics?

Or do you mean to say something else, Andreas?

Measured the potentiometer on the side of the distributor.
Changing the position of the meter did have an effect on the engine, (worse) but it now sits at exactly 0,7 volt at idle.

i wanted to say, that you will not come to a result by the way you work. You are skrewing on too much places at once. You have to work by System. Did your read in the WIS? You find it here in the SL pedia
This is a exzellent programmed repair Instruction.

ZitroniX wrote:

i wanted to say, that you will not come to a result by the way you work. You are skrewing on too much places at once. You have to work by System. Did your read in the WIS? You find it here in the SL pedia
This is a exzellent programmed repair Instruction.

Andreas, thank you for your clarification and understanding.
I confess, It has become a game of elimination instead of “programmed repair.”

That is why I have to pay up to 85eur per hour to bring the SL to a specialised garage, who have the tools, the place, the people and the education to follow the garage-manual/programmed repair.

I’m here to enjoy my hobby, and to communicate with other hobby-ists, that is all I can say.

As I already mentioned, when the engine is cold, the problem with the lean mixture is not (or much less) present.
As a test, this morning, I started the car, and immediately floored it on my private drive-way. And it revved up, and spun the rear-tires (on hard gravel, not concrete) do, even under load, AFR seems quite ok.

Is it possible that the fuel-distributor is fouled up inside, but in a way that allows enough fuel-flow when the engine is cold?

In any case, I think I checked and measured al componants that I can with my limited possibilities, so I think I am at the end of the road with my effort to fix the problem myself.

Hi Gunter,
just to explain this for you: everybody here tries to help anybody. No matter what's wrong with the guys' SL. 
But you need to work systematically. Right now you really work on driving any professional workshop nuts. 
Some of us here are professional car guys, or even learned a lot by doing it (the right way).
Don't get me wrong, first thing with problems like this is to get a clear diagnosis. Me, and also others here tried to lead you towards this. Please get a emission test sheet, at different rpms( like I requested before) and / or check the duty cycle first. Most of us here know by experience this will be the first and most important step. You will receive any possible support from here, you probably noticed this already. 
Rolling the dices will make nobody happy..
Cheers, Werner

Thanks, Werner for your explanation.


Bye!

Hello everyone,

I have been able to narrow "the problem" down to vacuum advance for the ignition-distribution.

I will explain later.

 

A second fault was found. The temperature sensor is broken (green 2 pin-sensor with 2 round seperate plugs above intake manifold). It shows 8200ohm resistance, at 7°C watertemperature (ambient temp after 24h) which is much too high. At 90°C I still measure almost double the required resistance. (410ohm measured, 247ohm required). It means that the EHA receives a signal to enrichen the mixture at a time when it is not required.
Next tuesday I will replace, and report.

 

I have received the new temp-sensor, and installed.
Idle is very good, and stable at 750rpm, in park, but problems as described before, are not gone.

I measured the resistance between pins and ground, and it was 250ohm at 90°C.

Elsewhere, I read that you can unplug the ECU, to effectively make a k-jetronic (purely mechanical injection) out of a ke-jetronic.

So I did just that.

Car started quite difficult, but then held idle. The problem with revving was worse, but did show the same symptoms; car did rev a bit when slowly pushing the throttle, but bogged when quickly pushing the throttle.

If the problem would be elektronic, then I would think that detaching the ECU would ameliorate the running, but it did not.

So I guess the ECU and EHA are trying their best to provide a correct mixture, but, once the throttle is openend quickly and suddenly, the systems seems unable to "follow that input."

I have ordered a obd1 scanner (I have the rectangular 8 pin diagnostic port), so I will be able to measure faults.
I will also buy a multimeter with duty-cycle measurment, to see what I can measure.

And then we can start all over again....

Well, I don’t see what eventually was the problem here. In my opinion, one of the very first tasks to undertake is a smoke test. You can’t very well achieve reliability of results with one or more vacuum leaks.

Hello,
I thought this thread was sleeping
I performed a vacuum-smoke test a while ago, that uncovered a few small leaks (the most obvious was a leak at the throttle plate axle) but nothing serious.
The SL is at a specialist workshop now.
They have uncovered a substantial fuel-leak at the plunger of the air-flow plate.
The complete assembly was removed, and sent to a renovation specialist, but after new installment, the plunger still leaks.
The mechanic has now installed a replacement air-flow assembly with a non-leaking plunger, to see if that is the source of the problem.
But it has clearly been determined that the problem is mechanic, and not electronic.
furthermore, the distribution-chain and guiderails have been replaced as well. They showed a slack of 7,5 degrees, with a maximum of 9 degrees allowed by the factory manual.
 

It has taken blood, sweat and tears, not to mention a few pedantic people on this forum, but the problem with my 560SL "Japan" has been solved!
The head-mechanic at the specialist where my SL is now, decided to change out the fuel-distributor and "luftmengenmesser" (with test-examples at the garage) because the plunger in the distributor still leaked after revision. 
But not only that, he also blocked of the fuel-feed to the cold-start-injector.

Apart from fine-tuning, the bogging or refusal of the engine to rev up after warm-up has been cured!!.

It is still difficult to understand what exactly went on in the inlet-trajectory, but the leak at the plunger combined with an apparant leak at the cold-start-injector caused the mixture to become so rich that there was no combustion possible anymore, especially when entering air sucked in fuel and fuel-vapors which accumulated in the inlet. When only slightly pressing the throttle, less air was introduced, so it sucked up less accumulated fuel and vapors, and as such, the engine did rev up a bit. 

I hope that my diary write-up can be of help to some-one else whe experiences the same symptoms, and seeks advice!

To everybody on this forum, even the so-called professionals who have forgotten to think instead of following procedures, thank you very much for your help!!

Greetings

Gunter