Late Model CIS-E Diagnostic / Duty Cyle Help
 

CAR: 1993 190E with 3.0L from 90' 300SEL (85k miles)


For the past several months I've been trying to diagnose/fix my air fuel mixture. I'll be as detailed as I possibly can. As I understand it from multiple sources I can read duty cycle from X11/3 port. If I want diag codes I have to do some other things and they will be output for me via that same port. I havent done this yet as I'm not sure about the procedure. I'll post what I've done so far. Hopefully someone can spot something...I'll post this across a few MB boards I'm on in case some people arent members there etc etc.

Heres the story:

In March, I just had a 3L m103 put in. It was running around 20%-25% Duty Cycle. I started the procedure to adjust it to the proper 50%. A friend of mine walked me though it. I increased the EHA about 1/6 turn. It brought the mixture to around 40-45%. Then I went to adjust the Lambda tower to fine tune it. I mistakenly did so with the car off (after the adjustment I stared the car, waited for the mechanical control to give way to electronics and observed the duty cycle) I did this twice. My final reading was still around 40-45%. As I was waiting for the mechanical control to switch over to electronic I decided it would be a good idea to clean my passenger seat area. Now whether this was coincidence or not, The duty cycle signal dropped off. It climbed up towards 100% and on the scope I had at the time went to 0v. My voltmeter tells me .09v.

And thats where I am right now.

I swapped EHAs with my old 2.6 motor thinking maybe it would do somthing (anything really to effect the voltage there) but it did nothing.

Then HefDMC over here clued me into the fact that, the 99% could mean something in the FSM. Sure enough I looked it up and the FSM said:

http://i98.photobucket.com/albums/l2.../Capture-3.jpg

I've checked:
Lambda Control. Everytime I adjust it. The DC will
appear then slowly fade down back to 100%.

Oxygen Sensor was just replaced and has no shorts that I can detect

Control Unit [CIS-E] was swapped into my buddy's 300SE and ran fine. His was swapped into my car and showed the same results.

EHA: has continuity between CIS-E and plug ends. Draws current fine. It considered to be working.

And they listed a procedure lower in the PDF on what to test if you have a 100% (non oscillating) DC: I'll lead you down the path I've taken.

It begins:

1) Test Adjustment of Lambda Control.
"On-off ratio can be regulated. Read Out Oscillating" OK / NOT OK.

2) NOT OK:
Disconnect 02 sensor signal. Check Voltage from CIS-E to ground: 450-1000mV.

3) OK:
Test Line to CIS-E control unit jack 8 for continuity.
Resistance approx 0 Ohms. Here is the diagram they provided for this:

http://i98.photobucket.com/albums/l2...gSchematic.jpg

Now, here is where it starts to fall apart. 1) my CIS-E doesnt have the same amount of connectors as this diagram. 2) I get no continuity at all between the CIS-E and o2 signal. However, I get the same result from a 90 300SE who is in working order and has good afm/dc

4) The last test in this category is the voltage to the heater lines of the o2. Mine Tested fine.

So this is where I am right now. I really dont know what else to do. I dont want to take it to the dealer and get reamed out the ass with high bills and what not.

If I were you, I would go ahead and build the LED tester to check for any DTC's. If you want any diagrams or help, let me know.
Other than that, good luck finding your problem.

slk,

Yeah. I think thats what I'm gonna do. If you could send/link me to the diagrams and procedure I'd appreciate it. Maybe that can lead me somewhere.

I'm still lost as to why the FSM says the O2 is connected to the CIS-E when I checked on two cars and got nothing.

Remember, the MB Duty Cycle is actually 1-measured duty cycle or the inverse of the observed duty cycle.

Does your DVOM have a duty cycle setting?

It does not. Its just a run of the mill DVOM. To get duty cycle I've been using this formula:

Duty Cycle = [1 - (V{pin 3}/V{max})] x 100%
where pin 3 is x11/3 and max is battery voltage when car is running at idle.


from http://www.landiss.com/mixture.htm

I used this schematic to build my LED Tester. I already had a momentary switch, some Banana Plugs, and some 18 ga. wire. The LED I purchased at Radio Shack was part #276-209, it has the built-in resistor.

The Red Plug with clip connects to the Positive side of the Battery, the Black Plug connects to pin #1 on the X11/4 (in front of battery), and the other color (I used white) goes to pin #3 on the X11/4 connector.

With this tester, I was able to retrieve DTC's, clear any DTC's, and put the CFI Module into Diagnostic Mode to check Duty Cycle.

so it just occured to me that in all these diagrams there a very large amount of pins and I only count ~35 pins on this unit I'm working on...

After googling the number it turns out that I've been working with the ABS controller.

Which explains quite a bit why there is no continuity between the o2 sensor signal and any pin 8. Yes. LOL indeed.

Update:

Upon my most recent revelation. I've redone all the continuty tests and verified that the O2 is infact connected to the CIS-E and has continuity, as does the EHA. I've also swapped the CIS-E unit into my buddy's 300SE and it worked fine. Duty cycle reading was oscillating as it should.

So from the FSM:

Lambda Adjustment too Lean Unknown
O2 sensor defective NO
Control Unit NO
Measuring Devices NO
No Current on Actuator Tested OK, not a lot of current though .02amps at time
Increased Idling Speed NO

Next thing: I'm going to swap EHA's with my friends 300SE and see where that gets me AFAIK its unadjusted where as both of mine are. So we'll see where that and messing with the lambda control get me. Maybe they'll do it.

I also wonder if my oxygen sensor is a problem. I've been using the Ford 302 sensor everyone has been telling me is the same as the Mercedes one just cheaper..

If all this fails to produce any results its on to the LED diag device.

What is your latest duty cycle reading?

I swapped EHA,CIS-E and OVP with my friends 300SE. duty cycle is still reading 99-100%. No oscillation. I am now stuck and have no idea where to go except that LED tester which is next on the list of things to do.

What the exact procedure for our 190's?

Ah. Case Closed.

Problem solved.

Lambda was way too lean.


It must have been really adjusted out of wack because I tried adjusting it at first a few time and it didnt do anything.

Please clarify, JD
 

Hello JD,

You say,

"Ah. Case Closed.

Problem solved.

Lambda was way too lean.

It must have been really adjusted out of wack because I tried adjusting it at first a few time and it didnt do anything."

I may be dim but I don't understand that.

What, exactly, was adjusted 'way out of wack?'

The EHA is adjustable but I've swapped EHAs (same part number for M103 and M104 engines, no change when I installed older M103 unit for newer, leaking M104 unit - no, it wasn't the grommets!), and never tried an adjustment.

The CO is adjustable (measuring duty-cycle) via the pillar above the KE mechanical MAS.

The lamda sensor is in the exhaust and feeds info to the rudimentary electronics which, in the KE, is the strangely-named MAS box usually stored behind the battery. I understand that this info fine tunes fuelling according to the O2 content in the exhaust.

Any clarification of your experiences would be much appreciated.

The best to all from rural France.

RayH

PS: Mine is a 300CE-24 Sportline - M104 engine with KE Jetronic. Suffering a few issues at present including:

1. Refusal to reverse sometimes but if switched of for a while, no problem.

2. Can be made to stall when hot it accelerator is 'flicked' up briefly to, say, 1500 then released. Might well stall. Otherwise, idle is dead stable.

3. 5-speed box behaves like a pig. Accelerates and changes well until 4th is called for. It engages 4th then will hold fourth for a second or two if powing hard. If not, it'll change immediately to 5th. Thereafter, it wants to stay in 5th and, reluctantly find 3rd (NOT 4th) when speed gets low enough.

My view is it's a vacuum problem. Investigating further but keen for input.

Sorry, I don't mean to hi-jack your thread bit if I have, I'll be outside the head''s office tomorrow morning, after assembly, for a carpeting and six.

the adjustment on the lambda tower was set too lean. I had to enrichen it to restore duty cycle readability.

Now, since you said your car is CIS-E M104 this stuff should apply to you as well. I know that a few early M104 engines were CIS-E based before switching to the newer stuff (Cant remember what its called exactly...)

From what you've said you seem to have the basic understanding of the system down.

The CIS-E "brain" looks somthing like this:
http://img527.imageshack.us/img527/8500/img0790zq1.jpg
Note: this is not the exact cis-e unit for your car. Its the same general shape. This one is taken from a 8 cylinder engine.

The oxygen sensor, which is in the exhaust pipe, is (signal wise) connected to it. There are two heater lines that go to the oxygen sensor, one of them comes from the MAS (fuel pump relay).

The general procedure as I understand it is as follows (this atleast works for the US version of the car, not sure about european)

1) First off you want to do this after the car has reached operating temperature ~80C. This way the oxygen sensor is fully warmed up. Attach duty cycle meter to +pin3, -pin2 of x11 connector on fender of car.

2) Start car, observe duty cycle. For approximately 30 seconds or so the engine will be under mechanical control and dc should be at 50%. Once the mechanical control gives way to the electronics the dc should fluctuate. If not, if it stays at a specific percentage for a long period of time then you have a error and need to consult the FSM for what you can do about it. In my case it was 100% which the oscilloscope I had didnt show because voltage dropped off to .09v.

3) If your dc is fluctuating then you can proceed with the adjustments.

4) If you're EHA has never been messed with. I'd recommend leaving it as is unless you have a serious dc issue and need the extra fuel.

5) Adjusting the Lambda Tower: this part may take sometime to master. I'm not sure where it is at on an early CIS m104 but on the m103 there is a tower that is next to the intake flapper thing. You stick an alan key down the mechanism until it reaches a set screw. If you push this set screw down too far you'll stall out the engine.
http://www.landiss.com/graphics/adjust1.jpg

6) You want to turn the alan key in small increments. On the m103 clockwise will richen the mixture, counterclockwise will lean it out. As you adjust it you want to let the engine settle back down to a steady idle and dc readout. From there continue adjusting until you've found you're setting. I know some people that set it for 40%. I tried to get mine around 45-50% I didnt have a dc meter to read it with so I used the formula to calculate : DC = 1 - (Vpin3 / Vbattery) * 100

Additional Reading: http://www.landiss.com/mixture.htm
http://www.importedcarparts.com/arts/1986_MERCEDES_BENZ_CIS_INJECTION_SYSTEM.HTML

Thank you JD
 

Yes, that's pretty much how I understand things. My only real problem has been at idle - it's better now than ever before but it annoys me that it isn't 100% correct.

That's aside from the transmission problem, of course.

Thanks again and atb to all.

RayH

Duty cycle stuck at 88% at idle, warm engine
 

I hooked up my fancy multimeter and ready the duty cycle and its stuck at 88% which according to the fsm is a faultly air intake sensor or wiring or fault in the communication between the 2 pcms. The car stalls when it reachs operating temperature. Did a fuel pressure test, passed everything. New mas, ovp.

digging around the recesses of my memory, you should (in order..)

check system pressure
check control pressure
adjust open loop CO
verify duty cycle varies when closed loop

there's more of course, but that's the quick and dirty.
One other thing to check/verify is that with the boot off, clean the air flow sensor housing and plate, top and bottom, using 3 .010 feeler guages, slightly loosen the bolt holding the plate and using the 3 feelers, center the plate and tighten the bolt.

While I haven't worked on too many CIS-E MB's, I've spent the better part of my adult life working on Audi's, VW's, and Porsches.

Bentley does have a good book on Bosch fuel injection systems,but you'll likely have to dig around for one.


erie