Duty Cycle Readings 91 M103 300E - PeachParts Mercedes-Benz Forum

JohnM. · #1 03-10-2010, 07:24 PM

Figured I'd start a new thread since I'm delving into CIS-E troubleshooting here. Basically my car failed emissions pretty grandiosely. HC 3.4 (Limit-1.5), CO 120 (Limit-20). I replaced the O2 sensor and my semi-stuttering idle, and cold/hot starts got a lot smoother. I took it back and failed even worse than before. I know there is something broken in the fuel control system.

I will mention that there are no driveability issues at this time. It's got power everywhere and idles smooth. There are no driving "symptoms" other than being a gross polluter. This is not a Cali car.

I don't have access to a meter than can do Freq readings ATM, so I just used voltage. With the engine off (ignition on) between pins 2-3 of X11 there is a 12.2V reading. I noted that deflecting the air flow plate DOES NOT change this number at all. Both WOT and partial deflection of the plate have no change on the duty cycle reading.

With the engine running (warmed, idle) the voltage between pins 2-3 of X11 is 9.4V and is fixed. The 9.4V will jump to 7V (50% duty cycle) only when the engine is revved semi-hard. If you raise the throttle to a steady, high RPM it still stays at 9.4V.

What's the next test?

400Eric · #2 03-11-2010, 01:18 AM

Another one, this time out of right field: Maybe the ECU hasn't adapted to the new, good O2 sensor yet? How long have you run the car with the new O2 sensor? I've been told it takes at least a week of everyday driving for the ECU to fully adapt to a major change unless you reset it. GSXR told me how to reset ECUs but I forgot. It's probably on his site though.

I know I'm reaching here but it's all I can come up with.
Regards, Eric

JohnM. · #3 03-11-2010, 01:58 AM

Quote:

Originally Posted by 400Eric

Another one, this time out of right field: Maybe the ECU hasn't adapted to the new, good O2 sensor yet? How long have you run the car with the new O2 sensor? I've been told it takes at least a week of everyday driving for the ECU to fully adapt to a major change unless you reset it. GSXR told me how to reset ECUs but I forgot. It's probably on his site though.

I know I'm reaching here but it's all I can come up with.
Regards, Eric

It's been over a week and 600 miles on the new O2. I really think it's the ECU. There are some tests I found to see if the ECU is really doing anything. I'm off to the Upullit yards tomorrow to try and score a 90-92 ECU regardless, and probably a spare EHA, throttle switch.

loepke72 · #4 03-11-2010, 09:29 AM

When you say you deflected the sensor plate, did you only push down on it? If so, you would not have seen a change in the duty cycle since the system is already at the rich limit and can't do anything more. Try grabbing the nut in the center of the plate and pull up while checking duty cycle. Pulling up on the sensor plate will lean out the mixture. If the duty cycle changes, then readjust the mixture at the adjustment screw on the air meter assembly. If not, then it will be necessary to measure the voltage at the O2 sensor and see if it changed with mixture and find out why the rest of the system is not working (failed ECU, EHA, etc.) All this assumes that the fuel pressure readings are in spec to begin with.

Remember, with a rich air-fuel mixture, the CAT will not have enough O2 to do its job, so it will look like a failed CAT at first glance.

ps2cho · #5 03-11-2010, 01:22 AM

Your duty cyle at those two pins should fluctuate. This is indicating you are not achieving a closed-loop system. As you said as well, the fact that deflecting the plate does not change it means there is a problem.
You don't need a frequency DMM for this. Voltage is good enough. 13.6v = 100% rich, 0v = 100% lean. You want a fluctuating voltage around ~6.5v

Quote:

Originally Posted by landiss.com

Assuming the "standard" 49-state controller (70% duty cycle with ignition on but engine not running, or about 4 volts average), the following tests should be performed before attempting to adjust the idle mixture.
1) Idle Contact: Deflect the air flow sensor plate. The duty cycle should decrease to 10%. If it remains at 70% test the throttle valve switch (Job 07.3-121).

2) Full Load Contact: Open the throttle completely. The duty cycle should decrease to 20%. If it only decreases to 40% test the air flow sensor potentiometer (Job 07.3-121).

Try the above and report back.

Also, pull the error codes because it might be something the system already has stored.

Of course, your cat could be suspect, but with your system not going into closed-loop, this should be something that needs to be addressed regardless.

JohnM. · #6 03-11-2010, 08:24 PM

I built the engine code reader today, man, that's nifty.

. It threw the following codes in this order.

13 Intake Air Temperature signal
2 Throttle Valve Switch (full throttle contact)
3 Coolant Temperature Sensor

Is 13, that stupid little "bung" sensor in the air boot, right before the air cleaner?

2 Is testable, but my car shows no symptoms of this switch being faulty. But I have not tested it per factory, yet.

3 is really interesting. That was one of my top guesses before I grabbed the codes. Because the ECU would think the car is always cold, and dump fuel accordingly. Thus preventing closed loop. I changed all of the sensors on the head when I did the head job two years ago, and thankfully I saved all of the old sensors! So it's just a matter of swapping that 4 pin bastard out! .....Lets see...

sptt · #7 03-11-2010, 09:08 PM

My 91 300E passed California emissions but idled horribly. Since then, I switched out the o2 sensor which made the biggest improvement and next was the water temperature sensor. Then I manually adjusted the air mixture and my idle has been great. However, I'm afraid I mat have screwed with the emissions now and lately the car has been hard to start (after many months with no issues).

hookedon210s · #8 03-11-2010, 09:10 PM

Quote:

Is 13, that stupid little "bung" sensor in the air boot, right before the air cleaner?

Yes. Its values are testable with a VOM.

Quote:

2 Is testable, but my car shows no symptoms of this switch being faulty. But I have not tested it per factory, yet.

Doubtful this is the problem

Quote:

3 is really interesting. That was one of my top guesses before I grabbed the codes. Because the ECU would think the car is always cold, and dump fuel accordingly. Thus preventing closed loop. I changed all of the sensors on the head when I did the head job two years ago, and thankfully I saved all of the old sensors! So it's just a matter of swapping that 4 pin bastard out! .....Lets see...

Its values are also testable with a VOM. Do a search on this site to find the range of values for a given temperature. Mark

400Eric · #9 03-12-2010, 12:46 AM

The CTS is only 2 years old. I find it hard to believe it is bad already unless it has been bad from the beginning. How old are the other two sensors?

Since you have plenty of power at full throttle, I can't help but wonder if that throttle valve switch hasn't failed in the "full throttle" mode. That would sure make conditions rich all the time and prevent closed loop operation too.

JohnM. · #10 03-12-2010, 01:44 AM

God I sure hope it's not the Throttle Position Switch. That would mean removing the entire FD/Air flow assembly. Not fun! But this is yet another good possibility! I was confusing this part with the "Decel microswitch" on the throttle linkage!

400Eric · #11 03-12-2010, 02:25 AM

Please post the results of your tests. We'll be waiting. This is way more exciting than the outcome of that stupid "Bachelor" show that my wife makes me watch.

JohnM. · #12 03-12-2010, 03:33 PM

OK. So I replaced the 4-pin coolant temp sensor and it seems to run a bit better yet.... Lambda reading is still weird though. Now, it fluctuates, but only by .5v. It's still reading 9.2V at idle. When you open throttle quickly it jumps to 7v, and then back to 9.2v at idle. If you stop the engine and restart...the idle reading is 7v again! Then as soon as you rev, it and let it drop back down to idle...it's back to 9.2V.

Don't really want to mess with AFR adjusting until I'm sure there are no broken parts that would skew the reading. What next....?

JohnM. · #13 03-12-2010, 05:47 PM

Decided that I should at least try to mess with duty cycle a little bit. Adjusted to 50% with warm engine idling, but it still jumps to rich (9.2-9.4) when RPMS are brought up to 2500rpm. Obviously still an issue.

ps2cho · #14 03-12-2010, 05:49 PM

Fluctuating by about .5v is right. Is it doing it fast according to the volts or slow?
When you open the throttle it should change, but 9.2v at idle is still running very rich. Give the mixture a 1/4 turn anticlock (lean) and see if it that helps bring the volts down to ~6.5

JohnM. · #15 03-12-2010, 05:55 PM

Quote:

Originally Posted by JohnM.

Decided that I should at least try to mess with duty cycle a little bit. Adjusted to 50% with warm engine idling, but it still jumps to rich (9.2-9.4) when RPMS are brought up to 2500rpm. Obviously still an issue.

Quote:

Originally Posted by ps2cho

Fluctuating by about .5v is right. Is it doing it fast according to the volts or slow?
When you open the throttle it should change, but 9.2v at idle is still running very rich. Give the mixture a 1/4 turn anticlock (lean) and see if it that helps bring the volts down to ~6.5

We both posted at the same time

. I also tested the TPS and it does appear to be indicating the WOT signal and the idle signal correctly. All that leaves is, maybe, the ECU.

It was really weird though. When adjusting the lambda, I went just a snitch lean from the previous adjustment (7.5v) and then the Lambda just started dropping all the way down....slowly.....from 6.5v to ~2.5v....then I adjusted it a snitch richer and it was back to 7.5v. Why did it free-fall there?