EHA, Duty Cycle confusion - '91 300CE - PeachParts Mercedes-Benz Forum

rgnprof · #1 02-25-2011, 09:03 AM

I have a '91 300CE Coupe, M104 motor, and I have been dealing with poor gas mileage issues for some time. I had replaced the EHA valve a while back and fooled with the mixture some, but the EHA started leaking (used one). Replaced it yesterday with a brand new one and now I'm having problems. Two observations: 1) this car has always run better when cold. 31 degrees here today and it starts great, accelerates smoothly, lots of power, no hesitation - feels like a new car. But, as it gets warm, it's starts and stalls, idles rougher, get's sluggish, etc.

2) With the new EHA valve in, I adjusted the mixture to right around 50% - but when I get it to 2500rpm, the mixture jumps all the way to 14-18% - runs very rich to where you can really smell it. I am assuming this is contributing to my poor mileage. Why is it doing this? New O2 sensor just installed as well...Any thoughts?

ryan

mbdoc · #2 02-25-2011, 09:36 AM

Are you measuring the duty cycle with the purge system disconnected & plugged off?

rgnprof · #3 02-25-2011, 10:54 AM

I wasn't at first, but then I did unplug it - didn't cap it off though...

rgnprof · #4 02-25-2011, 06:54 PM

OK, I really could use some help here.

Capped off the purge valve, and got the same results. Idle duty cycle around 50%, at 2500 rpm the duty drops to 10-18%, let off the accelerator and returns to 46-53% +/-...

Hooked up my EHA harness and took it for a drive - fluctuates around 0 mA at idle (-1 to +1, again, give or take some). At cruising speed, EHA mA between 6.5 mA and 9.0 mA's. Let off the gas and coast and mA's return to around 0 - hit the pedal and immediate jump to 8 mA's. In an old post, Steve B. syas that when you disconnect the o2 sensor, mA's should return to 0 - mine doesn't - stays around 1-2 mA's.

One other observation is the idle speed. I am having trouble - while performing these tests to get the idle to return to 650 rpm's. Sometimes, I have to really gun the motor to get it to idle back down - as it wants to stay around 1000-1200 rpm. This is new...

It really smells rich to me when I am sitting there in the driveway revving motor to 2500 rpm, with duty cycle hovering in the teens.

Any help?

ryan

JohnM. · #5 02-25-2011, 07:50 PM

When you start the car cold it's in open loop and will always have perfect AFR which is why it runs good. It's obvious once your car is in closed loop (warm) is when things are going wrong. Something in the EFI system is sending the wrong signals, causing your rich condition.

EHA is the main controller of mixture above 2500rpms. But since you've already replaced the O2 and EHA, I would test the TPS (or whatever its called on these motors), and the idle control valve. One of those may be causing the idle and/or the running issues when warm. Have you tested the main coolant sensor on the head? Also a bad ECU could be causing the problems in closed loop. Last but not least, there is a remote possibility your cat is plugged. But that's about the order I would look at things in. Everything in the fuel system is testable with a decent quality VOM. Good luck.

rgnprof · #6 02-26-2011, 01:52 PM

Some follow-up:

I tested the 4 pronged coolant temp sensor on the front of the head and it seems to be in spec. Around 0 degrees C this morning and ohms measured 5000 one way and 5500 the other, warmed up to 90 degrees C and measures 225 - 250 ohms.

But, following the Static tests according to the landiss.com page, I am getting some confusing results. If I put the black lead from my Sears 82139 into pin 2 on the X11 and the red lead in to pin 3, I get 70% - so far so good. But when I check the voltage, it reads 8.28 volts. Using the Landiss formula - 8.28/12 = .69, then 1 - .69 x 100 = 30%. HUH? (I did the same test on my '86 560 SL and it works according to the landiss specs). Going to step 2, when I deflect the air flow sensor plate, the 70% does NOT change at all. Again, on my '86 560SL, the duty cycle drops to 10% - as described in the landiss document. Back to the 300CE - the full load contact test is weird as well - if I read the 70% duty cycle (red in 3, black in 2) then the duty cycle jumps from 70% to 80% when I open the throttle completely - if I switch it around and get 30% (putting red in 2 and black in 3), the duty cycle jumps from 30% to 20%.

With me so far? Man, if anyone can sort this out - I would greatly appreciate the help!

Anyway, I think I have a bad throttle valve switch on the 300CE - but I can't find the part ANYWHERE. There is a part number on the switch that is attached to the throttle body - 201 540 42 45 - but this brings up the entire throttle body assembly - for $350!

Can you not just replace the switch? Any help?

ryan

Texholdem · #7 02-26-2011, 02:03 PM

I am selling an Mercedes 1990 300CE Fuel distributor complete with MAF.

It is $99.- at the moment. Return guarantee. Maybe it fits your 1991. Check it out.

http://cgi.ebay.com/ebaymotors/Mercedes-W129-W124-300CE-SL-Fuel-distributor-MAF-_W0QQcmdZViewItemQQhashZitem45f8f5c36bQQitemZ300529599339QQptZMotorsQ5fCarQ5fTruckQ5fPartsQ5fAccesso ries

JohnM. · #8 02-26-2011, 03:23 PM

Throttle valve switch, which I referred to as the TPS is located underneath the fuel distributor assembly and attaches to the intake manifold. It is not a cheap part new, I would recommend good used. This is most definitely the culprit.

rgnprof · #9 02-26-2011, 06:25 PM

John,

Your '91 300E has the M103 - this is the M104 motor. I can't find a Throttle Valve Switch for this car anywhere...

The switch is attached to the front of the throttle body - has a part number stamped on it - 201 540 42 45. ********** is the only place that number brings up a part and it's the entire throttle body, with the switch attached for $350+.

You can see the throttle body in this ebay link -

http://cgi.ebay.com/ebaymotors/Mercedes-Throttle-Body-1991-92-300SL-300CE-0021402453-_W0QQcmdZViewItemQQhashZitem3cb4db6724QQitemZ260732315428QQptZMotorsQ5fCarQ5fTruckQ5fPartsQ5fAccesso ries.

ryan

JohnM. · #10 02-26-2011, 06:28 PM

You may technically have an M104, buts its more like M103.5, your car uses all M103 fuel injection parts with very minor tweaks to the FD only. The throttle switch is not sold separately for whatever reason, you will need the entire TB assembly. The assembly on my 300E looks exactly like the one pictured. Like I said, a good used one will do you just fine.

JohnM. · #11 02-26-2011, 06:41 PM

I have done some research and can say for certain that the 300E throttle body is the same as the 300CE, so finding a spare one will be super easy and cheap. There is no extra part number for the 91-92 CE. Obviously, the later 24V HFM fuel injection used from 93-95 has a completely different TB.

rgnprof · #12 02-26-2011, 07:41 PM

Thanks, John - appreciate the help. On the Landiss page, there is reference to a test for the throttle valve switch - Job 07.3-121, from the FSM, I assume.

Any idea on where I can get this - or, any advice on further testing of the throttle valve? Would like to do some more troubleshooting, before spending the money and committing the time.

ryan

JohnM. · #13 02-26-2011, 08:42 PM

The throttle valve switch is very simple in operation. It has two contact points, Idle and full load. It sends one of those to the ECU and the ECU compensates with fuel. Usually they fail in full contact position which just dumps fuel all the time. The ECU thinks the pedal is down even at idle!

After re-reading your post on the test you did with deflecting the air plate. This is to test the air flow potentiometer which is the black thing on the side of your FD assembly. So its possible your potentiometer is bad, not the throttle switch. I thought the test for the throttle switch was deflecting the air plate, but I was mistaken. But I'd say its either of the two at this point.

Here is a FSM for the fuel system on these cars. It has all the factory tests for anything you could want. And I've also included two pages from the Bosch FI manual that talk about the operation of the potentiometer and throttle switches.

http://pdfcast.org/download/mbenz.pdf

rgnprof · #14 02-26-2011, 10:25 PM

So, is it possible that the Landiss manual has these backwards:

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

"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)."

rgnprof · #15 02-26-2011, 11:03 PM

I just built the 'homemade' LED code reader. I have a 16 pin board and I can get the LED to light up (pin 1 is ground and I have to connect the red to the +battery), it won't blink at all. My understanding is one flash means no codes...I followed the instructions from posts here. Any thoughts?

ryan