300E Failed Smog Check!!! -- Causes?
 

I just got back from getting a smog check, and my car failed. Any idea on what the problem is? Here are the failed items:

2500 RPM Test
HC PPM: 208 (should be < 130)
CO %: 1.15 (should be < 1.00)

Idle Test
HC PPM: 287 (should be < 100)
CO %: 2.65 (should be < 1.00)

It also failed the EGR test.

The car is a 1993 300E 2.8 w/ 175K miles. I replaced the 02 sensor about 20K miles ago. The only other noticable problem the car has is a slightly rough idle.

So what is this? Bad EGR valve? Defective O2 sensor? Gas milage is still good...

On that car when the CO is high the most likely cause is the O2 sensor not controlling the mixture.
You also need to find out why the engine has a rough idle, that is also a reason for high HC's.

My car was at the limit of hydrocarbons when it was last smogged and my mechanic said this is often caused by a catalytic converter on its last legs.

Greg
'84 300D
'90 300CE

have you changed your sparkplugs recently?

Yeah mine did the same thing about a year ago, I think mines the Catalytic Converter, I didn't feel like dropping that much money into my car. So I just paid the $300 for a "certified" tune-up and they had to pass me. Altho when i came back in, for the second time, they had to check if I actually got the stuff put in, so I suppose that takes out giving a gas station $20 for the papers saying I spent $300 on a tune-up, oh well:rolleyes:...btw, it still works for tint! give'em $20 and they'll look the other way on the dark tint;).
So if everything dosn't work, get the tune-up.
Good Luck

FIRST THINGS FIRST . . .

Start by curing the EGR problem . . . then proceed to the NEXT phase . . .

That actaully very well be my current problem...But when I asked about it, I was told I didn't have an EGR valve, seemed a bit odd, but I just said "OK". EGR also has to do with Gas Milage correct? I posted a thread awhile back asking about it, and someone also said i didn't have one.
SOOO, maybe i should just go outside a check it out when I get home.
Yeah, first things first, i agree

I'll have to look at the 02 sensor again -- the sensor itself was working last time I tested it, but the heater wasn't working. Perhaps there is a break in the wire. Anybody know where (i.e. what pin) the sensor goes into the computer? I could at least test to see if there is continuity.

Your EGR problem is likely a blocked "hot inlet pipe" to the intake manifold.

It is a well-documented problem on these M104 engines.

Search the archives and you'll see tons of posts on the problem.

The blockage is where the pipe meets the intake manifold.

You can either ream out the pipe, or replace it.

Stevebfl (Steve Brotherton) also points out the problem with pictures in the DIY link at the top of the homepage under the article Evaluating Electronic Engine Controls, or something like that.

Thanks Paul. I'll take a look and see what I can do.

BTW, that new throttle actuator you sent me is still working fine. I took apart the old one, and you wouldn't _believe_ the poor condition of the insulation in the control cable. I cut it open and it just flaked off in my hand. I'm working on a way to replace the cable (easier said than done b/c of the customer mercedes connector); I'll post pictures on here for everyone as soon as I figure it out.

EGR flow has no effect on CO or HC but will cause the NOx to rise!
High CO readings come from excessive fuel. High HC comes from un-burnt fuel, ie; misfire.

Obviosly take M.B.Doc's advice before mine!

However, it wouldn't be a bad idea to take a look at your egr valve and the hot inlet pipe.

Stevebfl has posted that he reams out this pipe as part of the 60,000 mile service.

There was also a TSB on the original EGR valves that commented on a sticking problem.

MBDoc-

So you're saying the CO is caused by an excessively rich mixture; possibly caused by a bad 02 sensor. The hydrocarbons are caused by another problem, but probably not the EGR system. The other problem may be a misfire... (it doesn't really feel like a misfire -- anything else it could be?)

So, did I get that right?

Could a malfunctioning O2 sensor cause a 3x HC problem? Don't think it can swing it that far. This is more like a misfire and unburned fuel getting into the exhaust.

Given that the problem gets better as the RPM's come up, suggest vacuum leak caused by injector seal(s) causing a lean misfire at idle. Could also be plugs, wire, cap, etc.

If the EGR valve has a small piece of carbon in the valve seat, it can cause
Poor idle
Unburned fuel from misfire
Rich mixture (the exhause diluts intake air, so less oxygen + same fuel = rich mixture)
Also it will cause a failure of the EGR valve. It is tested by closing the valve and checking the vacuum and mixture, then it open the valve and checks again. It should see a change. It won't if the valve is already open and stays that way.
SO replace the valve, clean out the EGR passages, and retest.


Paul

But these plug up not stick open

Actually, the original EGR valves were defective and do stick.

There is a way to check if your EGR valve is not working due to a plugged up hot inlet pipe, or if the valve itself is sticking (open or closed):

The EGR is located on the exhaust manifold on the passenger side of the engine. There is a small vacum line attached to it which comes from a switch-over valve under the front plastic cover in front of engine.

The ECU sends an electric signal to the switch-over valve which opens the EGR valve to engine vacum. The vacum pulls open the diaphragm on the EGR, allowing some of the exhaust gas to recirculate back to the intake manifold (thus the name, Exhaust Gas Recirculation), reducing emissions.

The ECU will not send this electric signal at idle or wide open throttle, or until the engine has warmed up.
So, if you hand pull about 12" of vacum on the EGR at idle with a warm engine, you are creating a very lean condition (vacum leak) and the engine will attemp to stall and run very rough. This tells you not only that the EGR is opening, but also that the hot inlet pipe is not completely plugged up. If, when you pull vacum and nothing changes, then you know that either the valve is not opening up, or the hot inlet pipe is completely blocked.

Now, release the hand vacum and the EGR should snap shut. Engine should smooth out. If the EGR sticks in the open position, the engine will run rough at idle until the engine warms up.

There is another possibility that the diaphragm in the EGR itself may leak and cause an internal leak. The test for this is to see if the hand vacum will hold the EGR open for a period. It is common on vacum leaks to not be as noticeable with a cold engine as the engine is running rich and that helps off-set the added intake of air caused by the leak.

Since you are not complaining of an idle problem either when cold or hot, the EGR valve might not be functioning when in closed loop. It could be either stuck in the closed position or the pipe might be completely blocked.

By the way, I had this problem when I first bought my car, and Arthur Dalton walked me through this process. I first reamed out the hot inlet pipe, but the EGR code still came up. He informed me of the TSB on sticking EGR valves, so I replaced it with the new one, and problem went away.

The saga continues...

I spent most of the weekend looking for the source of my problems. Found some interesting things, but not the solution. Need some more advice:

First went looking at the 02 sensor. The ECU is not supplying current to the heater, and doesn't appear to be reading the sensor either. The sensor itself tested good, but when I connected a 1.5V battery to the input and ground terminals going to the ECU, I didn't notice any change in the engine. (I understand the engine is supposed to go lean, to the point of almost killing it.)

So is this a bad ECU? I tested for continuity in the wires leading to the ECU, and they are all fine, so that isn't the problem. Do ECUs even go bad? Its all solid state electroincs...

Second, I went looking for my EGR problem. I removed the EGR tube at the intake manifold, and you wouldn't BELIEVE how gunked up it was. I took more than two inches of gunk out of there. It was completely blocked up, and probably had been for a long time. However, that didn't solve my idle problem, which (probably) means it didn't solve the emissions problem, particularly if the 02 sensor isn't being read.

Chris 17H,

First, they did not "PASS" your car for the $300 tune up, State allowed you to licence it, but it is still a "FAILURE" this is called a waiver. Wont happen this year though, rules have changed.

Second, 87 300e's EXCEPT for the 1.9's and 3.0's have EGR's.

"So you're saying the CO is caused by an excessively rich mixture; possibly caused by a bad 02 sensor. The hydrocarbons are caused by another problem, but probably not the EGR system. The other problem may be a misfire..."

Both high HC and high CO are caused by insufficient oxygen in the combustion, or insufficient oxidation in the cat. An engine running lean to reduce both also produces high peak combustion temperatures - and the result of this is an increase in NOx. The EGR valve leaks some exhaust gas back into the intake, and this reduces the combustion temperatures, and thus the NOx.

So, it is reasonable that the same source of your high CO is causing the high HC also. Find why your ECU is out of loop control, and that will likely sove your problem. Note that the O2 sensor is the primary feedback device, but the ECU will also not function if it can't detect engine speed, throttle opening, airflow or any of a select group of other input devices.

Steve

From reading the archives, I understand I can put a new catalytic converter in there to reduce the emissions, since (with 175K) mine is probably dead. However, W/O finding and fixing the source of the problem, the catalytic converter won't last, right? I'll have to put another one in there eventually, yes?

So, this brings up the question: does anyone know HOW LONG it will last? Six months? A year? Two? Five? Once again, my emissions readings are:

2500 RPM Test
HC PPM: 208 (should be < 130)
CO %: 1.15 (should be < 1.00)

Idle Test
HC PPM: 287 (should be < 100)
CO %: 2.65 (should be < 1.00)

I won't hold you to it, but any educated guesses?

I still think that you should investigate this from a misfiring angle.

Make sure that your ignition components are in good shape.

Put a fresh set of plugs in.

Spray some carb cleaner around the intake manifold and injectors and see if you can find a vacuum leak.

See if you can isolate the miss by "cracking" the lines on the fuel distributor one-by-one. If you crack one and the miss does not change, then you have found it. Replace the injectors and seals on the weak cylinder(s).

Guess I forgot to mention that earlier. I already went hunting for this. I replaced all the plugs a couple months ago (they have about 8K miles on them now) and sprayed starter fluid around this weekend. No vacuum leaks. I didn't check the injector seals, but it doesn't matter -- I've traced the problem to the 02 sensor not being read properly. Problem is, it looks like the ECU, so I don't want to go off and replace it.

Can I put a new cat in to take care of the problem? How long will it last?

You may well have a problem with the sensor not being read properly. Your converter may be shot. Your fuel distributor may be hosed. Your ECU may be brain-dead. But more than likely, you have a fairly basic problem causing your rich mixture. In the absence of good information, I suggested some simple and inexpensive tests to try to rule out the most common problems.

But do go back and read Steve B's posts on the limitations of the O2 feedback loop before you buy a new cat.

And good luck, I'm outa here.

A well tuned engine should output about .5% CO and 50ppm HC, so as you can see, your vehicle has more than a little something wrong with it.

Things not yet mentioned concerning high HC readings: Base timing, advance/retard function,CPS, leaky injectors, vacuum leaks.

High CO:air intake restrictions: ie clogged AF, PCV, fuel delivery,EVAP system

Have you adjusted the CO setting?

Going back through posts I may have missed a reference, but if you have no readings at the O2 sensor, have you checked the OVP? If the fuse on top is blown, or the relay is bad, you will have no power to the ECU. Usually I think the ABS lite is on, but maybe not if you have the 2-fuse model.

Also, it is possible you won't have ECU functions if another input sensor - such as crank position, airflow, etc, were malfunctioning. I wouldn't assume the ECU is bad, under the circumstances you indicate.

Steve

Providing the OVP, etc is OK, I just ran across the following method to set A/F for the CIS using the CO screw.

I have not done this, however, the reference states to connect a dwell meter to the pulsed side of the frequency valve or duty cycle test terminal and select 4 cylinder ( typo? perhaps should be set to 6 cylinder?) adjust CO screw in AFM to arrive at 45* dwell.

His car is HFM, not CIS, and idle, etc. can't be adjusted. HFM fuel injection systems are designed so that idle speed can't be adjusted. Idle speed is completely controlled electronically.

Who's "he" I thought we were talkin about a 93 300E? Am I trippin? Thats not a CIS?

Thank you for the insight.

I see what I did, I replied to Chris17H, with a 87 300E, before, and got him and the other 93 300E mixed up.

1993 300E's have the HFM 3.2 liter engine. They are mis-badged. Read my signature below. I don't know why MB didn't call this model a 300 E 3.2 (like they badged the 1993 300 E 2.8), or 320E (like the rest of the world), or even E320 (like 94 and 95).

1993 300E's and 1994-1995 E320's have the same VIN #, engine, and chassis code.

Yeah, its HFM. Its actually a 93 300E 2.8, but it doesn't change much. It has the M104 engine.

I already replaced the OVP earlier, and no dice. I just ran outside to check the fuse, and it is still good. The computer _seems_ to have power, because the diagnostic port works, which is wired directly into the ECU. This car is the Federal version, and doesn't have a seperate diagnostic model.

It could be another sensor causing trouble, but logic would have it that the heater for the O2 sensor would still get power, even if the computer was ignoring the readings. Since we've got no power on the O2 heater, and its ignoring the readings, I still think it might be the computer (although I'd LOVE to be proven wrong!)

There is one other possible clue to this puzzle: the computer is also giving me a code 11 on pin 14 of the diag connector. This looks up to be:

-Closed throttle recognition signal to engine control module
-Fuel saftey shut-off to engine control module

Normally I'd say the CTP switch in the throttle actuator is dead, but that ain't the case here. I get this code with a throttle actuator I got from sugnami, and it is still good. So it has to be something else.

I have no idea if this sheds any light on the problem (it didn't for me), but I thought I'd mention it....

My '90 M104 failed last year with pretty much the same high readings. Problem was the thermostat. Engine would not warm up fast enough which causes engine to run rich longer which causes catalytic converter to load up which causes high readings. This can also cause catalytic converter to fail. If you find this to be a problem, it could take up to an hour of driving to lower readings once corrected (need to clear out catalytic converter).