Check Engine Light 94 E320
 

I have a 94 E320 58000 miles. The vehicle operates flawlessly. Smooth idle, 22 MPG fuel ecnomy, no problems in starting, engine temperature stays between 80 and 95. Everything works fine, except the Check Engine Light. I always get DTC 19 code a few days after resetting and along with it either 3 or 4.

Arther and Suginami are helping me. I have the following checks planned. Pin 8 and see why code 19 is coming up. I will also check the fuel pressure regulator and try to clean the EGR Tube and see what happens.

Enigne Gasket, Aux fans, EGR valve and wiring harness have been replaced. 60000 mile service has also been done, new spark plugs, air, fuel filter, transmission etc.

Mercedes Benz dealer diagnosed my problem as follows:
Function Test W/HHT/ HFM Codes 035 O2 Sensor
Stop/Mixture to lean, 051 Self Adapt too rich, Replace O2 sensor still at Rich stop, Replaced Mass Air still reads flat, Replace over voltage relay, reading still flat, Diag Faulty HFM Control Module $2200.

If anyone else has any idea's or similar past experiences, please let me know. Steve and Jim seem to be the experts on the Bosch Control Modules on this site. I did notice that the LH-SFI module had some software issues.




http://www.peachparts.com/shopforum/showthread.php?threadid=15363&perpage=15&highlight=chip%20upgrade&pagenumber=1

http://www.peachparts.com/shopforum/showthread.php?threadid=4806&highlight=engine+control+module

<<<Function Test W/HHT/ HFM Codes 035 O2 Sensor
Stop/Mixture to lean, 051 Self Adapt too rich>>

AllData -code 035 e320-124 chassis 104 eng HFM-SHI

O2 sensor system operating at rich limit, mix too lean

Remedy/Causes:

Intake air leak
Fuel injectors
Reg pressure diaphram

BC you also see an AIR code fault, I would look at the vac hose
that goes from the engine intake , front of engine [ white plastic hose under front engine cover ] to the AIR/EGR Switch over valves... on 104s , these dry out and break, causing an air intake leak [ causing the lean condition] and that will bring up the AIR and/or EGR codes along w/19 ..just a possible, but common 104 problem....
Post what you find at pin 8 on HFM-SFI module...that will help
I don't think HFM-SFI computer had the problem of the LH-SFI

I have purchased all the neccessary stuff from Radio Shack to make the LED/SE tool. I also bought the Sears Multimeter with the duty cycel. I was trying to see if the multimeter can be used for the engine codes, but after researching, have found that the LED/SW tool is the best method to get the Codes.

Will post the codes as soon as the Check Engine light comes on again. Has been off since I reset it a few days back. It still has code 19. I believe the Check Engien Light only turns on when DM gets a second code like 3,4 or 5.

>

Don't have to wait for CE lamp indication to retrieve stored codes from other modules..
CE only comes up when faults effect emmissions paremeters..
So, go to both 8 and 14 and see what is in there..
Pin 1 is gnd [ neg], pin 16 is positive [ 12v] and the odd colored
wire goes to appropriate pin [in this case , both 8 and 14]

Arthur,

I got the following codes:

DM LED Code 1

PIN 8 Code 13 (O2 system at rich limit, mixture too lean, Intake air leak, fuel injectors, diaphram pressure regulator)

PIN 14 Code 5 (Stop lamp switch S9/1). The instrument panel light bulb warning had come on. I Replaced a brake light bulb a few days back. Maybe that's what this one is.

I opened the front engine cover and looks like all the vaccum lines are in good shape and connected.

I repeated the procedure a few times to make sure that no other codes were present, and the codes above kept on repeating. I did not clear the codes.

Now clear all codes and see what comes back...

Looks like everything is pointing to a fuel pressure check.

What dissapoints me most is the level of expertise at Mercedes Benz dealerships. William Auto World in East Lansing, MI has recommended changing the ECM after getting Code 35 (Mixture lean) without check the Fuel Pressure!

Rock Hill Imports in Kingsport, TN just changed the Wiring Harness, and informed me Friday that they could not cover the wiring harness under the goodwill policy, after talking to the factory rep, due to the age of the car! Delarship in Atlanta has done this for hundreds of cars. I will have to talk to the factory rep myself. Wiring harness did not need to changed at that time, the ECM Code has not changed.

I gave Rock Hill Imports the print out from this FORUM on the LH ECM, mixture too rich, on the second visit for the same Check Engine Light problem, and the tech concluded that changing the module was the the correct diagnosis based on DM CODE 19, wihtout realizing that LH is different from HFM! and no other testing.

I requested them to change the Aspirator Motor for the climate control and they changed the Cabin Temp Sensor, I had to tell the service manager Grag Durham where the motor was and he agreed to refund the labor and parts for the sensor change. Aspirator was on the service requested form for the Tech!

I had requested Williams Auto World East Lansing, MI to change the ACC cabin filters and they informed me that my car had did not have any!! They did not know that the filters are located below the windshield wipers.

Both the dealers above handle both Mercedes and BMW. Hard to believe!! They should have checked the fuel pressure after getting the code from ECM. None of them did, $2200.00 (ECM Change) MI dealer, I declined, and $884 TN dealer (wiring harness) which they did, and it did not solve the problem. If I had know about the harness issue, I would have gone to the Atlanta dealer and got it replaced under the Goodwill policy, but at that time, I was not aware of the this FORUM.

I would like to thank everyone on this FORUM again for the help and the information data base that has been created.

Arthur,

I cleared all the codes and got Code 1 from, both Pin 8 and 14. I started the car, turned it off and checked again. Code 1 again from both Pin 8 and 14.

Code 1 is No Codes stored [ simply meaning you have successfully cleared the faults from the memory of each module]

Now drive the car and go back in a couple of days ..

Don't wait for CE lamp, as that only comes on after other module
detected faults talk to DM and DM decides if that fault will effect emmissions..if it does not , CE won't come on even if other mods have stored codes..[that is why we go to other mods when you get ignition/fuel related probs that finally get to CE/emmissions paremeters]

Unfortunately, this means of diagnosis really is limiting. The info that is necessary is the adaptation numbers and you must have a scanner for that.

The code 19 problem happens in any system. It means you have reached the limit of electronic adaptation. The LH problem arose over LH systems in hot humid climates that got richer with great regularity. At lower partial adapt numbers of about .85 the limit was reached.

OBDII cars have the same problems with lean running AMMs where they bump the limits at 1.32 or so. They trip P0170 and P0173 codes. But notice the numbers. The early cars only had 15% correction built into the software. The MB solution was new software encased in new hardware. As Jim and others have found, that software rides on a 36k chip that is easily readable and new ones programmed. Why MB makes everyone buy the hardware over is beyond me. Anumber of companies allowed updated software by just replacing the chips (BMW, Volvo and GM come to mind)

Part of the point is that with failures the system logically can't correct for ever. The new ones with 32% mixture correction on both sides of the original equation 1.000 are really amazing.

So when dealing with this issue the amount of correction is one consideration, but solving the deviation is another. If the mix has gone rich due to pressure changes then fixing that is the point.

If the thing runs "rich" at .85 and still has closed loop control then software can solve the problem. New software (MB list at over 1800 usually) will allow that .85 car to go to probably .68. The 119 LH cars that were fixed with software seemed to get long term correction at about .82 from what I have seen and heard. In these cases the software is a good solution.

I am pretty certain I have had code 19 on HFM cars that were at the .85 not .68 level.

A few comments . . . .
 

I didn't know that E320 engines developed the "dreaded" DTC19 problem.

Of course, I know that V8s from '92 to '95 develop this problem and recently, it happened to me. The fix was to change the eprom which was successful. Monitoring the SA after the replacement eprom show a continued 'downward' trend to 0.8319.

At that point, I replaced the O2 and MAF sensors. After a month plus, the SA when back to 0.8534 at the last measurement.

But replacing BOTH of those parts is not cheap! An added bonus, was the increase in gas milage to 16mpg on the last fillup. At CA prices, it may pay for the parts, soon!

I believe that in my case it is border line. I reset the code in the ECM module, Code 035 W/HH and code 13 with impulse scanner(Mixture too lean, O2 at Rich stop limit), drove the car for 35 miles, checked the codes and got Code 1 (No system faults). Over a period of time, it probably goes a little above the 15% adaptation limit and then gives the code.

But, I agree with Steve, the proper fix is to correct the Air/Fuel supply issues.

My fuel economy is in the 20-24 MPG range. Does running lean have any adverse affects on the engine?

<<But, I agree with Steve, the proper fix is to correct the Air/Fuel supply issues>>

Interestingly enough, I have had 104 code 19s that were vac related [lean] and the fix on one was a intermittant /sticking open purge valve.. [ common part failure on these]. MOT on left fender well... feel for pulsations..archieves
This was finally found by plugging each vac system and driving the car for periods, waiting for a code.

The other was a sticking EGR valve .. this is also common on 104s and there is a TSB out on the part replacement..
Both of these vac leaks are a little tricky to find BC they are Internal leaks and intermittant..
The EGR was easier BC I also had EGR codes.. but the purge did not code out..
Granted , I did not have access to limit values at the time , but have not had a code 19 return on either since...[2 yrs]

Check Engine light came on, right after starting. Got the following codes:

DM Code 4 (Air Injection inoperative)
Code 19 (Fuel Injectors open/short or Emission system at adaptation limit)

Pin 8 Code 13 (O2 at rich limit, mixture too lean)
Pin 14 Code 1 (No faults)

WJM replaced the chip. I would like to replace the chip as well to increase the adaptation range, along with Air/Fuel supply investigation. Below is what WJM wrote about the LH chip mod:


"I have a copy of the Service Bulletin that I received from Steve some time ago. Post your fax # and I can send it to you.

As closure to this particular issue. It turns out the the difference between the computer Mercedes recommends upgrading to and the original is the software on the chip... and at that only 18 bytes of it.

I've run into a few additional people who have had this problem. We went ahead and simply replaced the chip with the newer version (got a copy from someone who had the new computer) and it has so far solved the issue and eliminated the check engine light #19 code.

Willy"

Willy, what is the procedure for upgrading the chip?

I think you need . . . .
 

Arthur's expertise for your car.

I can 'repair' this problem for V8s but do not have a chip for the 6's.

WJM (Willy) and a group of people, including myself, had problems with V8s. And I can supply a chip for any of these cars that fixes the problem. BUT NOT FOR 6s. Maybe WJM has this capability???

Because you seem to consistantly bring up the AIR fault code along with 19, I would give a fast check to the AIR system, specially the vac lines.
You mentioned that you have checked the main engine vac feed
line to the SOV that I mentioned in an earlier post, so I would next check the one that goes from the AIR SOV to the AIR Control Valve. This is the valve between the Pump and the exhaust ports 3 and 4 in the head that, when vac from the energized SOV reaches, allows the pumped air to get to the CAT. [for rapid cat heat and hc emmissions aid]
If this line is leaking/off the Air control valve, the AIR system will not work and a code will pop. At the same time , the line can cause a lean condtion.. enough to do a 19, ???? prob not ... but worth a look see.
The reason I mentioned the EGR and Purge valve leaks before is that these are both large lines and could pop a 19. [ too much for the stop limits to correct]
If these lines/ valves seem to be OK, check and see if pump is getting power, etc. for that code..

the Service Bulletin yo mentioned was for the v8 only. It said to replace the computer giving me the #19 code with another one with a different part #.

I found someone on this forum who had the upgraded chip, so we copied it and placed it in my computer. It solved the problem, as the only difference between the 2 computers was the software on the chip.

Anyway, if there is a newer version of the 94 E420 computer, maybe we can grab the chip from it.

If you get a chance, send me the BOSCH and Mercedes part number from your computer. Maybe we can see if the 95 had upgraded software that we can test in your car.

Just a thought.

Willy, I think you mis-read this thread...
 

the car is a '94 E320, not a '420'! Nuff said!

Yea, I knew that. It is a typo above. I meant to say that we should check to see if there is a newer version of the computer available for the E320... possibly on a 95.

Yes, maybe it's me who . . .
 

can't read? :(

This is the first instance that I have seen of a non V8 having this problem. My tech has never seen one either, and he has worked on Mercedes for 30 years!

Yes, if such a 'chip' is available, it would be advantageous to Zafar.

There is an ECU part# change , but Benz specifies changing
part# for part# [ same] when changing ECU..

Another item I did not mention that may be of interest on trim
code issues:
The led/sw access code system also has an ECU memory reset/reactivate procedure that allows the adapt memory to get back to "Mean". When changing parts/repair after codes , this should be done ..
I assume this was done when 02 sensor was changed, as I believe it is auto reset when done w/HHT...
But, for the benifit of theDIYers w/o HHT scanners....
- After one clears stored fault codes, code 1 will appear- verifying you have cleared the codes...you can now reset/reactivate the ECU adapt memory..
The Reset sequence is:
After bringing up code #1 [ No codes stored],
Wait 2 secs
Hold led/sw 6-8 secs
Turn off key 5 secs.
Turn on key 10 secs.
Start car

This is not to be confused with the simple 'Clear Code" procedure
of the Led/sw... that procedure does not get to adapt memory..
This procedure can only be done AFTER a code 1 Flash.....
Someone here may know more about this and/or have a better explanation...

Good info.

JimF... you need to add this to your web site if it's not there already.

Willy

excellent idea WJM. I will post the part numbers this evening.

Zafar

Steve,

According to Alldata, the 94 E320 ECM has 25% adaptation range.

"Mixture Preparation Self-Adaption

The Lambda control system determines fuel injection duration precisely so that the fuel/air ratio is kept constant at Lambda level 1 (equals 14.7 kg air to 1 kg fuel) under all operating conditions.

Should malfunctions occur in the form of:



Intake air teaks,
Injector wear or carbon build-up,
Engine wear,
Contact resistance in mass air flow sensor,
Defective diaphragm pressure regulator,
Defective purge switchover valve,
the engine control module automatically corrects the fuel/air mixture by adjusting the fuel injection duration.

The degree of correction is constantly calculated and permanently stored if the following conditions are satisfied:



Engine coolant temperature < 52 °C at start.
Engine coolant temperature between 80 and 100 °C while driving,
Intake air temperature between 0 and 55 °C.

Air/fuel mixture self-adaptation occurs at idle and under partial load. The maximum degree of correction towards rich or lean is 25%. After fuel injection system or engine repair work is performed, the engine control module will automatically adapt itself after approx. 10 trips. After eliminating a malfunction or after trial installation of an engine control module from another vehicle, the self-adaptation feature must be reset (see "Resetting and Reactivating ECM Memory").

Willy, I am familiar with that . . .
 

reset technique but it didn't work on my car and two others. It's supposed to but . . . .

So nobody has found a upgraded eprom for the 320 engine???

I got the following numbers from the ECM

Mercedes:
124 440 0410 (19-03-93) (29-03-94)

003 545 56 05

Bosch:
0265 106 080
CG284631541
013 545 8632

Ok,

Here's the scoop.

Try searching for a used ECU here:

http://car-part.com/

What you're looking for is a computer with the same "BOSCH" number (0265 106 080) but a different Mercedes part number. Try getting as much info fom the box as possible,

I have no particular info on the numbers but I can tell you that there is no way in H that you will find the same Bosch number but different MB numbers. bosch even creats a new number for a part when they sell the same part without the MB number on it.

Every change of any kind gets a new number. Software especially.

Years ago Bosch sent a flier around (to Bosch Service Centers)saying that they would provide no control units till 5 years after a car was produced as they (at that time - at least 15 years ago) couldn't tell until atleast that amount of time the final software the manufacturer intended to call appropriate for that year.

With the current software situation that can change daily.

According to Benz ADM-410 Notice"

94/95 E320-104

As of chassis C117898 , use 016 545 57 32

[There is one of those at WJM site - $300
They list it coming out of a '95 ..]

Before C117898, use 017 545 32 32

They do not recommend interchange...

Steve,

2 years ago, you were kind enough to fax me the Service Bulletin for the Code #19 problem that the 94-95 V8s were having.

In my instance, the old problematic ECU (code #19 problem) for my 94 E420 and the replacement one that Mercedes recommended had different "Mercedes" part numbers, but they indeed had the same exact "Bosch" part number. That is why I had assumed that the only difference was the chip's software.

I simply found someone that had the newer computer, copied the chip, and placed it in my old computer to resolve the issue.

FWIW

Willy

Based on all the information so far, I am planning to do the following:

1. Verify that a 95 E320 ECM adapats upto 32%, to get 7% more adaptaion than the current software limit. Find someone with a 95 E320 and figure out how to copy the Chip.

2. Replace the Diaphram pressure regulator.

3. Find out about the Purge switchover valve.

4. Have the Mercedes Benz dealer check for Intake air leaks.

5. Find a used speedometer cable (e-bay) to clean the EGR valve pipe.


From the Shell website, I found that the gasoline storage life is only about 6 months. After that it can cause problems. My car was not driven much from 94 to 03 (58K miles).

ECM controls the Air/fuel mixture by adjusting the fuel injection duration. I think that the ECM is opening the fuel injectors, but the injectors are not getting sufficient fuel flow, causing the lean condition.

Any suggestions to the above course of action?

I had the Mercedes Benz dealership check the Fuel Pressure. This is what they found:

3.4 bar at Idle
4.1 when accelerating
Remove all vaccum from regulator diaphragm and pressure max's to 4.2 Bar.

Check for leak on Intake Manifold

Checked intake manifold for vac leaks, found no VAC leaks at intake or any vac hoses

Any idea's?

For starters, I would run a good FI cleaner through it..
You may well have a gunked up injector [ as you mentioned long sitting]
The o2 sensor reads the leanest cylinder , so even just one screwed up injector can cause the ECU to get Lean message ..
I would also get the part # off the EGR valve to see if it has been change to the newer one.. there was a TSB on this for sticking valve...

Arthur, I read this at Alldata, Could the purge switch over valve cause the mixture to go lean?

COMPONENTS
The EVAP system consists of a non-vented fuel cap, vent valve (51), charcoal canister (77), purge switchover valve (Y58/1) and the routing hoses. The canister contains activated charcoal and a hose connects it to the throttle valve via the purge switchover valve.

OPERATION
When pressure in the fuel tank reaches 30 - 50 mbar (0.435 - 0.725 psi.) the vent valve opens allowing fuel vapors to flow from the fuel tank to the charcoal canister where they are absorbed.

With the engine running at a coolant temperature above 80°C (176°F) and a specific intake air mass range, the HFM-SFI control module (N3/4) applies a current to the purge switchover valve. The valve opens and vacuum from the intake manifold reaches the purge valve. Thus, the purge valve opens allowing fuel vapor from the charcoal canister to be drawn into the intake manifold via a port in the throttle plate.

In order to control the purge quantity, the HFM-SFI control module activates the switch-over valve at a frequency of 5,10 or 20 Hz (cycles) depending on engine load and other conditions . The purge quantity is determined by the duration of each cycle.

<<<Arthur, I read this at Alldata, Could the purge switch over valve cause the mixture to go lean?
>>

I mention that on the first page of these post..
The vave or any of the vac lines to it can cause lean condition.

To repeat.. I had a lean condition 104 that had a purge valve that was stuck in the open position.
These valves get gunk in them and stick..
A sign of a good one is to hold in in your hand and feel for a Heartbeat pulse...
It is on the left fenderwell and says 'MOT" on top.
Both the purge and egr can leak/stick and not show as vac leaks BC they are leaking within the system itself, as opposed to hissing / open to ambiant leaks..
I see you have AllData ... they have all these simple test procedures in there..

You will notice in the alldata info on purge that the ECU opens/closes the valve in little pulses.. [ that is the heartbeat feel I talk of]. The reason the ecu does this in little gulps is so that
the engine does not stall or run rough..it has time to absorb this
unmetered air in little shots, so you can see how a stuck open valve can upset A/F mix...
Usually, with both egr and purge, one will also get codes .. but not allways [ specially with intermittant sticking]
Some guys have success w/cleaning the valve with WD-40
[ archieves]

Arthur, I held the Purge valve in my hand and pulses were present. Does this mean that it is working properly or is it possible that it is partially stuck?

I hope that it is the purge valve, because the next items to check would probably be the fuel and air injectors.

PURPOSE
The air injection reactor system pumps air from a maintenance-free dry air filter mounted on the pump into the exhaust ports of the cylinder head to promote oxidation of CO and HC. The heat created by the oxidation process also brings the catalytic converter up to operating temperature more quickly.

OPERATION
Air Injection takes place for a about 110 seconds after start up and only when the coolant temperature is between 10 °C (50 °F) and 40 °C (104 °F), the engine speed is less than 3600 RPM, a specific intake air mass is not exceeded and the oxygen sensor is in open-loop.

Under these conditions the electromagnetic clutch of the air pump (125) is switched on by the HFM-SFI control module (N3/1). The HFM-SFI control module simultaneously powers the air injection switch-over valve (Y32) which provides vacuum to the air injection shut-off valve (126) in the pump circuit. Air is then pumped through the check valve (127) and into the exhaust ports of the #'s 3 and 4 cylinder in the cylinder head.

The heartbeat test usually indicates an OK purge valve..

I would still run some injector cleaner through it before changing anything..

Arhtur,

I added 16 oz of Valvoline Fuel Injector cleaner, and 16 oz of Valvoline Fuel system cleaner to half a tank of gas and after a few hundered miles of spirited driving (at speeds which the public safety departments of most states would not consider prudent), when the fuel warning light came on, I checked the Codes and got the following:

Code 13.

I got a new code: Code 20 (Self adaptation at idle speed too rich or lean).

I added another 16 oz of Valvoline Fuel system cleaner and filled the gas tank with Shell Octane 93.

I will check the codes again after this gas tank empties. I have been paying very close attention to the Idle, although the RPM needle barley moves, a slight periodic vibration can be felt, which seems to have gotten better with the fuel system cleaner. An Inline six engine is inherently balanced and is one of the smoothest engines except a V12, without the use of a counter rotating balance shaft. That was one of the reasons I choose the E320 instead of 400E. I was not expecting to feel anything at all during idle.
With transmission in D, nothing can be felt at all. The RPM needle stays rock solid.

Checked engine oil, after 4500 miles, still looks new and the engine did not consume any oil. I am planning to change the oil at 5000 miles with Valvoline 10W30. What oil do you recommend?
At the 60K mile Z3 service, the Mercedes Benz dealer had used the oil mentioned above. I was considering either 10W30 or 10W40.

I don't recommend any particular oil brand , as long as it 's Benz Spec.. I use Castrol GTX 10/40..
What I will recommend to you on the oil change is to do an archieves search that I posted way back on a TSB -re: oil level change on 104 eng.
They were having trouble with oil fill being too high and the TSB
changes the capacity [ I think it was about 1/2 qt less, but check for sure] ...

I would run this tank and maybe 1/2 of another w/o the FI cleaner and then change the oil, clear all codes [ write down what you have retrieved] and then watch for new ones

The reason I would clear the codes at oil change is you will then be assured of no gas contaminated oil [ which can bring on an Adapt trim limit code]]

At the same time . I would also do the trim memory reset procedure after you get the 1 Flash, no code OK ...
just to get back to Mean values

As far as idle is concerned , the 104 are famous for bad coil to plug connectors under the 3 coils ... if you have not done them, I recommend it , as they are short $$$ and most tech do them at plug change BC of their short life...[$7-8 ea]

The computer stops the injectors from opening when it detects a skip at the plugs, [ to protect the cat from getting raw fuel dumped into it],so these can also cause a lean reading back to the DM,,,
As you can see , all systems are related and affected , one way or another...

Mercedes dealer changed the coil to plug connectors. Service manager sketched the coils with the connectors and the bottom connector is the one which goes bad. Hopefully the Tech replaced the right ones, about $14 each and $20 labor. All three coils were done.

That idle condition seems to be better.

Since I was there, I had them change the oil as well, they had Mobile 15W40. Next oil change I will go back to 10W30.

I cleared all the codes, and checked them a few days later. Still getting code 13 (mixture too lean or rich).

What can be checked next. MB has added a layer of complexity for emissions controls. Some choices like delayed transmission shift can have adverse effect on a cold engine.

Again, it drives fine at all speeds, like a new vehicle.

Is it possible for the secondary air injectors to keep on running even after warmup, causing the lean mixture?

Some time back, after some hill climbing and decent, I had code 20 (mixture too lean or rich at idle or partial load) but that has not returned since. This was before the fuel injector cleaner, connector and oil change. Two other DM codes I had were code 3 (lambda inoperative) and code 4 (air injectors inoperator), along with 19.

Speedo needle vibrates a little below 40 MPH, can that have any effect?

Replaced the ECM. Problem fixed.