Having searched the archives and found a similar discussion, I was hoping to request further clarification. The originator of the original discussion is graciously assisting me in figuring out this problem, but I wanted the collective wisdom of all members. My check engine light came on, although the car seems to be running well. I pulled code #19, and have seen the TSB which calls for a replacement computer, particularly in the Jacksonville area. I live in Southern California, and this has apparently always been a California vehicle.

I obtained a replacement computer from a salvage yard bearing the same part number, despite knowing that a new part number was available, since this had taken care of the problem for another member. For nine days, the check engine light stayed off, but then suddenly reappeared. Clearing the code out and shorting out the 1 and 4 connectors did not prevent the check engine light from coming back on very quickly (less than 30 minutes of driving). I am sending the computer to this other member as he was friendly enough to try it out in his car to see if we could isolate the problem to either my car or the computer. Now that I have switched back to the original computer, the light has not come back on, even though I have made many trips, some up to 90 minutes at a time. While I fully expect the check engine light to return, I was hoping for an explanation as to why it would not immediately return, as opposed to this delay.

If the computer works normally in another E420, what are some of the other causes for this behavior? I believe that error code 19 suggests a fuel injector problem, but like I said, it seems to be running well in all respects.

Finally, the idle on the car seemed to be a little lower than normal (under 500 RPM), and it would very rarely stall at a stop light. I went ahead and used Redline Fuel System Cleaner on my last tank of gas, and it is now idles at about 700 rpm. I don't know if Redline would make such a difference or if this is simply a coincidence, or if, in my boundless optimism, a previously clogged fuel injector could have caused error 19 and the check engine light problem described above. I guess my real question is what is the normal idle speed for the 94 E420, and is there any adjustment? Thanks.

NO idle speed adjustment! However cleaning the throttle plate area with scotch-brite pad might help the idle speed to return to normal. To get rid of the check engine code(19) the modified LH ECU is the best fix.

. . . well documented . . .
 

screedchan, well written report!

Have a friend whose '94 S420 does the exact same thing! He, unlike me, is not concerned now that he 'knows' that it may be the computer and not an injector. His car runs as smooth as glass.

In all of the cases I seen on this forum, it always seems to be the ECU! Be interested in the conclusion.

Idle speed. . .
 

ALL-DATA CD-Rom says for engine temp 60- 80C, idle speed sb 550-750 rpm.

same problem with 400E
 

I have the same problem with my 400E. My check engine lights is and is idling to low at a stop. The problem with mine is that whenever I start my car, I have to keep my foot on the gas pedal to keep the engine running. Whenever I let go of the gas, the engine dies. But when I let my foot of the gas slowly and I tried to keep the idling, the engine will continue to run and I'm able to drive, but it would still idle low. Would this be a similar problem to the 94 E420?

Thanks

same problem with 400E
 

I have the same problem with my 400E. My check engine lights is and is idling to low at a stop. The problem with mine is that whenever I start my car, I have to keep my foot on the gas pedal to keep the engine running. Whenever I let go of the gas, the engine dies. But when I let my foot of the gas slowly and I tried to keep the idling, the engine will continue to run and I'm able to drive, but it would still idle low. Would this be a similar problem to the 94 E420?

Thanks

MBDOC

I understand that swapping the computer is the "best" fix, but that implies that there are other fixes available. Moreover, is there an explanation as to why the check engine light does not immediately come back on when the same computer has been pulled out and returned to service some days later? Finally, what exactly is going on within the original computer which causes the check engine light to come on, and what is different about the new computer? Is this related to heat, age, milage, humidity? Why are cars in Jacksonville more prone to this problem?

Sorry for all of the questions, but inquiring minds want to know! Also, inquiring minds are loathe to spend $2000+ on a new computer. Thanks.

. . . self adaptation
 

Is there an explanation as to why the check engine light does not immediately come back on when the same computer has been pulled out and returned to service some days later?

If it was the 'self adaptation' feature at its limit that caused the CE light, it takes a number of 'trips' to change to what it was. A 'trip' is defined as; 5 mins of engine operation; speed greater than 5mph and tach greater than 500rpm.

So starting the car and shutting it off is not considered as a trip. Also running the car for 90 mins is considered just one trip (assuming the other parameters were satisfied).

If an injector is bad, i would think you surely would notice it! The car should idle roughly.

So it does takes some time for this to happen but you seemed to indicate that you have satisfied the 'trip' requirement and light has not come back. Don't know why!

I'm the one testing screedchan's new computer. When I get it, I'll run it for a few days (at most).

I'll also replace the chip with the "updated" version, as I got it from someone else who went through this same problem some time ago and replaced the ECU to the new Part# recommended by Mercedes.

Hopefully it will work.

Willy

Since the fuel system is running *RICH* that means that the return voltage fron air mass sensor is too high!! OHMs law says that resistance can lower voltage, so yes there are other repairs! if you can figure out the return voltage from your air mass sensor to the ECU! The DM has to see the LH-ECU at that limit(.85) to turn the check engine lamp on.
CLEANING the throttle body DOES help with LOW idle speeds.

What is causing the car too run rich to begin with? What exactly is going on with the ECU to cause this? Also, how difficult is it to remove the throttle for cleaning? Thanks.

AS the service bulletin says the problem is in the ECU's chip program. MB screwed-up & should have recalled as many of these cars fail emission testing! At out dealer we notified our customers & installed REMAN ECU's when they were still under warranty. MB stopped with the reman/exchange units as soon as they were all out of warranty!
The throttle body can stay in the car for cleaning, only the air mass sensor & connecting rubber boot have to be removed. WE use a fine scotch-brite pad & air intake cleaner, make sure to wear rubber gloves. I always vacuum out all of the crud before starting.

wjm, who contributed to this thread and graciously agreed to test out my replacement ECU, mentioned that he obtained the new chip from a ECU which superceded the ECU in my car. Does this mean that switching out that chip will take care of the problem, or is there another chip which controls fuel delivery? Thanks.

. . . MAF
 

. . only the air mass sensor & connecting rubber boot have to be removed. WE use a fine scotch-brite pad & air intake cleaner, make sure to wear rubber gloves. I always vacuum out all of the crud before starting.

Does this help the fault code '19', the low idle speed, or hopefully both? I think it just helps the low idle speed. The reman ECU fixes the DTC 19? Or as Willy said, the new EPROM chip?

Like clockwork, exactly seven days after I put my original ECU back in, the check engine light came back on, again reading fault 19. I really don't understand why it would take so long to come back on, especially since I probably put 1000 miles on the car in that seven days. Now that has reared its ugly head again, I know that it will continue to come back on within 30-60 minutes after I clear the code. Can anyone explain this strange behavior? I am still holding out some faint hope that the revised chip from the newer computer will help with this problem, but who knows with something as screwy as this. If anyone has any ideas, I am all ears.

I see a couple possibilities. One is that you have an intermittant fuel pressure problem causing the adaptation problem and the second has to do with the time interval after sitting.

The adaptation thing really needs an explanation first to see whats happening. The car comes with a calculation based upon sensor readings and it starts running the car. The computer monitors the mixture at the exhaust O2 sensor. In early models we had to adjust the base mixture so that the ability to correct had equal distance rich and lean. With adaptability the car now corrects the base mixture by averaging its prior immediate history. If it is adding more fuel than the base calculation it moves the middle. It has only so much capability to enrichen or to lean out the mixture. When the average can't be corrected any more either rich or lean, the code is set.

Most of this is a software issue and leaving a computer disconnected for extended periods resets fully all the adaptions to base calculations. This does not happen by just resetting the code.

There also is a service bulletin pointing out that scan data can not be used for evaluating MAF sensors as the computer will creat a substitute value if its other values calculate a problem. The ability for the computor to recalculate for suspected problems can also sometimes pass through code resetting. A scan tool can reset adaptation values to base settings but long term disconnection or battery disconnection with the pos and neg cables put together to remove capacitor memory, can also reset the values.

I don't know what the new controller does but I bet its a totally software issue. They have probably increased the adaptive range or decreased the rapidity of corrective movement.

. . . self adaptation
 

Self adaptation can be reset by grounding pin 1 to pin 4 for 8 - 9 seconds on the X11/4 38 pin diag connector provided that there are no other codes in the system. This will 'reset' the ecu to 'mid' range. It then 'learns' and adapts as you drive.

I've done this reset for the '94 S420 described previously and it works for about 4 - 6 days (not sure how many 'trips' that is). So if you want to reset it, you needn't pull the battery leads.

But it always comes back as yours does. So far it's the ECU!

BTW, what happened with Willy's tests of your ECU?

[Edited by JimF on 07-04-2001 at 01:32 AM]

Willy is trying out the new chip but has yet to give me the verdict. I do intend to have the fuel regulator checked as well, since that might be causing the problem. I don't know if this is related, but I did mention a low idle problem, and today, noticed that it is more pronounced when coming to a moderately hard stop. The tach needle dips below 500, and most of the time it returns to somewhere above 500, but every so often the car stalls. Does this suggest some sort of related problem with fuel delivery?

Replace the {slide}

What?

. . . same as run.exe
 

Your problem seems to be similar to run.exe (Scott). Maybe he is 'monitoring' and can give you (us) a run down of what fixed his problem.

His idle was also low (below 500rpm) and would stall on hard turns left or right. His car however is not the same, a 500E (or E500).

Do a search on RUN.EXE to review the thread. I think it has about 50 or more replies!

. . . injectors?
 

This DTC states "Fuel injectors open/short or self-adaptation (SA) in LH-SFI control module (N3/1) at limit".

In my friend's S420, the cars runs well and seems to idle perfectly. But I wonder the injectors could be leaking (a little bit) from all (or some of the injectors) such that the SA is forced to its lean limit.

Since it's just a little bit from all of the injectors, couldn't that cause this failure mode? It seems as though it could.

Maybe the opposite is happening; the injectors are cutting off a bit (due to age, dirt, etc), so it forces the SA to the other end of its limit (rich).

My point is that, so far, I haven't seen where any/all of the injectors were replaced. The ECU is $2200 retail, whereas the injectors are about $18 wholesale each (I think). I'm trying to get my friend to try this.

as a good newbie, i did a search and it brought me to this thread (and a couple others). my girlfriend's cousin is flat out sick of her 1994 E420 because it just stalls at random. she even GAVE it to her kids (30 somethings) and they don't want it.

so, from what i gather, i should first clean the throttle body. worst case is that i need a new throttle body. the car only has 29,000 miles on it. in this post, it says that the computer was not properly calibrated from the factory. does this cause the car to stall or just run rich? i hear chip talk going on here. this seems like a far cheaper idea than a reflashed ecu. or can i give my ecu as a core and get a "corrected" ecu pretty cheap?

wouldn't a rich running car eat O2 sensors and converters like no tomorrow?

can i get an aftermarket maf that could be calibrated to adjust the rich condition? we ford guys have pro flow, i even have on on my SHO.

i am learning that all cars are fairly similar. being this mercedes E420 and my SHO are both pre-0B2 (pre 1996), we have a lot in common. on my SHO, erratic idle can be caused mainly by the idle air bypass valve. it's a mechanical spring loaded sensor that continuously adjusts idle for smooth operation. problem is that as it gets old, the spring gets out of whack which results in erratic idle and stalls. i know my gf's 1997 audi A4 has an IAB, does this mercedes?

does this stalling problem result in erratic idle (jostling 250 rpm or so until stall) or does the car just flat out stall. if the car dies out of nowhere, can it be the cam position sensor or the crankshaft position sensor?

i am very eager to hear your responses. i sure hope i can nab this car as my SHO now has nearly 123,000 miles on the clock. time for a new sled.

see ya,
Robby

Update to DTC 19 problem . . . .
 

Wow, this post was started in Nov 2000. Time goes by quickly when you're having fun!

The final resolution to this problem was the replacement of the fule computer's eprom with the newer version that contains the opened limits. They were changed from 0.85 and 1.15 to 0.68 and 1.32. Quite a difference.

Replacing the eprom has fixed EVERY car that had the dreaded DTC19 problem including all 'S' and 'E' class cars! I believe the count is something like 6 or 8 cars.

About 4 months ago, it even occured on my S500 (104K miles), so I replaced the eprom. S/A is running about 0.85.

If you're interested in permanent fix for the DTC 19 problem, send me email for the details.

Re: Update to DTC 19 problem . . . .
 

awesome!

basically, what is involved? is it something a non mechanical person such as myself can replace? i do have friends that are very car capable though. where do i buy the eprom? how much is it? any reason to find a newer one from a junked MB (cheaper)?

hey, you aren't the guy with the MB ATI supercharged coupe?

Robby

Robby, check your email. . .
 

for a private reply.

Basically, you need to disassemble the fuel computer module and replace the internal eprom with the newer one that contains the enhanced limits. It's not that hard but worth the effort provided that the problem IS the DTC 19 code.

Jim,

Wow... unreal that this thread is still alive. It's been a while since I've popped my head in the forum.

You still burning chips for folks with the #19 error code? I have not done one in a while. I have not had the error since I reprogrammed my chip... works like a charm.

Hope you're doing well. I see your website has MANY additions since the last time I was on.

Willy

Still driving his 94 E420... have about 85k now.

Willy, glad to see you. . .
 

still kicking! Yes, it has been awhile.

My S500 turned 104k miles recently and it's doing fine although I have a interrmittant 'ticking' lifter.

The dreaded DTC19 happened to me at 102k. So replaced the eprom and it fixed the problem nicely. In fact, the chip has fixed ALL cars that had the problem, about 15 or so!

A few weeks ago, I replaced the O2 and MAF sensors. This changed self adaptation [went higher] and gas mileage to around 16mpg city. Have left the new eprom in.

Re: Willy, glad to see you. . .
 

ticking? ouch. what happened? your engine has hydraulic valves, correct? makes finding/fixing the tick a PITA?

my SHO has solid actuated valves. if there's a tick, a shim is probably out of spec. replace it and be on your way.

is this common in MB V8s or am i just being neurotic?

Robby

Do a search on 'ticking' and. . .
 

you will find this post:

http://www.peachparts.com/shopforum/tech-help/37931-tick-tick-fixed-tnx-donnie.html

The end of these oil tubes blow out and cause the lifters to loose oil supply. Had this happen to me as your read from the above post.

Now the tubes are fixed and it appears to be a worn lifter that needs replacing but I don't know which one! So, since it's intermittant, I live with it for now.

This is NOT a common problem as far as I know.

The "Dreaded Code 19" - FIXED!!
 

All,

I ended up with Code 19 as the reason for my now-working (another story) CEL. That usually means a new ECU - very expensive fix. A guy named Joey on the Technical Help forum at Mercedes-Benz Club of America F0RUMS gave me a great lead: see Jim. I then checked with JImF who had an engineered solution that avoids putting in a new ECU. He's at http://pages.prodigy.net/jforgione/MB_S500.html. His replacement E-PROMs are programmed identically to those in the newer ECUs from Mercedes, and contain the software fix for the adaptation limits. These limits were too tight in the earlier ECUs and are usually the cause of the Check Engine Light being turned on by Code 19. The dealer would charge $2200 plus installation. Jim's E-PROM is in the $60 range, and the longest part of the install is driving to a tool dealer for the proper Torx bits (if you can, bring the ECU with you to ensure you get the right sizes). It's a doggone miracle! :)

Alan Case
87 300E (for sale on eBay)
94 E420

You're welcome!
 

As the self-proclaimed author of this fix, I'm glad Jim was able to help you out.

Another happy customer! :-)