CEL Code 26 revisited (Cold upshift delay)

[pryce27]

Hi guys,

3 months with the "new" 500E and it has been terrific. AC now working fine (needed new compressor) and the only thing left to do is a finicky check engine light that keeps coming on.

I have had it checked out by three mechanics. First guy said it was a "false code" and that I should replace the entire ECU for $4k to make it go away. I didn't like the sound of that.

Second guy said it was maybe some sort of minor vacuum leak but said just "not to worry about it" and offered to disconnect the bulb. I don't really like that either, I like things being fixed.

Third guy I dug a little deeper and he said the CEL was throwing "code 26" and said there is nothing to do and that I could read more about it on the internet. Of course I came here and read through it- sounds like sometimes it may be just a simple vacuum connection and other times it can be a bigger deal. Mine is not related to cold weather, it comes on about a week after I clear the code (have cleared the code 4 times now). Car runs perfect.

Question: the only modification I made was the same day I purchased the car when I had a first gear start installed- could this be the cause of the code? If so, anything you guys suggest to fix it?

Second question: if not from the FGS, should I specifically ask one of the mechanics to test the hoses and connections for leaks using a test device? I haven't been telling them what to do but maybe they just don't know the things you all do!

Any other suggestions appreciated, I hate having an awesome car with an illuminated CEL!

Thanks-
pryce27

 

[gsxr]

First guy is an idiot. $4k for something that won't fix it?!! Don't go back there.

Second guy has a viable option, but I agree it isn't ideal.

Third guy is wrong, there is something to fix, you just don't know what that is yet.

1) Changing to FGS valve body should have no effect on this. And, I've not read any reports of other people having code 26 after the FGS upgrade.

2) The fault is coming from the DM, which controls the CEL. It senses that the engine RPM's are not remaining above the minimum expected RPM, for a long enough time period, after a cold start. This means something in the cold upshift components is not working normally. After a cold start, the RPM's should hang around 2500-3000 before allowing an upshift... if this doesn't happen, code 26 appears along with the CEL. You'll need to check the electrical and vacuum plumbing. This is all DIY stuff or you can pay a shop $$$$ to trace it down. Worst case, the control cable on the transmission is defective, which is quite a bit of labor to replace (miserable job).

3) The modified EPROM that I sell should fix the problem symptom (click here). This EPROM allows disabling the cold upshift delay, and for reasons I don't fully understand, prevents triggering the CEL. It should help in this case. There is NO negative effect to using this chip besides the cost.

This PDF document explains how the system works. The last page has a diagram with component layout: http://www.w124performance.com/docs/mb/transmission/trans_cold_upshift_delay.pdf

 

[gerryvz]

I would test and replace the vacuum connectors and line at the switchover valve just aft of the brake booster. I had this same code a long time ago and it ended up being a split vacuum connector. Often it is a very simple thing like that. Less often, the switchover valves themselves go out.

 

[pryce27]

Thank you guys so much. I will pursue these options and let you know what happens.

 

[pryce27]

Thank you guys so much. I will pursue these options and let you know what happens.

Found a new mechanic who seems more willing to look into the CEL issue. I will be bringing him the car Monday morning.

In the meantime, something new has come up- last night I was cruising on the highway at about 70 mph and suddenly the ASR light comes on solid yellow and I lose all power. Floor the gas pedal and nothing happens. I pull over to the side of the road, turn the car off and on again, and everything works perfectly, ASR light is no longer on, I drive the rest of the way home with no issues. I can't seem to find this issue in my search, anyone have any ideas? Thanks in advance.

The red CEL is still illuminated, as before.

 

[gerryvz]

You need to pull the codes and see what happened. Clearly you want into "limp home mode."

You should not have had a "nothing happens" condition when flooring the gas pedal. There is an emergency mechanical linkage between the throttle pedal and the ETA, so that if the car goes into "limp-home mode" (which yours did by describing the symptoms), that if you press the gas pedal hard enough, you can have enough throttle (say up to 50-60 MPH) to give it enough gas to get home.

If you are saying that there was literally no way you could rev the engine at all by pressing the gas pedal to the floor, then you have SERIOUS problems with the mechanical linkage between the pedal and the ETA, and you need to see a mechanic ASAP.

The most common cause of "random limp home mode" is the neutral safety switch, mounted on the side of the transmission. There are many other causes, but that is the most common one. The ETA can also be bad with wiring on the outside, or on the inside (wiring and/or mechanical components) which is another major cause.

Random limp mode problems often go away when you turn the car off and then on again, but it's only a short matter of time until they re-appear.

 

[pryce27]

Update and solution:

The new mechanic I brought it to is a 50-year Mercedes mechanic who has seen it all. He was very familiar with the code 26. He tested the vacuum and the connections and it was all fine. I told him about going into limp home mode while driving on the freeway and he was able to map the problem to the ECU part # 1245450732. This was mis-communicating about the upshift delay (causing the code 26) and can also trigger the limp-home mode error.

I found the replacement ECU at a specialty shop in Orange County and he installed it- no more codes, no more errors, no check engine light. So far everything is golden.

Thanks to gerryvz and gsxr for the help- will update if any further issues.

 

[gsxr]

Hmmm. Definitely keep us posted. While the E-GAS module (1245450732) may b e the cure for the limp mode, it should not have any effect on code 26. At least, I don't think so, unless you *only* had the CEL along with limp mode...

 

[gerryvz]

I too do not believe that the CEL (if indeed illuminated with a code 26) issue and the random limp-home mode issue are related. Two completely different systems.

If the CEL re-appears (and sadly I think it will after a few more cold starts), you can check code yourself via the pushbutton and LED underhood, unless your car is an early 1992 model and is not so equipped.

Remember, too, that the upshift delay only happens during the first few minutes of engine operation from a cold start, and is employed through a vacuum actuator on the side of the transmission. It basically just holds things in gear a bit longer until some pre-set parameters are met. I believe the LH (fuel injection) computer controls the upshift delay, not the E-GAS (ETA control) computer, together with the switchover valve behind the brake booster (to supply engine vacuum) and an upshift delay switchover valve on the passenger side of the transmission.

I hope the E-GAS unit helps fix the random limp-home mode issue. It may indeed, but a bad E-GAS is not the most common problem we’ve seen for RHL mode over the years.

Typically, random limp-home mode has been caused by a bad ETA itself (i.e. the internal or external wiring, or internal ETA components are worn out) and even more commonly by the neutral safety switch mounted on the driver's side of the transmission. These NSSes tend to wear out through repeated friction of their internal electrical contacts.

I’ve personally experienced both of these issues and indeed lived with them for a while and been frustrated by them. In the end both were fairly simple to fix. I hope you have the problems licked.

One question I forgot to ask — what type of first gear start did you install? BergWerks, FGSSwitch or a 400E420 transmission valve body?

Another question — were there any other codes found that would have tripped the CEL besides a 26?

 

[pryce27]

I too do not believe that the CEL (if indeed illuminated with a code 26) issue and the random limp-home mode issue are related. Two completely different systems.

If the CEL re-appears (and sadly I think it will after a few more cold starts), you can check code yourself via the pushbutton and LED underhood, unless your car is an early 1992 model and is not so equipped.

Remember, too, that the upshift delay only happens during the first few minutes of engine operation from a cold start, and is employed through a vacuum actuator on the side of the transmission. It basically just holds things in gear a bit longer until some pre-set parameters are met. I believe the LH (fuel injection) computer controls the upshift delay, not the E-GAS (ETA control) computer, together with the switchover valve behind the brake booster (to supply engine vacuum) and an upshift delay switchover valve on the passenger side of the transmission.

I hope the E-GAS unit helps fix the random limp-home mode issue. It may indeed, but a bad E-GAS is not the most common problem we’ve seen for RHL mode over the years.

Typically, random limp-home mode has been caused by a bad ETA itself (i.e. the internal or external wiring, or internal ETA components are worn out) and even more commonly by the neutral safety switch mounted on the driver's side of the transmission. These NSSes tend to wear out through repeated friction of their internal electrical contacts.

I’ve personally experienced both of these issues and indeed lived with them for a while and been frustrated by them. In the end both were fairly simple to fix. I hope you have the problems licked.

One question I forgot to ask — what type of first gear start did you install? BergWerks, FGSSwitch or a 400E420 transmission valve body?

Another question — were there any other codes found that would have tripped the CEL besides a 26?

Interesting, will see what happens in the next week or two and if the CEL comes back. No other codes have been tripped to my knowledge. The FGS was a E420 transmission valve body, would that have anything to do with it?

Thanks for the informative explanation.

 

[gerryvz]

Generally, when a CEL appears and is reset, it takes a bunch of cold starts (I'd say 5-8) for it to re-appear. If you drive the car regularly say for a month, and you don't see another CEL, then it's probably safe to say that it isn't coming back (at least for the Code 26).

When I had CELs (most recently Codes 5 and 6, pertaining to the air injection and EGR systems), I would only have a few times driving the car before the CEL would come back. This ended up being a vacuum hose that had become disconnected from the smog pump vacuum valve (code 5) and a broken vacuum line down in a sheath where it wasn't visible (Code 6). I suffered with these codes for probably 6 years on my E500. The point is, that BOTH codes were caused by simple things, nothing complex or a mechanical failure. In my former white E320 wagon, with the M104 engine, a persistent Code 26 ended up being a cracked rubber connector to the vacuum line running from the intake manifold to the upshift delay vacuum switchover valve. It wasn't visually apparent until I removed the rubber connector and closely visually inspected it. Again, a simple thing, not a complex thing.

The CEL is generally related to the emissions system plus some basic but important engine inputs, such as coolant temp, knock sensor, cam timing, O2 sensor feedback loop, and intake air temp, not much else. Typically, these systems involved the delayed upshift, oxygen sensor, air injection and smog pump, exhaust gas recirculation (EGR), intake air temp sensor, fuel injectors, purge valve, knock sensors, or engine coolant temp sensor (for the fuel injection system, not for the dashboard gauge).

There are a few items relating to the ETA (which the E-GAS unit controls) that can trip CELs and store a code. These mainly pertain to the idle speed regulation, and wide open and closed throttle positions. You would probably know if something is wrong with the idle speed regulation (it would not idle at the normal ~500-600 RPM). The other wide open and closed throttle positions, again you would probably sense something was wrong.

For code 26, the MB diagnostic tree points directly to testing the LH/fuel injection system (as I mentioned earlier), NOT the E-GAS. I just don't see how a bad E-GAS would cause a CEL without a noticeable malfunction of the ETA itself (i.e. idle speed or throttle response). A Code 26 is going to point directly to something in the upshift delay system .... not anything else. It's pretty narrowly focused.

 

[gerryvz]

A couple of other notes on the Code 26 and CEL.

In addition to the switchover valve aft of the brake booster, it is not uncommon for the vacuum actuator on the passenger side of the transmission (which the switchover valve is directly connected to) to not hold vacuum, and thus also be bad and a cause for the Code 26. This can be replaced, if faulty (and you can run a test on all of these components with a Mityvac to see if they hold proper vacuum), by removing the rear transmission mount, and wedging the side of the transmission over a bit, so that you have better access to the passenger side of the transmission. You can do it with the car jacked up, but it's easier on a lift.

The upshift delay "feature" is indeed directly controlled by the LH (fuel injection) computer's programming. It is dependent on several variables: engine RPM, engine temperature, and engine load below 50%.

One other slight revision to my earlier statement: if the upshift delay system is not working, the Diagnostic Module (DM) will illuminate the CEL after TWO failures. The first failure will store a code, and the second failure will illuminate the CEL.

Also, if you warm up the car to a certain engine temp for a few minutes, the parameters to activate the upshift delay will not be met, and the LH unit will not use it, thus avoiding it.

 

[pryce27]

4 days out with two cold starts per day and so far no CEL and car is running fine. I haven't been warming the engine up, I just take it easy after cold starts. If it comes back your above info will be super helpful for next steps- will let you know!

 

[gerryvz]

They can take a little time to re-appear. If you've gone a month or so with no CEL, then you may be out of the woods on it.

 

[pryce27]

Further update: after about 6 weeks with no CEL, the light came back on. I took it back to the mechanic and same code was being tripped for upshift delay. He diagnosed that the auto trans kick down cable (part 140 270 0573) was not holding vacuum. He replaced it, which unfortunately required partially pulling the transmission (rear motor mount, oil pan, and rear drive shaft). $1000 fix, almost all of it labor.

But- I notice the difference! On cold starts it revs higher now before upshifting (upshift delay is working), and shifts are definitely smoother too. Seems to be running perfectly now. Hopefully this was the problem.

Hopefully this helps others who might face a similar problem.

 

[gerryvz]

thanks for the follow-up. Always appreciated.

 

[Kyiv]

@pryce27, who's your mechanic in Los Angeles?

 

[pryce27]

@kiev: Herbert at German Motor Car Service

We’re sorry. We can’t find the page you’re looking for.

The link has expired or can’t be found.

yelp.to

 

[luckymike]

Further update: after about 6 weeks with no CEL, the light came back on. I took it back to the mechanic and same code was being tripped for upshift delay. He diagnosed that the auto trans kick down cable (part 140 270 0573) was not holding vacuum. He replaced it, which unfortunately required partially pulling the transmission (rear motor mount, oil pan, and rear drive shaft). $1000 fix, almost all of it labor.

But- I notice the difference! On cold starts it revs higher now before upshifting (upshift delay is working), and shifts are definitely smoother too. Seems to be running perfectly now. Hopefully this was the problem.

Hopefully this helps others who might face a similar problem.

Am I to understand that the auto trans kick down cable referred to above is related to the upshift delay?

My '95 E420 always skips second gear. Unless I floor it from a stop, then it will spend a few seconds in second. And I've noticed lately that the kickdown to pass is not working unless I bottom the accelerator pedal.

This is next on the list after passing smog... Still waiting on rotors from FCP. Lost in transit.

 

[pryce27]

That is my understanding from what the mechanic told me. I am no expert by any means.

 

[gsxr]

Am I to understand that the auto trans kick down cable referred to above is related to the upshift delay?

Yes.

My '95 E420 always skips second gear. Unless I floor it from a stop, then it will spend a few seconds in second. And I've noticed lately that the kickdown to pass is not working unless I bottom the accelerator pedal.

That isn't right. Upshift RPM's from a dead stop should increase proportionally with throttle position. I.e., at 10% throttle, each upshift (at operating temp) may be at ~1500rpm. At half throttle each upshift may be at 3000rpm (roughly), at 75% throttle it might be 4500rpm, etc. If you are not seeing a distinct change in upshift RPM based on throttle position, something may be screwy with your Bowden / control pressure cable adjustment.

This is next on the list after passing smog... Still waiting on rotors from FCP. Lost in transit.

Hopefully just delayed from the winter storms last week? FedEx / UPS are still drastically backlogged in many areas.

 

[luckymike]

Yes.



That isn't right. Upshift RPM's from a dead stop should increase proportionally with throttle position. I.e., at 10% throttle, each upshift (at operating temp) may be at ~1500rpm. At half throttle each upshift may be at 3000rpm (roughly), at 75% throttle it might be 4500rpm, etc. If you are not seeing a distinct change in upshift RPM based on throttle position, something may be screwy with your Bowden / control pressure cable adjustment.



Hopefully just delayed from the winter storms last week? FedEx / UPS are still drastically backlogged in many areas.

Am I to understand that the auto trans kick down cable referred to above is related to the upshift delay? YES (gsxr)

--Sounds like there's a new part and some rack time in my future.

That isn't right. Upshift RPM's from a dead stop should increase proportionally with throttle position. I.e., at 10% throttle, each upshift (at operating temp) may be at ~1500rpm. At half throttle each upshift may be at 3000rpm (roughly), at 75% throttle it might be 4500rpm, etc. If you are not seeing a distinct change in upshift RPM based on throttle position, something may be screwy with your Bowden / control pressure cable adjustment.

--This isn't what happens at all. I can use any amount of pedal and the car is going to skip 2nd. The only time I get 2nd is by manually flicking the shifter quickly into neutral the back to drive, or from a DEAD stop, giving lots of pedal. Even in the latter instance, 2nd isn't held very long.

One last thing, If this kickdown 'cable' (isn't it actually a vac line?) doesn't hold vacuum, is it automatically bad or can one remove and repair the part?

Oh. I see now that it has both. But same question, can a vac leak inside the unit be repaired? Obviously one can replace the elbow.

 

[gsxr]

--Sounds like there's a new part and some rack time in my future.

Remember the Bowden cable is a huge pain to replace - you don't want to do that unless you can confirm it's defective. First verify both ends are connected properly. If you disconnect the ball/socket at the throttle linkage and pull on it, you should feel resistance, and it should pull back when you release. If not, the cable may not be connected to the linkage rod inside the transmission.

Click here for a How-To on this job... enjoy!

--This isn't what happens at all. I can use any amount of pedal and the car is going to skip 2nd. The only time I get 2nd is by manually flicking the shifter quickly into neutral the back to drive, or from a DEAD stop, giving lots of pedal. Even in the latter instance, 2nd isn't held very long.

How about the 1-2, or 3-4 upshifts? If those vary correctly based on throttle input, the Bowden cable may be ok, and you have other issues with the valve body which are only affecting the 2-3 shift. The Bowden cable position should affect all 3 upshifts in about the same way.

One last thing, If this kickdown 'cable' (isn't it actually a vac line?) doesn't hold vacuum, is it automatically bad or can one remove and repair the part?

Oh. I see now that it has both. But same question, can a vac leak inside the unit be repaired? Obviously one can replace the elbow.

The vacuum portion is only used for the cold upshift delay. Vacuum doesn't affect normal shifting operation. And it can't be fixed... if the vac diaphragm is bad, and you need the cold upshift delay functional, the whole thing has to be replaced - NOT fun, as seen at the link above.

 

[luckymike]

Remember the Bowden cable is a huge pain to replace - you don't want to do that unless you can confirm it's defective. First verify both ends are connected properly. If you disconnect the ball/socket at the throttle linkage and pull on it, you should feel resistance, and it should pull back when you release. If not, the cable may not be connected to the linkage rod inside the transmission.

Click here for a How-To on this job... enjoy!



How about the 1-2, or 3-4 upshifts? If those vary correctly based on throttle input, the Bowden cable may be ok, and you have other issues with the valve body which are only affecting the 2-3 shift. The Bowden cable position should affect all 3 upshifts in about the same way.



The vacuum portion is only used for the cold upshift delay. Vacuum doesn't affect normal shifting operation. And it can't be fixed... if the vac diaphragm is bad, and you need the cold upshift delay functional, the whole thing has to be replaced - NOT fun, as seen at the link above.

Wow, nice writeup on the Bowden cable. Have access to a lift but still hope I don't need it! I will test for vac and resistance soon.

What about the 1-2 upshift? It's non-existent. The trans shifts directly from 1 to 3. Every time, cold or hot, at about 23mph. Every time. I can only experience 2nd gear if I apply hefty throttle from a dead stop or, at some point before 23mph, if I flick the shifter briefly into neutral and back to drive. Then the trans goes from 1-2 and the 2-3 shift follows normally.

Actually, there is one other situation where the trans shifts from 1 to 2. There's a 90 degree turn near our house where the road goes from level to a very slight downgrade. If the light at the intersection is green and I can roll through the corner, the trans will shift from first to second, every time. It's the weirdest thing. And very repeatable.

Except for this odd 2nd gear miss, the trans shifts just fine.

 

[gsxr]

Do you mean that even at light throttle, fully warmed up, it won't shift out of 1st gear until 23mph (3000rpm)? That isn't normal either.

I'd almost think the kickdown is stuck on (energized), but if that happens I don't think it will ever upshift to 4th.

It sounds like something is screwy with your valvebody. I'd drop the pan and remove each of the VB side plates, and check for any broken springs... if nothing looks awry, I'd try swapping the entire VB.

What you describe is really bizarre.

 

[luckymike]

Yes, even at light throttle, no 2nd gear. Just straight to 3rd at 23mph. Again, unless it's one of the three situations where it will shift from 1-2.

Easy enough check springs one of these days.

And I'll check the vac for the upshift delay and the friction of the Bowden.

 

[gsxr]

Fully warmed up, it should be shifting out of 1st gear at ~1500rpm, with light throttle. Something is definitely screwy.

For grins, you can try temporarily disconnecting the Bowden cable at the back of the intake manifold (throttle linkage), and/or disconnecting the wire at the kickdown solenoid. If there is no change at light throttle, something is wrong internally - most likely with the valvebody.

I really need to get a video of my E420 instrument cluster showing upshifts at light, medium, and full throttle.

 

[luckymike]

Agreed, something is definitely wrong.

Disconnecting the Bowden cable is an easy test. I'll try to do that this weekend along with the other tests.

It would be great to see a video of your car showing upshifts.