M104 regeneration valve

[evo-number-one]

I have started to experience a problem with my 1994 E280 Estate (approx 90,000 miles) which I hope someone can kindly throw some light upon.

When starting the car from cold, the car settles and ticks over around 650/700rpm - possibly a little low - but still nice and smooth. As the car warms the regeneration valve pulses away. The car pulls and operates perfectly at all speeds.

Once the car is completely warm (reading just over 80 on the temp gauge) the car still pulls as it should, but seems to be not quite as smoothly and when finally slowing to approach a junction or at a standstill, the tick-over is now down in revs as if its about to stall (although it never does). Coinciding with this the regeneration valve no longer pulses. Should this continuously pulse?

When the regeneration valve is not pulsing, the tick-over speed never hunts, it just stays low - almost as if it is going to stall. Initially, I replaced the regeneration valve believing it to be faulty, but the new one has made no difference.

At all times, the car runs on 6 cylinders and is not misfiring (in the past I have experienced those symptoms with coil and harness problems – both of which have been sorted – and the current problems in no way have the feel or restrictions that I experienced with those issues).

Could somebody please describe how the regeneration valve works within the emission/evaporation system and explain to me what series of events cause it to pulse. If the charcoal filter (which although I have a stamped service book, I do not have any invoice/written evidence has been changed) is blocked and no longer efficient would this cause the valve to not operate?

Maybe I am going up the wrong path with the valve/filter - but the car runs perfectly when its pulsing.

If somebody else has experienced the above and cured this issue I would greatly appreciate some guidance as to the cure.

Regards

Bob

 

[grober]

In my experience Mercedes dealers rarely change the charcoal filter altho they are supposed to at around 72k miles. There are 3 connections to the filter which is normally found behind the nearside front wheel arch liner. One connection goes to the petrol tank such that any petrol fumes from the tank are trapped on the activated charcoal. One is a vent to air and the third is connected to the inlet manifold via the regeneration valve. This valve is controlled by the engine ECU. If the following conditions are met
coolant temp >70c
2 minutes elapsed after start
engine not deaccelerating.( throttle plate position switch??)

then the valve is pulsed open to evacuate the charcoal filter of fumes.
I don't recall ever hearing that valve operating but then I have never listened for it. There could be lots of reasons why the engine is not running normally which have nothing to do with the valve. However it would be worth checking all the flexible pipe connectors and pipes leading to and from it and the charcoal filter for leaks since this might effect the mixture strength. First test is remove the(black??) plastic pipe to the regeneration valve from the inlet manifold and block it off. If there are any leaks upstream of this then the fault should disappear. If not the fault lies elsewhere.

 

[evo-number-one]

Good Man Grober! Just what I was after - how the valve is activated

You can't actually hear the valve but when it is active you can physically feel it pulsing.

I will start with pipe-work investigation as suggested.


Bob

 

[Vlad]

I've had similar problems with a M104 engine and this is what I did.

1) Cleaned throttle body
2) Replaced OVP Relay
3) Changed Charcoal Filter
4) Replaced Reg Gen valve.

 

[kth286]

evo

You may have read my post on other forum similar to grober as we read the same Merc manuals.

However as I state you are going about this incorrectly without reading the codes first.

Then take action according to any codes that may be set.

Based ? I can read the codes for you in Reading area.

 

[evo-number-one]

Over the past month the issue appeared to become much less frequent, but at the end end of the week the problem reared its head again and it now appears to be more frequent than ever.

Just to recap, the problem is that the car occassionally idles quite roughly. It never hunts up and down, nor stalls - it just ticks over at 550/600 rpm. It sounds and feels quite lumpy, but more as if its out of tune or fuel related than its not running on all 6 like I experienced when I had the loom prob (which I think Waterhouse mercedes put in the coil repair sections - not 100% as it was sorted by the Seller, but I know they definately sorted it and the coil connctions are like new).

More often than not, if I pull upto a junction and it happens, when I pull away it seems fine and at the next juntion it ticks over low, but is not rough/lumpy. Last couple of days though the probs been been pretty constant whenever at idle.

Now, armed with my blink tester I am recording just the one error - code 17 on pin 8, which I believe is injector / cylinder 4. I have cleared it, used the car and its run fine - checked it again and its not showing.

However, next time I use the car (hot or cold) it idles roughly, re-check the code and its the same - code 17 on pin 8.

Has anyone had experience of this code showing up on this engine and can point me in the next (right) direction?

The car only gets used at weekends, but was hoping to use her for our summer trip to the West Country, but looks like we are going to have to play sardines in the Focus :-(

Thanks for any help.

E-N-O

 

[w124coupe]

Hi, from your posts it seems your harness was mended, not replaced?

You may therefore have problems with the harness to the injectors as the "repair kit" just sorts out the wiring to the coil packs.....this part is cooked the most so dies first but the rest will not be far behind.

Unfortunately if you have harness issues, working on the car tends to shatter the dried out insulation and you end up chasing endless expensive faults.

I would check the harness first by cutting along the insulation and visually inspecting the insulation on the individual wires inside.

If its gone, either replace the loom (about £600 and not a difficult DIY) or get rid of the car.

Assuming the loom looks OK, you could swap two injectors over and then see if the fault code 17 moves (to determine if its the injector at fault).

Or maybe as a long shot try some injector cleaner through the next tank.

 

[Dieselman]

Check the wiring loom then swap the injectors over and see if the fault remains on the original one or moves.

 

[evo-number-one]

Dieselman / W124coupe

Thanks for the advice guys.

Probably won't get a chance before the end of the week, but I will summon up the courage (and the scalpel) and report back - hopefully that its the injector itself!

I can but hope......

 

[evo-number-one]

Loom of Doom strikes again!!

Update: My local garage (who service the car) put the loom to the knife today and found not a lot........ was left!!!

So, they have taken out all of the bad wiring (and beyond) and replaced with new and the car at last again runs exactly as she should.
:bannana:

Obviously only got her back this evening, but they appear to have done a great job - the rapidly approaching Family trip to Cornwall will be the test - but I am going to treat this as a definite temporary measure and get the loom replaced in its entirety - but for now their work relieves any immediate stress and I can budget for it being done properly at service time (with less than 90k on the clock, my plans are to keep this car for the long term, so this will be sorted).

Just one quick question - has anyone any experience or have heard of the MB replacement looms playing up? - This is to gauge whether to go original replacement or get the auto electrician to replicate. Obviously paying for an experts time is not going to be cheap, but quite feasibly cheaper than the £568 plus vat that MB want for the original part.

I would post up a photo of the state of the wiring that was removed and replaced, but its x-rated stuff and there maybe readers out there that are
squeamish and easily shocked.

Thanks for everyone’s help and advice from the forum on this.

 

[Vlad]

Replacement looms are made from a different insulation.

Re the state of the loom, I have images of bare wires, powdered insulation etc etc LOL!....well at least you can thank your lucky stars that you didn't fry your ECU and coil packs in the process

 

[kth286]

evo-number -one

To me the solution does not match the code 17 fault.

Lots of questions to ask if you want to know what really happened, and therefore increase your knowledge base for future.

Ask them how their solution related to the code 17 fault ????????????

You seem to have rolled over too easily - much cost to you ?????????

 

[evo-number-one]

Sorry, forgot to mention that when the car went in, it was driving even rougher - full time! I read the codes again before i delivered her and code 17 had been joined by code 18 (injector/cylinder 5).

Initially my local garage concentrated on injectors 4 and 5 and after could not pulse to the ecu on the former and only on the latter if they fiddled with the loom. On investigation they found damage to each injectors element of the loom and with an initial 'temp' fix on each proved that the car ran correctly and the codes did not return.

This to me does fit with what I was dreading regarding such an injector/cylinder error code(s), if it was not the injectors themselves and not the fuel supply. Obviously there is a bit more detail now, but what was it from my previous post that made you think that the loom may not be responsible for such a fault ? As you say - worth any info to aid my knowledge of these things.

They then delved further and found more loom deterioration (not as bad as initially on 4 & 5 - but also more damage to those 2 too further back) so as further precaution against the issue arising on any of the others, stripped back for all injectors well beyond the damaged areas, replacing anything they could find that appeared to be either already amiss or with apparent weakness looming (no loom related pun intended).

I don't consider to have "rolled-over too easily" on the above and thought that the price of a couple of hours labour (less than 50/60% of a local MB Dealer rates) plus the cost of the wiring/insulation was very reasonable - especially as I know they spent a bit longer than this on my car what with investigation and then the repair.

Hopefully, this has cracked it (doh! -no loom related pun intended here either) but I am aware there have been plenty of other posts where the whole loom has been replaced (by owners themselves, indi, local and Dealers) and either not cured the original problem (at this moment in time it has fixed mine) or at some time later the Owners original issue has sadly returned -hence my question about the replacement loom.

Because its been so long since it ran consistently smoothly, I am sure I will spend the next xxx miles driving with fingers crossed still and a slight anticipation that the rough idle will return - but time will tell!

Thanks again.

 

[kth286]

evo-number-one

OK you have supplied more info.

BUT the real problem with the looms is under the top metal engine cover where the coils sit under.

It is the insulation of the coil feeds that suffer and short out.

The rest of the loom may go a bit stiff but does not cause a problem with running issues.

The reason for the problem in the coils area is that the harness has to be moved every time a plug change is required every 18,000 miles.
This entails the coil feed wires being bent back on themselves as they are normally left connected to the coils when the coils are lifted up and moved back out of the way to access the plugs underneath down in the holes.

So, at 90000 miles this will/should have happened 5 times, and you end up with totally cracked wire insulation. When you see how tight the wires run next to each other you can see why the current finds it easy to jump from one wire to another on engine start up, via the cracked and fallen off insulation. It happens under the feed wires outer loose covering and is not seen until the outer covering is snipped away with small scissors. So, an inexperienced tech will not even know he is causing damage when doing a plug change as he cannot see it happening.

Conversley the injector feeds do not get moved at all. AND they are not enclosed in a very hot environment. They are on the cold side of the engine away from the exhaust side.

I would suggest that your garage is not a specialist Mercedes expert and that by fiddling previously with faults that they did not understand had perhaps left the injector plugs not fully plugged in and was causing an occasional open cicuit.

 

[evo-number-one]

Yep, know about the coily bits - been there done that.

So these looms never go at any other point than in connection to the coils?

Local mechs must have slipped around while I was asleep and wiggled No 5 off too!

Priceless