W124 E280 loses power on hills

[gasgas4]

I purchased a nice shinny 1996 W124 E280 estate last week. Car is great apart from one major issue, it loses power going up hills, or at least that’s how the problem first presented itself. After a week of driving around and various checks etc I have learnt the following:

If I drive carefully (less 2500rpm) all is OK, I can drive around for ages. It also seems to be ok on the flat. But if I push it going up a hill, all of a sudden it just goes flat (not very convenient on the motorway whilst overtaking). If you keep your foot down or try to accelerate it will just slow down but if you back off it seems to recover. Knowing this, I drove around for a few days without any problem. Anyway, knowing that this wasn’t going to solve anything I started to see if I could recreate the same problem on the flat by holding higher revs in 3rd gear. I couldn’t, it seemed happy to be driven on the flat at around 3500-4000rpm without any problem. So I found a nice quiet road and floored it several times, holding it to max revs for as long as possible. This was interesting as the first thing that I noticed was that there is some hesitation between 3500-5000rpm (old car came on much stronger in the mid range) and the second was that after it made the initial problem much worse. Now the car would slow up just going over a motorway bridge. I suspect that if I drive slowly again as previous the symptoms will slowly subside until I floor it again.

I would describe the symptoms as fuel starvation or a weak or faulty fuel pump but the car had a new pump/filter 6 months ago. So far I have swapped the MAF (from my other E280, same model/age etc with damaged wiring loom/ECU failure, will be used as donor for new car in the short term) but it made no difference. Fitted new plugs, checked wiring loom, coils, leads etc… Someone at work suggested that it could be a partially blocked cat (bits dislodging/moving around) but I’m not convinced.

Out of interest I built a blink tester and the fault it comes up with is code 9 “O2S 1 (before TWC) - voltage too high, circuit open or voltage implausible”. I’m not really familiar with the terminology but I think this is saying that the lambda sensor is faulty or open circuit. I thought that there was only one sensor and it’s fitted to the cat. Does anybody know if this is correct and if yes whether it would cause the above problems?

Another small observation is that the gear change between 2 & 3 (hard acceleration) seemed very slow.

Other than the above the car is faultless it ticks over like dream, the fuel consumption is normal, starts first time, temp normal.

If anyone has any thoughts or suggestions please post a reply?

 

[neilrr]

Welcome to the forum.

It sounds a bit like the problem I had recently with my 190E 2.6. See the first post - the problem would have been completely solved by changing the fuel filter, IF I'd remembered to adjust the EZL back to where it originally was.

http://www.mbclub.co.uk/forums/engine/76479-190e-2-6-has-become-gutless-less-willing-rev.html

 

[andymerc]

sounds like an MAF senor is the donor one 100% good

 

[Dave Lewis]

The blink code and symptoms point to the the O2 sensor being kaput. Try substituting the one from your other car, it's on top of the exhaust just before the cat, and plugs in under the driver's footwell carpet. Also, when were the spark plugs replaced, and are they the right ones?

 

[mercman15]

i had that problem sounds strange but worked for me

i blew fuel line back with air line .

worth a try

 

[kth286]

It sounds as if it may be going into limp home mode.

the fault code showing is correct for lambda problems - either not switching properly or the lamda heater broken.

Are you sure there are no other codes ?

Why not clear this code, then drive, and check again to see what code/s come back.

It could be that was an old code and was not cleared previously..

 

[w124coupe]

The O2 sensor heater throws code 11 if its broken - code 9 is O2 sensor voltage too high/implausible/open circuit (TWC in the descriptions means "three way catalyst" by the way). A digital code reader would access the three codes "under " the code 9 and tell you which of the 3 faults is being seen.

kth286's suggestion to clear the code and then read ALL codes is the right first step.

(press 2-4 secs and then read a code and repeat the press/read until you get back to reading the first code)
(clear codes one after the other by pressing for 6-8 secs after each code)

If 9 comes back you have nothing to lose (except possibly some knuckle skin) by taking the donor's O2 sensor (and checking the wiring to it is OK at the same time). The connector is under the NSF carpet and should be pristine assuming no water has leaked in...

Make sure you check codes and complete any further diagnosis quickly after the new sensor goes in or a fuelling problem might damage the new one as well...

With code 9 set the car will run on a default map (as it can't use the O2 sensor to confirm fuel/air mixture) and set limp mode which cuts fuelling at about 4000rpm - effectively moving the rev limiter downwards. This assumes the default map is close enough (given wear and tear and possible other issues) to give a usable mixture!

(slow 2->3 will be the ECU using the torque limiter plan used to blur all changes but without the O2 sensor control this will err on the safe side to avoid gearbox damage)

If you have a reasonable meter, you can access the connector on the floorpan and check if, when hot, the O2 sensor reading "flips" between about 0.2v and 0.8v at about 1 cycle per sec - if it doesn't then its bad OR there is a severe fuelling fault meaning the ECU can't influence fuelling to keep the O2 sensor cycling.

 

[gasgas4]

Took kth286's & w124coupe's suggestion and cleared all codes, went for a drive and checked codes again and got the following:

Code 9 & 11 which confirms the O2 sensor again plus code 40 which I believe is "Transmission overload protection switch - open/short to ground or B+ or open or closed or implausible)".

So, based on the confirmation of codes 9 & 11 I'll change the O2 sensor.

Not sure exactly where the transmission overload switch is or what it means but I guess that it may explain the slow change between 2nd & 3rd. Any thoughts?

 

[grober]

The transmission overload protection switch is vacuum operated and a large air leak might cause implausible readings on the lambda sensor. Might be worth checking all the engine manifold vacuum lines and flexible rubber pipework are OK?

 

[w124coupe]

I would swap the O2 sensor, clear all codes AND adaption settings, go on a few trips to let the ECU adapt and then see if you still get the transmission overload reported before spending time/effort on the valve (no problem checking the vac is connected while you are "underneath" though!).

(code 11 means the O2 sensor heater is dead so you will waste lots of fuel each cold start waiting for the O2 sensor to heat up via exhaust gasses and go closed loop - best to change it).

With O2 sensor out of action, ECU will use default fuelling map and also less aggressive ignition timing. If you have wear and tear etc the default map will be a fair way out and so will affect the ECU being able to moderate torque correctly during the gearchange.

Reset adaption (ignition on, engine not running):
Clear codes until single flash obtained
Press 6-8 seconds
Ignition off min 5 seconds
Ignition on min 10 seconds
Start engine



This will make sure you start from default fuelling instead of where the ECU ended up while the O2 sensor was getting old.....

 

[gasgas4]

OK so took w124coupe's advice and changed O2 sensor for a new one, cleared all codes & reset adaption. So far I've spent couple of days driving around including motorway hills at 80mph etc and all seemed to ok; in fact it was a big improvement (brilliant advice thankyou). However this afternoon had a small hiccup whilst going up a long hill on a local A road. The wife was driving so it's difficult to say exactly what happened but we had to slow down due to traffic, when she speed up it kicked down in to 3rd, then backed off the throttle so it changed back into 4th, Then on acceleration the car seem to labour for bit (not kicking down) before caring on as normal. Hopefully it's the wife's driving (too soft on the accelerator) but when we got back home I got the blink tester out again and got the following:

Code 40 Transmission overload protection switch - open/short to ground or B+ or open or closed or implausible (same as before which I expected)

Code 21 Ignition output 3 or ignition coil for cylinder 1 and 6.

I assume that this could mean anything from a dodgy lead/plug or coil pack to wiring or the ecu. I’m beginning to wonder whether I have an intermittent misfire on one cylinder which could explain the above hiccup & perhaps some previous behaviour. I know what two cylinders feels like as this renders the car pretty much un-drivable (experience from old E280). Would you get code 21 for a faulty lead or plug, or is it only likely to be coil pack or worse?

In the mean time I’ll keep driving to see what happens.

 

[w124coupe]

The ECU monitors the way voltage decays on the coil supply as a way to determine if a spark happened. Its possible you have a duff plug or a duff plug "boot" under the coil itself.

To protect the TWC (cat), the ECU turns off fuel to the misfiring cylinder(s) to prevent raw fuel overheating the TWC.

If it was me I would:

Clear the code 21 and swap (just) the coil with one of the other 2 and then repeat the heavy load (hillclimb) and see what happens.

If the fault re-occurs and moves then coil is the suspect, if it re-occurs and is still 1 or 6 then start with the boot.

Are the plugs confirmed as OK and definitely the right temperature spec? (I know I go on about this but misfires happen if the plugs are the wrong temp spec as they either run cool and foul or become hot and pre-ignite).

The plugs may have taken a hammering from incorrect fuelling owing to the lack of O2 sensor control over fuelling....

A look at the ECU-to-coil LT wiring (where it passes through the groove to the side of the coils) to check the state of the inner insulation would also be a good idea as, if its degraded, you might blow one or more of the ECU output stages.

 

[gasgas4]

OK many thanks for info. I'll do as you suggest.

Out of inerest are the boots main dealer only or can you get them from eurocarparts? Also were is the transmission overload protection switch located?

 

[w124coupe]

just off out but - boots dealer only (cheap). Switch is on the transmission - you probably saw it when you did the O2 sensor (and hopefully didn't disturb it )

Just out of interest, if you get the codes read digitally (rather than just the blink codes), there are several "sub codes" under 21 which will tell you if its:

061 cyl 1 misfire
062 cyl 6 misfire
063 coil current value fault

Similar re the overload error 40 which has digital codes 093-096 inclusive to further determine the error. (I would check the wiring etc is undamaged as its in a vulnerable place down there...)

 

[gasgas4]

The latest news is that since swapping the coils everything has been ok with the exception of the code 40 transmission overload protection switch. I guess it's a matter of time though. In the mean time took a brief look at the transmission overload protection switch when I moved the rubber boot a bit of oil leaked out, so I guess the diaphragm within the switch is damaged. I didn't disturb it for fear of losing more oil.

Does anyone know where to get a replacement or whether it's a main dear only part and how much it will be? Also is it easy to change, are there any complications?

 

[gasgas4]

Well thanks to all of the fantastic advice I have now managed to get to the stage where there are no fault codes & I can drive the car pretty much as normal. There is however one underlying problem and that is I still get some hesitation between 3500-5000rpm. I would still say that this car still doesn't have the torque of the old car and I'm sure that this problem is linked to the lower performance. I was wondering about the VVT, could this be a potential problem. I have swapped the MAF from the old car and this certainly improved things but it didn't get rid of the hesitation. It will rev right though max RPM.

Can anyone offer any advice?

 

[grober]

Your car has variable inlet resonance length by the use of flaps. It sounds as if these aren't working as they should. They are designed to increase torque together with the varaible inlet camshaft timing.
The variable inlet length resonance device is controlled via 2 ECU controlled solenoid activated flaps. Different combinations of these gives 3 different effective lengths of inlet manifold depending on engine revs. This helps inlet gas flow which markedly improves torque and driveability across the rev range.

p.s. if its an estate there could be 2 fuel pumps

 

[gasgas4]

Wouldn't this flag up a fault code. I notice that 38 is the Adjustable camshaft timing solenoid - open/short to ground or B+ and that 46 is the Resonance intake manifold switchover valve - open/short to ground or B+.

Anyway I guess I can swap the cam solenoid but I don’t know where to start with the variable inlet length resonance device. Where is it?

 

[grober]

The variable resonance manifold is normally controlled by 1/2 vacuum operated actuators. 12/4 and 12/5 These are actuated by 1/2 solenoid operated valves Y22/6 and Y22/7. If there is a vacuum problem/leak or the actuator diaphrams are gone or the flaps are broken/jammed internally the electrical solenoid valves may operate correctly and thus no error codes will be flagged since its only the electrical operation of the system that is monitored? Your best bet is to look for a couple of vacuum actuators on the inlet manifold and trace back the vacuum lines to the 2 valves. I will try to find some pics. This is not well documented.

EDIT:-
This is for the diesel but I suspect the petrol will be similar
http://ww2.justanswer.com/uploads/eurotec/2009-11-16_203013_Variable_intake_system_function.pdf

 

[jonnyboy]

Doesnt the whole vacuum thing tie in nicely with the gearbox, too ?