190E 2.5 Starting Problems - HELP!

[Alex Crow]

yes, compressions are quite good for a car of this mileage.

i very much doubt carboned inlet valves, not usual to see this on these engines. also there is more than an edge missing here, i estimate output at around 100bhp, as opposed to around 170bhp (it is an old engine after all!).

 

[meanie]

yes, compressions are quite good for a car of this mileage.

i very much doubt carboned inlet valves, not usual to see this on these engines. also there is more than an edge missing here, i estimate output at around 100bhp, as opposed to around 170bhp (it is an old engine after all!).

I'd agree that it's not usual on these, but as it's a leggy auto, for the time involved it may be worth half an hour with an endoscope stuck through the injector ports.
As you obviously know your stuff, we can take fuel pressures, senders etc for granted. Good compression would seem to rule out valve clearance and/or chain stretch issues.
I mention the carboned inlets because when I first knew my 24v 300 several years ago, its then owner moaned about a lack of power. I can't really tell you on here how I cleaned them, but if you still have some of the SI bulletins from 20 odd years ago (!!) you'll find the answer!

 

[television]

Timing chain wear and timing out ??

 

[Alex Crow]

thanks malcolm, checked that one staright away as the cosworths had a known fault with jumping teeth. as this is the 2.5 model and has a twin row chain this problem is not typical, but checked anyway to be sure. timing absolutely spot on and all valve gear and drive in very good condition, even distributor gear.

 

[television]

Exhaust blocked they have CATs

 

[Alex Crow]

well done malcolm. this is our conclusion, although it does not have cats. the system is genuine and not all that old - there are some dents in the front pipes they are not bad enough to cause such power loss. if it had been the original one, or if it did have cats we would have removed it for inspection with the endoscope. as it was it was deemed unlikely, so we concentrated on the 432 other things that could cause it, but we are now suspecting the exhaust after a process of elimination.

 

[television]

Just curious,,is it a genuine exhaust

 

[Alex Crow]

yes, it is genuine and not old. not that i have anything against the youthfully challenged, malcolm

 

[television]

You will get old one day , hopefully I do not do now, what I did back then,

It is odd years ago I have run a business, built a house, produced kids, fixed cars, all at the same time.

Just pushing the keyboard keys can get hard enough these days.

 

[W.Roger Williams]

i have been out today doing 50 miles. It seems the pedal has to go a long way down before the throttle starts to have any effect. also theres a distinct take up of slack (clunk!)when the car changes ratios,both up and down the box.
What would be useful is to have all the data gathered last week and see how it compares with M-B works settings and recommendations.
i look forward to your thoughts, and thanks for your response. Roger Williams

 

[Number_Cruncher]

As above, checking for an exhaust blockage would be a good place to look - it might be possible to diagnose with a simple vacuum gauge check on the inlet manifold.

You may be getting a clunk when changing gear because the throttle is being held wider open than usual, hence less vacuum. Phrased another way, fix the low power problem first, and there's a very good chance the clunking will go away as well.

 

[Alex Crow]

What would be useful is to have all the data gathered last week and see how it compares with M-B works settings and recommendations.

that is a big ask roger, most measured values were within MB specifications, although compressions were just below par.

please note, it is 11pm and all this is from memory...

ignition timing 28deg at 4,00rpm with vacuum - this is correct.
valve timing exactly right as per markings on camshafts and crankshaft.
main system fuel pressure 5.8 bar - correct.
lower chamber pressure varies with the EHA but measures around 5.25 to 5.3 bar, which is correct.
exhaust gas CO2, CO, HC and O2 values are within the given range, although HC was towards the top of the range but well within reasonable limits for a car of such high mileage.

the above are just the most obvious things to check, but here is a partial list of the other checks and tests done.

sensor checks of: 2x air temp sensors (the cosworth has one for the ignition too which is basically just a switch which retards the ignition slightly at high air temps), 2x coolant temp sensors, throttle plate potentiometer, rpm signal to ecu, ezl compensating reference value, gearbox protection switch, crank sensor (although this and certain other items were never suspected as the symptoms were not indicative of ignition faults - although ignition timing and timing compensating elements were scrutinised), resistance of EHA, throttle switches (2x for idle, 1x for full throttle). some other sensors too, but it is late!

fuel pressure of main and lower chambers was monitored on road tests, and a substitute EHA of the same part number was tried (these are the same as fitted to the 103 engines oddly, not the 102). it could be seen from lower chamber pressure that the EHA compensation was functioning correctly, also deceleration shut off was noted as correct. it was even tried without the ecu connected, the ke-jetronic will invariably perform well with the ecu disconnected as it is essentially a mechanical system with some electronic control - no change.

various pin outs were checked at the ezl and ke modules, all sensor checks were done at the modules as per good practice. all power supplies, earths, signals etc etc were correct.

there were obvious mechanical checks done too, such as making sure the throttle was opening fully , checking the air filter, checking the air flap travel and feel (should have smooth travel with equal resistance all the way), a visual inspection of the exhaust as already mentioned, oil/coolant level/condition etc are always checked at the start.

as we have seen faults introduced to vehicles by occasionally misguided previous owners we checked the part numbers of some parts against the parts catalogue, these included the ke module, the ignition module, the EHA, the metering head, the fuel pump relay. all are correct.

the ignition system did not come in for great scrutiny, other than the various timing checks (at idle/4,000 rpm and with/without vacuum), and the condition of the plugs which were a very reasonable light biscuity colour and all the same. due to the obvious lack of any misfiring we ruled out ht breakdown as a cause. gearbox protection switch was a suspect at one point as it can retard the ignition, but this was checked on road test and was fine.

the OVP and fuel pump relay came in for no great scrutiny either as both were found to be working throughout testing. equally we know from experience that injector faults are no capable of causing this odd fault and so the injectors were not tested - correct testing of the injectors is time consuming and their symptoms, if faulty, would be more pronounced at idle speed.

it is not possible for us to test the metering head unit, if i suspected this to be likely i would recommend sending off for testing - it is however still an official suspect, although the symptoms just don't fit and pressure tests suggest the membrane is not split and the flap test shows the pin is free from resistance (other than fuel pressure).

i am sure i have missed a few things out here, so allow that i may add a few to this list at a later date! we like to think that our customer brought his car to the right place as we are very familiar with these cars, and would normally expect to isolate these faults much sooner! naturally this problem has caused us much thought and research beyond the time allocated on my job card, i expect i have spent at least 10hrs on it one way or the other.

where to go next? testing the vehicle on a dynamometer would be a good move, it may not pinpoint the fault directly, but knowing the dynamic mixture values and ignition timing would be useful! (although ignition timing has been adequately tested, a double check under load would be reassuring).

note that this car has no lambda sensors, no knock control, no sequential fuel injectors, no variable valve timing and comparatively simple control systems with limited diagnostic facilities (there is a diagnostic output from the ke ecu which gives a duty cycle value from pin 3 in the round port on the l/h inner wing, this was checked and the output value was 50%, which indicates no faults present).

ok, there you go! being the type of person i am i would love to know what is wrong with this one, it is always hard when one beats you like this. so to finish off for now, main suspect is a damaged silencer causing excessive back pressure and consequent running faults - this is impossible to test without replacement, but could be removed and disassembled and checked with the endoscope.

 

[Number_Cruncher]

AP - put your endoscope away!, and consider a simple vac gauge test for exhaust blockage;

http://www.autospeed.com/cms/A_2393/article.html

I know it's dreadfully old fashioned these days, but, it's not a long duration, expensive, or difficult check to make!

 

[television]

Stick a vacuum cleaner on the end

 

[Alex Crow]

thanks number cruncher and malcolm,

the thing is, the car is only down on performance at the higher output range. if the exhaust was blocked enough to show up on idling and throttle blipping tests WITHOUT any load, it would be flat as a pancake. i estimate it is down by perhaps 40%. if we could get it on a steady state dyno and do measurements under load that would be very useful.

 

[Number_Cruncher]

>>i estimate it is down by perhaps 40%

For an exhaust to be giving such a serious drop in power, it has to be seriously blocked*.

I would do the vac gauge test, and when snapping the throttle open and then shutting it at high engine speed, note the maximum vacuum reached.

Then, drop the exhaust off at the manifold to front pipe connection, and repeat the test. If after releasing the throttle, the vacuum peaks at a higher level, you've got a firm diagnosis.

* Blocking the exhaust is a very poor way of regulating engine output, I've done some tests on this using a Ford Zetec engine on an engine dynamometer, and exhaust restrictions are less effective than restricting the inlet by about the same ratio as the compression ratio of the engine!, i.e., beyond the obvious issue of choking up and sticking, there's a good reason why petrol engines are throttled in the intake and not the exhaust!

 

[Alex Crow]

for sure mr cruncher that is a great suggestion, and if we get the car back i will do it - as you say it will not take long.

i understand what you are saying about exhaust restrictions not having the same effect as intake restrictions, valve sizes demonstrate this to some degree with exhaust valves having less time area but flowing much greater volumes of gas. i am hoping mr williams wil be able to get the car back to us for further checks, but still doubt that this test could be conclusive without visual inspection. naturally if we had another 2.5 in at the same time we could do side by side tests, with probably definitive conclusions, but sadly this is not possible. will keep you posted.

 

[Will_uk]

Just seen this thread and had a quick read.

Without going through it all again, what were the readings from the compression check?

Has the car ever had a head overhaul - when were the valve clearances last adjusted?

If it's never been apart at 160k miles it's well due an overhaul.

Will

 

[Alex Crow]

ok, we had the car back today. having read number cruncher's vacuum testing link, i thought i ought to do a check of manifold vacuum in the following ways, all with warm engine. tests done with mityvac vacuum pump/gauge.

idling...........around 58 cm/hg with very slight fluctuations.
blipping throttle (instant full throttle) to 6,000 rpm - vacuum initially drops to very near zero, then goes up to 72 cm/hg when throttle closed, then returns to 58 cm/hg when it reaches idle. i also tried the test of raising the speed to 5,000 rpm and holding it there. the vacuum initially drops (the amount being quite sensitive to the throttle opening) before quickly returning to around 57/58 cm/hg at the held 5,000 rpm.

next i dropped the entire exhaust, and tried the tests again - although i had to be quick as the hot exhaust gasses shouldn't be allowed to exit in the lower engine bay for any longer than absolutely necessary.

the results were EXACTLY the same, no noticeable difference, so either the exhaust was ok, or the test was not going to be conclusive.

HOWEVER, on dropping the exhaust i had instantly noticed the inner skin of the front downpipe was extremely deformed and blocking over 90% of the cross section of the downpipe!!! colour me happy . note that the entire exhaust was relatively new, the oldest date being 12M06 (december 2006), it was all genuine, rust free and in very good condition - at least on the outside. we had already done a visual external inspection fairly early on in the diagnosis, and had no reason to suspect a problem like this, other than through a protracted process of elimination. also note that the spark plugs all showed similar colour.

shame i did not have my camera to photo it for the thread, but will take a picture on monday if we still have the section with us (the boss is quite keen to take it to mr williams to show him!). we obviously cannot report on the effect of fitting a new section yet, it has to come from germany. seeing is believing, but i could hardly believe the state of this genuine part, which seems to have been faulty from new! will keep you all posted, but am feeling vindicated for taking a logical and reasoned approach, and being sure in the many checks and tests i had made.

 

[Number_Cruncher]

Well found AP; a testament to your patience and tenacity, I hope the new exhaust fixes the fault.

I'm a bit surprised the vac gauge didn't show anything, and I'm sorry if my suggestion to use one wasn't helpful.

As an aside, I've had quite a bit of success with this type of old fashioned diagnosis. Early 1990's Vauxhalls had lots of trouble with sticking valves, and, the quickest way to see this was with a vac gauge - the needle would be bouncing all over the place at tick over!