The complete automotive resource for buyers, sellers, and owners like you.
Car, Truck and SUV Forums at Roadfly
+ Bentley Forums
+ BMW Forums
+ Cadillac Forums
+ Chevrolet Forums
+ Ferrari Forums
+ Jaguar Forums
+ Lamborghini Forums
+ Lotus Forums
+ Mercedes-Benz Forums
+ Maserati Forums
+ MINI Forums
+ Porsche Forums
+ General Discussion
+ Marketplace Forums
Test your diagnoatic skills (long) (archive)

[ Follow Ups ] [ 5-series (E28) Message Board ] [ Msg. Board FAQ ]

Posted by John Millward on June 18, 1999 at 08:27:28:

This is the history of a problem I experienced but have now solved. I offer it in case other people are having the same problem, but also as encouragement to keep on searching. You may like to jump to the end when you think you have a diagnosis.

The car is a 1985 525e, yes I am in the UK. Is this the same as a 528e in the US ? anyway it has an B27 engine with the following symptoms. It would start OK but then almost immediately start to misfire (over rich ?). Revving the engine would clear the misfire but the neighbours were beginning to complain about the blue / black smoke. (The car has done 156K). I might expect a little blue smoke on start up after this mileage but this was ridiculous. Once warm the car didnít run too badly but felt, smelt and drank
over-rich (approx. 20 mpg on motorway). Even when warm evving would give clouds of black smoke. Plugs and exhaust pipe all confirmed over rich. Idle speed was also difficult to get high enough when cold without being too fast when hot.

I concluded that I had two problems, 1) the car produced an alarming amount of smoke when revved and 2) the fuel consumption was too high. These problems may or may not have been connected. Some of the following ideas are
clearly aimed at only one of these problems.

I had recently bought the car (cheaply from a dodgy dealer) and it was in pretty good condition (leather seats, air-con - fairly rare in UK). I got some history and a few bills which showed that exhaust, centre steering tie bar, fuel filter and fuel pump had all been replaced recently. The pump was quite noisy but, as it was new, I assumed this must be normal.

Where would you start ? I took off and cleaned the crank-case breathers - even took off the inlet manifold to check the steel breather pipe wasn't full of muck. (swine of a job because it would not slide down into the block like
its supposed to.) All looked pretty good - no real improvement.

I checked the idle mixture using a colour-tune and was unable to get the mixture right even with the by-pass fully open. Suspicion fell on the air-flow-meter, temperature sender, throttle switch and fuel pressure regulator. Tried to clean out the AFM by-pass, no effect. In the UK we use
'Haynes' manuals - useful but not always 100% accurate. This gave figures for AFM and temperature probe (the one that feeds the Motronic) resistance.

The AFM resistance was bizarre (See another FAQ by me on this subject) anyway I bought a new one - better but still not right. (I still defy anyone to explain the AFM resistance characteristic - I have a degree in aerodynamics and I used to design engine management systems - either my meter is broken, my brain has seized up with age or the
relationship between volumetric flow and flap angle is horribly non-linear. Even if it is the latter I fail to see why Bosch don't do the linearisation inside the Motronic processor - after all that's what digital processors do
best. Then again Bosch are still going strong and the company I used to work for has been sold off .... Hmmm...)

Back to the plot. I thought I had found the answer when I measured the temperature sensor resistance - open circuit. Turns out I was measuring the temperature switch (the one used in the idle speed control circuit - my car has a vacuum operated idle speed control valve). When I measured the right sender it had a resistance of about 2K cold. The Haynes manual says it should be a few hundred ohms cold and almost zero hot, so I tried running with it shorted out. Maybe slightly better hot but the car wouldn't start
from cold. I also noticed that the sender looked new - had somebody replaced it with the wrong one ? I managed to get hold of the equivalent sensor from 528i and found it had very similar characteristics -my suspicions grew. I eventually persuaded the local BMW dealer to let me measure what he thought was the right sensor and, yes it had the same resistance as mine. So I concluded that the manual was wrong.

I checked the throttle switches were working correctly (closed and wide open throttle). The car is a non-cat model and therefore will be set up to richen the mixture at WOT, thus if it thinks WOT is always true it will run rich all
the time. I am not that familiar with US spec cars but I believe that WOT also causes cat cars to run slightly rich (just as they do when cold). Obviously the cat doesnít work properly under these conditions but I don't think it damages the cat so they do it because you get a slight power increase and WOT is not included in the type certification test (it certainly didn't used to be, maybe CARB have spotted this by now !).
Anyway all was well here.

I disconnected the return from the fuel pressure regulator and connected it to a fuel tank. Sure enough, when I ran the pump fuel was being returned to the tank. I even blew down the return line to check it wasn't blocked. Obviously if the injector pressure is too high, too much fuel will be
injected for a given injector pulse width (which is what the Motronic controls). I also pulled out the injector rail and injectors then pressurised the system to check that none of the injectors were leaking. This would have explained the high fuel consumption and possibly why there was so much fuel in the inlet manifold when starting from cold. However
there were no signs of leakage (Actually good news considering the price of injectors)

I also checked for fuel pipe leaks (obvious but worth checking) and air leaks into the inlet manifold. My car has a number of vacuum connections and some of these did look loose - particularly the brake servo pipe. Where there was enough slack I cut 1 cm off the ends of the pipes so that the stretched bit was removed. I also did this trick with some of the fuel lines. I'm sure this did make a slight improvement but the major problem was still there.

By this time I was starting to despair, was it really very worn valve guides / piston rings (the smoke would therefore be oil and the fuel consumption just normal for the car). I checked the compression (should have done this first but I hate bad news). I was surprised just how much compression I got - almost off the gauge with a cold engine. All cylinders were reasonably close.

The Answer

Measuring the compression encouraged me to check another pressure - yes the fuel pressure. Although I had checked that fuel was being returned to tank I hadn't checked the actual pressure, largely because I didn't have the
equipment. Anyway I made up some pipes and used the tyre pressure gauge from my foot pump (see last paragraph).
When I ran the pump I was surprised to see the gauge needle hit the endstop - over 100 psi ! I didn't trust the gauge but the needle did give a slight kick as it fell at about 40 psi - just where it should control. Also I would have expected the regulator to hold pressure at its control point when the pump stopped, mine let it fall all the way to zero.

Well I replaced the regulator and bingo, what a difference, goes like a bird. Haven't checked the consumption on a run yet but I'm getting 25 on a congested motorway (M3 - M25) run into work. Incidentally I opened up the broken regulator. I think the poppet valve had wedged itself half open / half closed (witness marks on one side of the seat and the poppet) so I guess this is not a common failure mode.

Guess what, the fuel pump is now quiet ! A blatant clue and I missed it.
Incidentally I have seen discussions about noisy transfer pumps, my car doesn't have one. I don't know why US cars need one and UK cars don't. ( See I was trying to find out whether the pump was OK, perhaps we should include sound clips with FAQs !!!)

The Moral

The moral of the story is 'donít give up'. If something is wrong it can be found using intelligence and perseverance. The BMW dealer may have better tools and so find the problem easier, but with the collective experience from fellow owners and personal motivation, I would bet a lot of owners would get the right answer for the least expense.

A word of warning however. I did a lot of messing about with fuel lines and my fuel pressure gauge set up was distinctly iffy. The fuel rail has quite a large volume and the pump can move a lot of fuel in a short time. Combine this with the sparks generated when hot wiring the pump (I used a lead between the battery and the pump fuse) and you have a recipe for disaster - SO BE CAREFUL and think about the possible consequences of what you are doing.

Finally I would like to thank all those who unwittingly helped by placing their experiences on the net, it gave me both ideas and encouragement.


Follow Ups:

[ Follow Ups ] [ 5-series (E28) Message Board ] [ Msg. Board FAQ ]
Questions, comments, or problems, please visit the Roadfly help desk. Logo © 1997 - 2018 Jump Internet Inc. All rights reserved.