91 Hyundai Excel Running Rough - Smoking

What color is the smoke? White, blue, or black? And what does the smoke smell like? Is it a sharp smell, or a toast smell, or a sour smell?

--scott

Definitely black on startup - leaves a sooty patch on the driveway where it comes out of a pipe I've attached to the exhaust. Smells like exhaust from a normally running vehicle except stronger. A sweeter smell than burnt toast.d

Anybody? All comments greatfully received. A similar problem with a similar vehicle? How likely is a fuel pump problem - as it stalled completely while running it seemed at the time that it might be that. Could a replacement of the fuel filter have solved the problem briefly? What is a good diagnostic for fuel delivery problems - preferably less work than the cure? What is the cure?

Thanks

That black stuff is partially-burned fuel. It COULD be because the engine is running super-rich, and there isn't enough air to burn all the fuel properly. It could also be because the timing is screwed up, too.

Start from the beginning. Do a compression test on the engine, and make sure there isn't something goofy going on. Get out a timing light and check the ignition timing. Take off the shroud of the timing belt or remove the chain cover and make sure all the notches line up perfectly.

The oxygen sensor will be bad... but it is likely to be bad BECAUSE of all the junk pouring down the tailpipe and not causing it. Even so, pull all the codes from the computer and see what could possibly be causing it either to be running super rich or to be pulling the timing off. Get a real scanner and start looking for irregular waveforms or any sensors that are out of range.

It's not a ring issue or an oil issue... these will cause blue smoke that smells like burning motor oil.

IF the fuel filter is partially clogged, the system will tend to be running too lean and not too rich. That's not going to produce so much black smoke but it will lead to some interesting smells at the tailpipe which are very characteristic of a lean engine. You aren't seeing that so that is not your problem.

A clogged return line or a failed fuel pressure regulator can cause all sorts of odd problems, though... and they will tend to be worse at idle than at running so it's easy to find them with a fuel pressure gauge attached to the rail.... too much fuel pressure means way rich operation.

--scott

I'd be looking for a fuel injector stuck open or a blown o-ring on one of the fuel injectors allowing fuel from the rail to be pouring into the engine. Unplug the fuel injector's electrical connections one at a time until you find the one that makes no difference in the idle, remove the injector and either preemptively replace it (they cost about $100 for most cars) or get hold of the test procedure for it. If the rubber o-ring(s) around the injector come out damaged, you've found the problem and the fix is just new o-rings.

The o-rings on a fuel injector would would not cause a rich condition. The lower o-ring, if bad, would cause a vacuum leak. The upper o-ring would cause an external fuel leak. The only way an injector could cause a rich condition would be an internal problem, such as being stuck open. Either way the inctor would need to be replaced or rebuilt.

I agree with your opinion about the injector itself, but respectfully disagree about the o-rings. I have experienced it first hand on a '91 Nissan

I believe you may be talking about a throttle body injection setup. If it was throttle body injection a bad o-ring would leak fuel like you claim but most, if not all, fuel injection sytems today are now port injected The port injected sytems would work as I described above.

Thank-you all for listening to the story and for the practical, reasoned suggestions. I will be counting dead mosquitos :> I will re- post whatever the outcome.

The experience I am talking about is on a car with port injection (separate injector over each cylinder). A bad o-ring on an injector on that car allows pressurized fuel in the rail to squirt past the sides of the injector into the intake. If you check the old 240sx FAQs, it is a common problem that usually happens when replacing injectors on that vintage Nissan. Its easy to kill new ones as well. Unless the o-rings are properly lubed, sharp edges in the hole will cut them on the way in and the fuel just pours through.

Granted, the OP doesn't have a Nissan, but since his car is also Japanese & the same model year it might have a similar injection rail setup. Worth a look anyway.

It looks as if the timing belt was serviced, although the notches line up exactly. TDC = notch on cam sprocket = protrusion on casting. If the timing belt had broken, are there valves that could have been damaged or is this engine clear of that problem?

The next step would be to take it to someone with the equipment - what constitutes a real scan? Can I learn anything using a scope with say

The fuel pressure is 50 psi at idle which drops to 38 psi when a vacuum of 25 inches is applied to the fuel return vacuum port. This is as specd by Haynes ch 4.

A funny thing happens though (funny to me) - when the engine stalls, the throttle control solenoid continues to adject the throttle. (chatters) How can this be - there is no signal being fed back to the ECM. Is this normal?

Thanks again for the education.

Did that - unplugging each injector causes a stumble and slightly increased rough idle. I took one injector out of the rail and the o- ring looks fine. I know you're supposed to renew, but when I put everthing back together, the engine behaviour was unchanged. There is only one ring on the rail (top) part of the injector and a rubber grommet seated in the intake manifold. Was someone talking about multiple o-rings? The grommet in the manifold seems loose fitting - is this normal? Thanks

Valves can get damaged in all kinds of ways. You need to do a compression test before doing anything, as I said in the first posting.

If the notches line up, this is a good thing. You still need to check the ignition timing and see if it's off too, though.

No, you need a box that has an OBD interface that will talk to your car computer, and display the values of the sensors that the computer is seeing.

It's entirely possible you'll see a good value on the scope, while the computer is seeing something entirely invalid because of a bad connection or bad A/D in the ECU. That's why you need to use a scanner that will actually show what the ECU inputs are.

That's good.

I don't know, that's a question for someone who knows the algorithm in the ECU for that particular model.

But first do a compression test to make sure you aren't just wasting your time.

--scott

I put the compression test results in my first posting; When I bought the vehicle I performed a compression test with the following results. All plugs were fouled equally.

A leak down test was performed with the following results;

I didn't know that timing would be a problem with electronic ignition so I broke/lost/threw away my timing light when I junked my last adjustable ignition vehicle :< Are we looking for 10 deg BTDC? - I dn't think Haynes specifies what it should be.

I am not sure whether the vehicle has OBD and I know that it does not have OBD II which came out in 1995/96. Do they still sell these scanners?