Hi;

Yup the same one. Still fails on NOx.

We have basically checked everything. The EGR was tested by not only
false triggering it by ground the solenoid wire, it was also removed to
check for free flow when open and proper sealing when closed. This a 5
wire type, so it has a position transducer so if the computer told it
to actuate and it didn't, it would know.

The IAC was checked the most functional way I can think of, turn key
on, then unplug IAC, start engine. It will idle high, plug in IAC with
engine running and RPM drops. It was A OK.

Had a MAP sensor handy so just changed that. Put in a brand new O2
(lambda) sensor.

Now it achieves closed loop operation and after the codes were cleared
has not set any new ones.

Still fails on NOx !. So we put a new cat on it, and it STILL fails on
NOx !

I am starting to think the problem is caused by excessive cylinder
imbalance. That either by partially clogged injectors or vacuum leaks
one or two cylinders are running too lean.

The way I explained it to my buddy is this : When you got a lean
cylinder you get NOx and HC, on a rich cylinder you get HC and CO.
Mixing these two in the exhaust does not make them go away.

At least that is my reasoning at this point. There is really not much
else I can think of. You ?

I am leaning toward vacuum leaks because the car has good power and
even though the idle is not perfectly smooth, it does not have a clear
miss. Off idle or under load it smooths out perfectly.

The other thing is that it does not read all that low on HC and CO. It
doesn't fail, but I have seen alot better readings from cars that run
this good. Especially with a new cat.

In fact my Dad's car which has the 3.1 version of the same engine read
better with a totally blown out cat. It was replaced because it got
loud, not to make it pass, it had already passed.

Now I do remember how they used to check for vacuum leaks, by spraying
starting fluid on the suspected areas. Is it still done this way ?
Should I disconnect the IAC while doing this ? That would seem logical.

There is one other thing. Cars still jump time. If the cam lobe design
is anything like what the norm used to be for a passenger car, jumping
a tooth would actually increase real compression, and lower average
vacuum. Given the conditions, I think it is a real possibility. IIRC,
the system in this car does not use a cam sensor because it does not
need it. Firing the injectors 3 by 3 and having 3 coils it simply
doesn't need a cam sensor. Firing the coils 2 by 2 andf the injectors 3
by 3 there are waste sparks as well as waste (kinda) injections.

On a more advanced engine like a 3.8 it needs both cam and crank
sensors because the injectors are fired seperately. These engines can
detect an error in cam timing but not a 2.8/3.1. I remember my other
buddy with one of the last 2.5 "brick ****house" engines that jumped
time. It would run, so he takes it in to Mr Badwrench and the machine
that costs the national debt said to replace the engine.

I looked at it and after much pondering told him "tear down for timing
chain". Later he showed me the almost non existant tensioner and the
chain with the burrs all over it. After a set of timing irons (my pet
name for the chain and tensioner and maybe gears) he was back in
business.

To support that theory, the 2.8 has very little backpressure at idle. I
mean even with the EGR tube open it is not all that loud. I took it off
and pulled the plunger manually and I just didn't think there was
enough coming out of it. That is, however, a subjective judgement.

Since it is in closed loop without an EGR code, I can only assume the
plunger is moving when told to by the computer. We are sure it is not
clogged. We are also sure the EGR is not leaking.

So what, in your opinion is the better way to go ? Spray for vacuum
leaks, check cam timing or look seriously at the injectors being partly
clogged. The idle sounds a but like a lean misfire, and the power is
good and smooth under load. By my reasoning the idle is affected more
than the power range. Doesn't this point to a vacuum leak more than
injectors ?

If someone is going to up and say "injectors", would injector cleaner
be a good idea ? Normally I avoid it like the plague. When the engineer
designs those injectors there is a paper that says what they are to
inject, and there is one word there ; gasoline. I also shun injector
cleaner because when the injector is really clogged, the cleaner will
never get to it. However there is always an exception. These do fire,
but perhaps flow is decreased in a couple of them.

In a 3.8 I would take and ground each injector line from the computer
and see how much it takes to flood any given cylinder, but they are
wired together in two banks in the 2.8/3.1 so there is no easy way to
do that. There are only two wires at the computer and you can't really
get to the other end because they have the top plenum over the
manifold.

Anything helpful is greatly appreciated. This guy is just back to work
and has to drive all over the place. He has to periodically go begging
for license plates and/or like last time, sign the car over to get
another temp tag. We let him use a spare set of plates for a few
months, but we got rid of the car and it would have to be tested to
renew them. With the new parts and the new cat, even if he could afford
another car right now, that would waste everything we put into this
one. He does construction labor and this being a station wagon, it is
the perfect work car for him. It's a breeze to drive compared to a
truck and on the rare occasion he has to carry a ladder it will work.
It locks and his tools fit.

We are getting to our wit's end here. We have even considered some
skullduggery, like running it on pure ethanol for the test, or even
test driving a car with the same engine and real quick switching the
enigines and returning it. We don't want to do either. When I was a kid
we would consider stealing a car and popping the engine in it, but no
more. I/we don't want to do anything like that. Getting caught is not
the problem, we could get away with it, but we do not WANT to. It is
immoral. Well the ethanol isn't but that doesn't help him when test
time comes around next year. We are in our 40s and want to do things
right. There is no good reason this thing cannot pass.

Well there is, but we are having trouble finding out what that reason
is.

Anyway, don't you just love my posts ? I know how it is. As much as we
tell people to be thorough and give details as to what you have already
done/tested instead of "Da thing don't work", I seem to be at the
opposite end of the spectrum. Sometimes at work I am called upon to
help another tech, and sometimes I say this, they reply "checked it", I
say that and they say "Replaced it". While that tells me they are doing
something, it doesn't make it any easier.

So as you read my post and are about to say "______" and the next
paragraph tells that I already did, don't get discouraged, please.

The other thing is that I think of time as money, not that I need pay.
This is almost a personal challenge. Like I said the guy is just back
to work, I bought the cat. I'll loan him the money for gaskets, but do
we need the intake set or the timing case set ? Actually he has a 30
day now, by the time it runs out he'll have a check, but I still do not
want to steer him wrong.

What, in a way, I really need are ways to definitively check these
things without a teardown. Got any ? For example, can someone say that
at a certain reading of the timing marks (if any) that a certain valve
under the front valve cover should be fully open, like at the peak of
the lobe ? Or is there a way to plug up the exhaust and pump air into
the intake to detect leaks ?

I am always looking for new ways to test what is seemingly untestable.
Pros have a few tricks, some of which they'll give up, others maybe
not. For example I have extreme shortcuts when it comes to fixing TVs.
We need that to make money.

I din't want to make this too long, but I have had to adapt to figure
out how to find the fault in TVs that are almost impossible to service.
The scope is the first thing for me. Dead set, got SMPS standby, put
probe near transformer. see RF ? move on to main SMPS and turn on, run
for a sec and shutdown ? Get to the flyback, but set to 10uS triggered.
See a five fingered waveform ? Bam, throw in a fly and wipe it's ass
and ship it. By wipe it's ass I mean give it a tuneup, clean the lenses
etc. and touch up the convergence. Auto diagnosticians surely have the
same shortcuts.

For example what alot of people don't know is before you buy a high
mileage car there is a test. Take the oil filler cap off and put a
piece of paper over it. It should stick at idle. There should be vacuum
there. If it is blowing out the engine has compromised sealing of
compression to crankcase. Don't pay alot for it, it is worn out. I do
this before even thinking of looking at the odometer. I know a few
things having been around cars alot, but the pros know more. I even
know about the brand new defective radiator caps, that thingy in the
middle should be spring loaded to the rubber seal. In some cars this is
very important, especially bigblocks, if you happen to still have one.

Kinda what I am looking for is how to check what. On the other side of
the screen I know you can't troubleshoot it from there. I seem to be at
a loss for testing techniques at this point. If someone can just up and
tell me what the problem is it would be great, but I don't really
expect that. Also, better techniques will come in handy. As cars get
more and more complex it ain't getting any easier.

Anything you got is very appreciated. Give me your address and I will
at least send you some beer, if you want it. This is just one of those
important things to get done, and it has become somwhat personal. I
can't let it win. If it only knew that if it wins it goes to the
crusher :-)

Everybody have a good year. I await your wisdom. Now I am off to scan
the recent posts to see if I can spead some around, in the spirit of
"pay forward".

Have a good one.

JURB