98 Prelude SH running rich, need advice

I'm trying to help my friend fix his 98 type SH (everything stock), and I
wanted to see if anyone here with a lot of Prelude experience has any
ideas. Heres what I have:
Got an o2 code (P0135) and a rich code (P0172). We replaced the 02 sensor
first, hoping it was the root cause of both problems. It wasn't. Now we
have a rich code, and the data list figures confirm it. After running long
enough to get into closed loop, the car idles low and stalls. It will run
at speed okay sometimes, but once in a while it will also cut out while
driving. We checked fuel pressure, and the #s are within spec. The fuel
damper is new. Drained the gas tank to be sure we're working with good
fuel. The map sensor is reading accurately. TPS voltage curve is linear
with no spikes or abnormalities. 02 sensor voltage occiliates from .2v -
.8v normally.
The fuel trim #s on this puppy are both around -18%, but of course, we
can't figure out why. Today we pulled the fuel rail and turned the pump on
to see if an injector may be leaking, but didn't see any leaks. I'll be
checking timing next as a matter of due course, though I don't suspect it
because the problem seems to come and go. If timing was the issue I'd
think the problem would be baseline and never leave.
We are considering injectors, but I'd like to know for sure that we
haven't overlooked anything as these suckers are about $500 for a set!
Anyone out there have any ideas for me? Thanks.

 

----------------------------------

Wow. One of the best questions I've ever read.

If there's any air in the cooling system the computer will do goofy
things with the mix / timing because it doesn't know the engine is
already hot. Have you made sure it's air-free?

'Curly'

 

One of the regulars here not long ago reported on his 250k
mile injectors on his 1991 Acura. The car was not running
well, yada. They were squeaky clean. He replaced them with
junkyard ones, which were also in similarly excellent
condition. No correction to his problem.

An automotive technology instructor I heard this past week
similarly said the injectors just don't foul in any serious
way these days. Gasoline is too clean.

Heard a fuel induction cleaner salesman say the same today:
The fouling happens not in the injectors but elsewhere,
because gasoline is so clean these days (at least for
injector ops).

For the record, is your new O2 sensor an OEM one?

 

well, for the record, my 89 automatic gets a flat spot in acceleration
occasionally which invariably clears after a tank of injector cleaner.
i don't think a cursory exterior inspection of a nozzle that shows it to
be "clean" has /any/ bearing on the interior workings of the injector
and whether it opens properly to give the correct spray pattern.

 

is the new sensor oem or aftermarket?

after the oxygen sensor, coolant temp sensor is the #1 prime candidate.
but also check the ignition wires - "weak" spark gives incomplete
combustion, poor idle, etc.

 

In an effort to pursue the simpler and more well known
causes of such behavior first (complying with Occam's razor
yada), how old is the ignition coil? Is it OEM?

'Cause, apart from the code, these are the symptoms (though
I am sure not exclusively) of a dying ignition coil. (Though
I could rationalize that code coming from a dying ignition
coil, too.)

snip

I should have clarified that I (only from reading here), and
of course you and the many other regulars here, are aware
that drippy (due to fouling) fuel injectors seem to be
common but typically are easily corrected via a bottle of
Chevron Techron fuel system cleaner. We agree on that.

But I also did not mean a mere cursory examination would
indicate the condition of a fuel injector. I meant that,
from what Tegger wrote on his experiences (and the injector
part of his investigation seemed to me rather extensive),
and the way this instructor and this sales guy were talking,
ISTM that injectors are not likely to fail.

I haven't seen the "hands on" side. It's only an impression
based in reading meant to give the OP some ideas,.

I am entirely for an exacting tuneup before the OP proceeds
further, "exacting" meaning making sure parts are OEM and
ensuring he does all the usual air, fuel, spark checks. Plus
throws a bottle of Chevron Techron into a near empty fuel
tank, fills, drives, etc. This establishes a baseline, and,
no, I do not consider it a shotgun approach but rather
sound, efficient automotive problem solving.

Check that ignition coil, too.

 

oops, what i meant was, check the output of that sensor. if there's
insufficient coolant, it can send the wrong signal, so make sure it's
full inside the radiator as well as checking the expansion bottle.

 

Thanks for the input, Jim. The o2 sensor is OEM, and I graphed the ECT
which showed linear temperature increase to about 196?F where it hovered
at idle. Looks normal to me. Closed loop is reached in a few minutes
without a problem.
Concerning the injectors, my suspiciion isn't so much a drip drip leak,
but maybe a pintle that isn't able to close quickly enough .
I hadn't been suspecting the coil so much in all this because Id think
incomplete comustion would not lead to negative trim numbers. Since the
oxygen sensor measures oxygen in the content in the exhaust stream,
wouldn't a negative fuel trim trend (indicated by too little measured
oxygen) mean that complete combustion is indeed taking place, albeit with
too much fuel? Seems like incomplete combustion would lead to unburned
oxygen in the exhuast stream, leading the computer to detect a lean
condition, thus displaying positive fuel trim #s in data list.
But anyway, checking the coil is a good course of action to establish
known parameters (especially before $500 injectors). I'd also like to hook
up the 5 gas analyzer to see what is really coming out of the tail pipe.
Thanks for all your suggestions, lay some more on me if you have them!

 

theres a lot of latitude in the ecu to allow for that. if it detects
excess fuel, it can keep shortening the injection window, so i would
think that lag shouldn't be a problem. besides, unless you have big
mileage on this vehicle [accords regularly go 300k miles without a
blink] the chances of genuine injector issue are very small.

unless you've had an igniter problem that's allowed the coil to bake,
the coil is lower on the suspect list. i'd check/replace plug leads,
distributor cap etc. first.

by definition, excess fuel means incomplete combustion.

i think that bit's right. as i understand it, i believe it's free o2
that's detected. incomplete combustion, regardless of lambda, would
read free o2. but there's more than one input parameter to an injection
width pulse.

definitely check the coolant content! you can't rely on glancing at the
level in the expansion bottle because any leakage allows air to suck
back on cooling, not fluid as it's designed to do. this is an issue
because any bubbles in the coolant stream do a number on the tw sensor
and that makes the ecu inject rich. also, the head gaskets on preludes
don't enjoy fantastic reliability, especially when driven hard, so check
into that. unlike detroit iron, gas leaks straight into the coolant, so
there's none of the "old fashioned" gasket leak symptoms like mayonnaise
under the oil cap.

 

Not a bad idea. I have a few co gas detector strips in my kit for checking
the head gasket. I'll try that a little later if I have time and report.

 

As a former Prelude owner I found this site very useful for info
http://www.absoluteprelude.com/