BIG Problem - HELP

Hello everyone,

As most of you probably know by now, I have a 1990 Honda Civic LX; 1.5L
DPFI 5-speed. I'm having a HUGE problem with it now and need help from
the experts. ;-) Here's the problem in a nutshell:

After running at a 2,000RPM idle for about 15 seconds: (high RPM idle
mentioned below)

Check-Engine light comes on, throws code 16 (fuel injector circuit)
Engine will not rev higher than 3500RPMs, if you try to, when the tach
reaches that level, the engine will "stumble" at that point, just as if
you have let off the gas. So, if you hold the gas pedal steady, the
tach will jump back and forth between 3,000 and 3,500 RPMs. (Based on
the sound of the engine RPMs speeding up and slowing down during this
fluctuation, I'm sure the tach reading is correct)

Engine falters and backfires severly during acceleration.

Prior to this problem, for the last couple of days I noticed a
scratchy-sounding noise coming from under the hood. This morning, my
engine stumbled hard, almost to the point of shutting off, the noise
stopped, and the problem started. :-(

I talked to 2 mechanics and they said to replace the throttle body
because it sounded like the throttle angle sensor was bad. Well, I took
their advice, and did so. (Luckily, I did it myself..wasn't too hard to
do either) ;-) Now, I still have the same problem, along with the idle
fluctuating between 1,800 and 2,000RPMs. I checked the idle adjusting
screw, and it is all the way out, so that's not causing the high idle
speed. I know I could have another bad throttle body, as the one I got
is a used one, but this does seem to be an electrical problem, and not a
mechanical one.

Does anyone have any ideas? Nobody around here seems to know. :-(

Thanks in advance!

Jonathan

 

I've never personally seen this on a Honda, but the wild fluctuations in RPM
could happen when something is wrong with the cooling system.
Make sure you don't have air bubbles in the system (maybe bleed it). Since
the thermosensor feeds into the computer, it can cause weird things to
happen when it doesn't see the right things.
I wonder if the weird noise you heard could be the coolant pump. Maybe the
injector code got thrown because the mixture is way off base but that is
pure conjecture.

This is just one possibility -- I am sure others here will have other things
you'll want to look at.

Remco

 

I've never personally seen this on a Honda, but the wild fluctuations in
RPM

Come to think of it, that termosensor (Engine Coolant Temperature Sensor or
ECT) actually sounds like another possible thing to check. You should be
able to read the resistance with the sensor on the car, so it is an easy
check.
I don't know what it should read, but a repair manual should have that
information.

 

Not an expert on this system, but the symptoms make me think of a vacuum
leak - a big one - into the intake manifold. That would account for the high
idle (the throttle can't control the air the engine sucks in) and the fuel
injector code (the ECU thinks it knows how much to inject for the throttle
opening it has ordered, but the extra air is throwing a wrench in the
works.) It doesn't explain the noise or the problem as the throttle is
opened farther, though.

A less attractive thought (that really doesn't explain the high idle) is
that the timing belt jumped a tooth. That also doesn't explain the noise,
but the stumble suggests it as a possibility.

I don't feel comfortable with either of these explanations, but it's food
for thought.

Mike

 

My son, Rob, has a suggestion. Possibly a defect that has caused the
ignition timing to become very far advanced. That will explain the runaway
idle and the 3500 RPM behavior (particularly the backfire out the intake),
and in turn the code. Best of all it is easy to check - if you see more than
about 30 degree BTDC when you try to rev it, you're on the trail.

Mike

 

There are two ways to correct your problem.
First replace every part that someone suggests is the problem, sooner or
later you'll hit on the right one and fix your problem. And think about it,
you didn't give in and let those ripoff mechanics get the diagnostic fee
they wanted (probably some outrageous figure like $40 to $60), but you only
spent $1200, but you got new, junkyard or different parts than before you
had the problem.
Or, you give in, pay that guy his diagnostic fee and find out it was only a
$50 thermo sensor or something similar, or worse, it costs more than the car
is worth and you don't get the chance to buy a bunch more parts for your
car.
Tough choice!

 

Thus far, I've only spent $65 + tax on the throttle body that I
purchased. Unfortunately, junkyards don't allow returns, only exchanges
for the same part, if the part I bought is defective. So I am still
caught between a rock and a hard place.

Jonathan

 

The best approach to problems like this is indeed two-pronged, but neither
path you suggest has much of a future. On one hand follow the positive clues
as far as they take us (high idle, backfiring out intake and exhaust when
trying to rev, injector circuit code thrown, sudden onset while running, and
the mystery noise before it all started). The other approach is to rule out
common or possible problems - valve or ignition timing, vacuum leaks....
This is going to be a hassle but despair would be premature. Let's follow
this until we fix it or have to take a different tack.

Mike

 

How do you tell what the "degrees" are, and what does BTDC mean?
(Sorry, I am a novice at this. I have enough mechanical inclination to
change parts, but I'm still new to all the technical specifics,
acronyms, and jargon) Do you suppose my timing belt could have been
making the scratching noise? Like I said, the instant the problem
started was when the noise stopped. (I still haven't heard the noise
again since this crap happened) Thanx for your input.

Jonathan

 

Sorry - didn't mean to use jargon. The ignition timing is measured in
degrees from Top Dead Center on the #1 cylinder by connecting a timing light
to the spark plug wire for the #1 cylinder (normally the one closest to the
front of the engine - the end with the timing belt and alternator belt,
etc.). The edge of the pulley on the crank has marks that are compared with
a reference mark or pointer on the timing belt cover when the timing light
flashes. Although the markings are often hard to see and vague in meaning,
for these purposes we only need to be sure the reference mark or pointer
points within, say, 1/4 inch of at least one of the marks when it is idling
and doesn't run way out of sight when the engine is revved. The ignition
timing normally advances with higher engine speed to allow the fuel time to
burn, but going to more than about 30 degrees Before Top Dead Center (BTDC)
means the ignition can actually occur before the intake valves close. If
that happens, the engine will backfire out the intake... and that's what
makes this interesting.

(Note - I don't know if your 1.5L is single or dual overhead cam.) The valve
timing will be related to the ignition timing, but can be checked without
special test equipment. If you remove the top part of the timing belt cover
(not terribly hard in most cases) you will be able to see the cogs for the
cam(s). By turning the engine *counterclockwise* (never clockwise - if it
overshoots more than a little bit go around again) until the timing marks I
mentioned in the previous paragraph roughly line up and the compression you
feel suddenly stops resisting and starts to help you, you can carefully stop
the engine where it just balances between wanting to turn CW and CCW. This
is TDC (Top Dead Center). It can be more accurately determined by taking the
#1 spark plug out and feeling the top of the piston with a long screwdriver
as it reaches the peak, but you won't have the feel of compression with the
plug out. With the engine in exactly that position, look at the cam cog(s).
There will be a dot forged into the cog(s) that should be lined up with a
matching dot on the engine block. If one or both (if dual cam) cogs is off
by a tooth or more, that's your problem. I hope that isn't the case since
that failure, while common if the timing belt hasn't been changed on
schedule, usually causes serious damage to the valves. I'm not expecting
that to be your problem, since the mechanics of the failure (cams tend to
fall behind the crank timing, not jump ahead) normally cause the timing to
be later and the engine won't start or idle. But it is a relatively easy
thing to verify.

BTW - there is no gasket for the timing belt cover - there isn't supposed to
be oil in there.

Mike

 

Thanx for the info. I'm still at a loss for what the problem is. I'm
99.5% sure it's electrical. Something with a sensor of some sort, I
suppose. I know I don't have a vacuum leak as mentioned before, because
vacuum leaks don't develop based on the start and stop of a scratching
noise, and vacuum leaks don't cause the ECU to throw an error-code 16.
When it starts, it definitely doesn't run right to begin with, and it
takes about 15-30 seconds for the check-engine light to come on. I wish
the 1990 Civic had a port to hook up to the ECU to take readings of all
the variables that the ECU measures during engine operation. Correct me
if I'm wrong, but this is how the ECU operates: The ECU has a look-up
table programmed into its ROM chip with a valid (or optimal) range for
each variable. When one of these variables goes higher or lower than
the value or range specified in the ROM chip, the check-engine light
comes on, and depending on which variables are exceeded, or under-shot,
the LED blinks a certain number of times. Of course, the number of
blinks corresponds to the assigned category or sensor failure, as
defined in the ROM chip. (i.e. Category (error code) 1 = Oxygen
content; optimal range of resistance is 1-4 Ohms; if Ohm level goes less
than 1 or higher than 4, check engine light comes on, and the ECU's LED
blinks once)

Sorry for rambling there... But, with all that computer logic in mind,
if there is something mechanically wrong with my car, then whatever the
problem happens to be is what is triggering the ECU error. However,
since the check engine light appears within 15-30 seconds, that's hardly
enough time to diagnose a mechanical problem that is causing an "extra"
electrical/ECU problem. *sigh*

Anyway, thanks again for your attempts to help me out. I appreciate it.

Jonathan

P.S. I'll probably have to take my car to a dealership to get this
problem fixed. I'm going to talk to a few more "Honda-specfic"
mechanics before taking it to a deaker though...not looking forward to
paying out the yin-yang.

 

That's pretty much how it works. Some of the measurements are indirect,
though, like inability to focus on a correct mixture is probably responsible
for the error code 16. The ECU says it can't control the injectors when it
may mean it can't control the fuel being injected in proportion to the air
being inhaled. (The Taurus with throttle body injection turned on the "check
engine" light because the injector seal was leaking and dripping gas into
the intake.) The delay would be about right for that error, since the ECU
has to try to gain control before it says it can't. Some codes are
no-brainers for the ECU and come on immediately, like loss of continuity on
a sensor that should only vary in resistance. Others can take hours to set,
like the "evaporation system" codes produced by a loose gas cap.

Probably not a bad idea to take this to the experts. The set of symptoms is
a bit of a head-scratcher. Good luck!

Mike

 

I started my car today, the check engine light didn't come on, but when
I revved the engine, it backfired really bad, and almost cut off. I'm
wondering if the idea about the timing belt slipping a tooth is the
case. Isn't it true that "bad timing" is what causes backfiring?

Jonathan

 

To be more precise, backfiring out the intake is always (I think) caused by
a timing problem. The cylinder has to fire before the intake valves close to
cause that.

Remember the timing is twofold - ignition and valve. Both have to agree with
the crank position. The distributor is driven by the camshaft, so the
ignition timing should be following the cam timing. If not, something is
wrong in that area.

I haven't heard of it in Hondas (probably because I don't get around
enough), but way back when I drove a Dodge that behaved pretty much the way
you describe except the idle was only lumpy. It could reach 40 mph downhill.
The problem turned out to be the metal gear on the distributor that engaged
the cam - a couple of teeth had broken off. Just a thought, although it may
be worth unfastening the distributor cap and trying gently to turn the rotor
with your fingers. It should either be steady, or may turn a bit in one
direction and spring back if the distributor has mechanical advance in it -
I don't think many modern distributors do that any more... do they? Anyway,
if the rotor is happy to sit in more than one position, that is not good.

Mike