‘89 CRX won’t start - frustration

My ’89 Honda CRX Si won’t start.
-Engine turns but no start so battery and starter are ok.
-No computer codes. Computer is operational but it gives no trouble codes
so computer does not detect any malfunction.
-Fuel Pressure is ok - I checked fuel pressure with fuel gauge meter, at
37psi within specification.
-Spark seems ok. I hooked spare spark plug to each cylinder and there is
spark (not a strong spark but I don’t know how strong spark is supposed to
be).

-Had left car in garage for 2 weeks before this problem started.

I’ve had an occasional no start intermittent problem with this car for the
past couple of years but now the problem is permanent, which I’m almost
happy for since perhaps I can now find out what the problem is. The
intermittent problem drove me crazy (the car would not start about once a
month) and I never found out what the problem was/is.

Any ideas would be greatly appreciated.

Thanks.

 

First candidate is the main relay.

See
http://www.tegger.com/hondafaq/startproblems.html#mainrelay
..

If it seems to be this, it is not expensive to fix.

 

Spray some starter fluid directly into the air intake.

If the car wants to start now, you might have a fuel delivery problem.
Since your pressure is ok, your fuel pump is clearly running. I'd look
in the direction of the injectors (this thing has injectors, right) -
make sure they are all clicking. You can listen for that with a
automotive stethoscope.

If the car does not start, you must have an ignition or perhaps air
problem. Ignition is more likely. Since you are getting a possibly
spark, power is getting to the coil. Check the timing, coil and
ignitor.

Remco

 

Thank you for your replies
? I know how to spray started fluid on a carburetor car but how can I
spray starter fluid in a fuel injection car, there does not seem to be an
access point to the air intake.

Thanks for the link to the main relay description. It seems to me that the
main relay should be ok since I have fuel pressure, so the relay and
computer are feeding current to the fuel pump correctly (I can also hear
the fuel pump running for 2 seconds when the ignition is turned on). As
far as the power to the injectors goes I assume that if the main relay
were not supplying voltage to the injectors that I would get a code from
the computer (or will I not? ), but as I said I get no computer codes.

 

for other diagnostics, go to tegger.com and check out the no-start
faq's. despite your main relay test, intermittent no-start usually /is/
main relay in my experience, especially for a civic/crx this age.
usually accompanied by a code 16 though. re-soldering is cheap if you
have the tool. i recommend properly de-soldering to make sure all the
old detritus is removed before re-soldering. '89 crx main relay
location and removal is the same as the 88-91 civic which is very well
covered in tegger's faq's.

sounds dumb, but you could also try changing the plugs. i was /given/ a
free civic last year by someone that had spent a fortune hunting down a
no-start just like yours. a new [used] set of plugs and it roared into
life! don't use bosch, use factory ngk's.

when testing spark leads with an old plug, open the gap right up -
sparking distance is different at atmospheric pressure vs. cylinder
compression pressure.

 

[[Top posting corrected]]

How can it be the relay if there's 30 odd psi at the fuel line?

Have a peek at the Rover 216GSi stuff on my website.

What colour is the spark you got? Is it weak and yellowish or
sharp and bluish?

 

I could not tell the strength of the spark because I was inside the car
turning the ignition and wathcing the spark through the winshield. Could
have been a weak spark but there was spark on all 4 cylinders. I hear that
spark is a lot easier at atmospheric pressure as opposed to the compression
pressure inside the cylinder and that makes sanse to me (I'm a physicist).
So could it be that the spark is strong enough to spark at atmospheric
pressure but not inside the cylinder?

 

Can't you come up with somebody to turn the key for you? You need two people to
do a home mechanic's spark test.

You use distance to compensate; if the spark can jump 1/4-1/3" outside the
combustion chamber, it'll have no problem jumping the gap in the spark plug.

 

I believe all Honda FI will put out code 16 if the fuel injector power
supply is missing. You might try to see if it's a flooded or listen for
the clicks out of the injectors.

 

This is pointless if you simply want to check for spark.

The correct way to check for a faulty plug or coil output is to
check the spark jump at the distributor. This is easily done alone.

 

Like Graham W said, if there's fuel pressure the main relay is fine.
That is how the circuit works.

 

Injector operation (that they ARE operating, that is, not necessarily
that they're operating PROPERLY) is easily checked: crank the car over a
bit, then immediately take a plug out and see if it's wet with gas.

Have you tested the compression?

 

er, ok, so when your *intermittent* fault happens to be /not/ evident
when you test, and faulty when you're trying to start, then what? fact:
for this age civic/crx, you're well inside 3 sigma on relay faults.

 

no it's not. because it's harder to spark a small gap than a larger one
at atmosphere, you may not see a spark, even if the system is ok.

not on a honda and trying to use a normal plug lead it's not. the
connector is too recessed and the hole too small for the lead to fit.
if you're /not/ using a plug lead, there's too much chance of leakage to
spark reliably, particularly if you're trying to spark on a plug without
the gap opened to work at atmospheric pressure. besides, if the
distributor cap or distributor arm are faulty you need to test
"downstream" of those components.

 

I have not tested for compression but I have no reason to believe that
anything is wrong with it. The car was working fine then I left it in the
garage for 2 weeks and now it does not start.
I know that the cranckshaft is turning properly, no broken timing belts
etc.

 

Just to eliminate some things:
Have you tried the cure listed at
http://www.tegger.com/hondafaq/startproblems.html#flooded
for flooded cylinders? This may happen with leaky injectors
after the car sits awhile.

I see your comments on your main relay check, but I am still
concerned about whether the pump continues running. You say
fuel pressure is there, but is it continuously there? From
looking at tegger.com's description, I am still suspicious.
See for example the more detailed discussion at

http://www.tegger.com/hondafaq/mainrelayoperation/badmainrelay.html

(Sorry, it's hard to tell exactly what's going on. It's the
price of diagnosing via posts.) If the main relay has never
been replaced or re-soldered, I'd sure be inclined to take
it out and re-solder.

The main relay is such a common problem on older Hondas. My
91 Civic's died in 1999, after about a year of intermittent
no start conditions.

As for codes, the computer and code setoff system are not
very sophisticated on these older Hondas. I do not recall
seeing codes set in the past when people have reported a
no-start problem and re-soldering or replacement of the
relay fixed it.

How old are your ignition wires? Are they OEM?

 

Yes, it's unlikely you'd lose compression on all four cylinders after
sitting for just two weeks - okay VERY VERY unlikely - but it's worth
checking just to cover all the bases. Low compression combined with
other issues could be the root of your problem.

 

Fuel pressure is there. After cranking for 20 seconds, no start and fuel
pressure still around 37psi. Also after cranking and turning the key off
the system seems to hold pressure, at least for 5 minutes which perhaps
indicates that injectors are not leaking (am I correct to assume so?)
I also took out all 4 spark plugs and cleaned them but again no firing.
The spark plugs were not completely dry when I took them out but I cannot
evaluate if they were too wet. I assume that after cranking for a while it
would be normal for them to be somewhat wet. I'll repeat taking out the
plugs and inspecting them after cranking.
I'll take out the main relay and resolder it tomorrow. I'm pretty sure the
pump is ok but the relay could still not be feeding voltage to the
injectors properly. Although one would think that that would result in a
code there is no description of the logic that the computer uses to detect
an injector problem condition...

 

Oh the ignition wires are pretty old and they are OEM. But could all 4
wires fail all of a sudden?

I'll do some more methodical testing tomorrow...

 

Anyone with near 20/20 vision can see the spark on a plug. To bend
the plug only ask for trouble. Each time it bends it has the potential to break.
If you don't see the spark then there is an ignition problem. Better to not
bend it.

Actually, trick requires just a paper clip... no plugs needed.

If you have difficulty checking for a dead plug at the distributor like the
way I do it on Accords, Civic then the timing light is your best friend.
For other's reference: Hook timing light to plug wire, no light = dead plug.
One dead plug = flooding. Flooding = no start.

You need to test both "downstream" and "upstream."