check your plug wires, sounds like you're not running on all 4. My last car was VW and it had a very similar issue, luckily for once the retarded electronic diagnostics did me some good and gave me a code for misfire on cyl. 3. anyway....if you checked the plugs, now try the wires. even if you didn't get a CEL, try running it for error codes. If you go to autozone they will hook up a cheap OBDII scanner for free, its worth a shot.

 

Okay. I was thinking about changing the wires but figured what the hell. Thanks!

 

The exhaust doesn't smell right, smells bad. I can notice this after I pull into the garage

Your cat converter may be bad "smells bad" like what? It could also be a bad O2 sensor. That can cause the converter to foul and smell bad, run erratically.
There's always the extreme measure of asking somebody with both an OBD reader and some REAL knowledge. ;D

 

Well thank you but I had really no clue where to start looking. Tomorrow I'll take a drive to autozone unless anyone else has suggestions.

 

I just had one of these in the shop last week similar issues.

This machine has a dual-fuel, (can run on gasoline or propane), 3.0 liter GM four cylinder engine in it. It did set a DTC though, so it was pretty easy to track down with the laptop. The number two downstream heated O2 sensor was bad.

The exhaust smelled strongly of ammonia of all things.

Typically, you'll get a general "burnt" odor, or rotten egg smell. But nope, this thing stunk like ammonia. Owner swears up and down he's been using quality gasoline, no additives, etc.

 

If you had a trouble code, I would think your Check Engine light would be on, which you did not state that it is. Also AutoZone only scans OBD II which is some 96 and all 97 and up. Yours is OBD I and AutoZone will not scan it. So here is the info you need to do it yourself. So operating on the assumption that there is no trouble code, I would go about it like this. I would start out with a spark check on the dirty cylinder to make sure the plug and wire are good. I looked up your car and it used a distributor cap and rotor, so unlikely that the other 3 would be firing good if there were a problem in there, so isolate the dirty cylinder to make sure it has good spark. Just pull the plug out of the cylinder, ground it on the block with the plug wire attached and crank it over. If you are by yourself you can start it (it will run on 3 cylinders, which it sounds like it may be doing already). If spark is good then you need to verify compression, specifically on that cylinder. If a valve is carboned up and not closing fully you will drop compression immediately and even with spark it will still not produce much if any power from that cylinder, so if the spark checks out you need to check compression. Compare it to at least one of the other clean cylinders. I personally would check all 4, but if you are trying to shortcut 2 will be good. If your reading is low, squirt a little motor oil into the cylinder and test it again. If the compression goes up, that means your rings are bad. If the compression stays low, that means your valves are bad. If the compression matches the other cylinders without adding any oil, then everything is good in that respect. I am willing to bet one of these tests will reveal a problem. AutoZone will rent you a compression tester if you do not own one. Report back your findings from these tests and we can go from there.

I would not suspect converter at this point. Also I would not suspect 02 sensor at this point either. My hunch is a carboned up valve seat, but lets see what the tests say.

 

That's a pretty good write-up for checking compression, but as PaPa-G states the problem is intermittent. A mechanical problem (like stuck rings or a dirty valve seat) would remain constant, where a leaky cap or wire would fire intermittently.

I would put money on a plug wire or most likely the cap and rotor, PaPa-G.

 

My '97 Nissan Pickup did the same stuff.

New rotor button, cap, wires, and plugs fixed it right up.

 

Generally, troubleshooting works best if you dont just start spending money replacing things until you find what works. :nana:

 

Wow, didn't expect to find all this info. Thanks guys...

Looks like I'm in for a fun day tomorrow..thankfully I have off.

 

I am looking to modify my sig. Something along the lines of Orange is a freakin' god would be good. Lets see what we can work out tomorrow. ;D

 

Sweet! Hey uh, what you doin next weekend? Come down to indiana...i have a ton of front end work that needs done to this POS..

 

It sounds like you have more than one thing going on, and it doesn't sound uncommon with these models. If you did just have a bad plug or wire or cap, or....if you did have an intermittent compression issue, none of those situations would cause the idle to vary so much. The symptom you mentioned about the idle increasing to way above normal, and then back down again, points towards a vacuum leak, or issue with the idle control. Or it could even be slightly low on coolant, which is confusing the control system into thinking the engine is warm/cold/warm/cold, etc.

From a Honda forum:

http://www.honda-tech.com/showthread.php?t=2719160

There are quite a few more examples out there....

 

Spark is good. I'm doing the EUC test now.

 

So...I did the jumper thing. There's 6 holes you can plug the wire into, only the bottom three look like terminals. Anyway, I tried 1&3, 1&2, and 2&3 and neither combo gave me an engine light error code flashing array.

 

That dont sound like the right plug, all the descriptions say it is only a 2 terminal plug. Take another look, one of those articles actually mentioned the 2 wire colors also. The pic also looks different than the one I posted. Similar, but different.

Edit, the 3 terminal plug is called out in that diagram pic I posted, it says DO NOT JUMPER that one.

 

Ahhh you were right. It was just buried up there.

OK, so the ABS light blinks. It blinks 4 times, pauses then blinks two more times. The sequence repeats.

 

What the hell, that doesn't even match up with any error codes.

 

That would be an ABS error code:

94-97 ABS CODES (thanks to JimBlake):

You pull these codes the same way you pull check-engine codes, but it's the ABS light that's blinking. Here's the sequence...

Insert the jumper, then turn on the key.
Long (2sec.) flash - long (2sec.) pause - this isn't the code yet.
Short flashes for the first digit of the code - 1sec. pause - short flashes for the 2nd digit.
Long (5-second) pause, then the next code.

(94-97) ABS DTCs...
1 - ABS pump motor over-run
1 2 ABS pump motor
1 3 high pressure leakage
1 4 pressure switch
1 8 high pressure system
2 1 parking brake
3 1 pulser RF
3 2 pulser LF
3 4 pulser RR
3 8 pulser LR
3 12 different diameter tires
4 1 wheel sensor RF
4 2 wheel sensor LF
4 4 wheel sensor RR
4 8 wheel sensor LR
5 - rear wheel lock - R/L
5 4 rear wheel lock - R
5 8 rear wheel lock - L
6 - fail-safe relay - F/R
6 1 fail-safe relay - F
6 4 fail-safe relay - R
7 1 solenoid - RF
7 2 solenoid - LF
7 4 solenoid - R
8 1 ABS function
8 2 CPU comparison
8 4 IC self-check


That matches with Left Front Wheel Sensor. While this is good for you to know, it should not cause any issues with the way your engine is running. You better do the compression test next.

 

whats the best way to do that?

 

Guess I will need a guage...ugh

 

Yup, AutoZone will rent u one if you do not own one. Make sure you have a little oil can too, or some way to convey a squirt or two of oil into the plug hole for the second part of the compression test if you need to.

 

Compression test done.

From left to right looking down at the engine.

C1: 160
C2: 145
C3: 100
C4: 150

 

Was #3 also the dirty plug? You also need to test #3 with a shot or two of motor oil to see if the # comes up or stays the same. You are getting warmer.

 

Ok my brother stopped by and I had him watch it while I was cranking to give a more accurate result....the gauge pressure drops after you cut the engine.

anyway.

C1:180
C2:150
C3:150
C4:180

 

What a crappy gauge, the pressure should not drop at all, it should hold peak pressure until you release it. In any event, are you confident in these new readings as being accurate and repeatable? And were all 4 of these done dry of with oil added?

 

Ok I'll try again.

 

Make sure you start it and run it for a minute to burn the oil off.

 

It will be, my brother left. So im going to have to do it myself. I need to put the car togther and head to the store for some batteries for my flashlight. Im going to record what it reads on my camcorder so i can see it for myself.

 

Rock On. :rock:

 

Ok. I ran each 2-3 times. FINAL results

C1: 180-200
C2: 180-190
C3: 130 runs 1 and 2...180-190 after i put in some oil.
C4: 180-190

 

Well that sounds like there is some definite ring wear there in cylinder #3, or possibly even indicate a cracked ring. A difference of about 30% is pretty significant, and could cause some issues in the way the engine runs, and definitely could cause a reduction in power, but if you can generate 130 psi repeatedly dry, I have my reservations about saying that is the root cause of your problems without further troubleshooting (not to mention a ring job is a pretty major project).

Are you sure your check engine light works? I.E. does it come on for a second or so when you start the car and then turn off? Just want to make sure there is not a code in there and the bulb isnt burned out.

 

Yes sir. It comes on when the car starts, and it came on when I did that test you had me do with the jumpers. It stayed on for like 2 seconds then went off.

 

Well, I am no Honda expert, so I do not know if they have the same engines as Accords, but being the same year I would bet they at least have the same types of engine control's. So if you have that service manual handy, I would start down this road now:


Hey everyone, I was working on my 95 honda accord today trying to fid a
surging idle peoblem that has gotten worse lately. I start up the car
and the rpm goes from 1000 rpm to 3000rpm sometimes.

I logged on to google groups to find help from other people that have
had the same problem. I FINALLY FIXED IT! Since this is a common
problem in honda's I figured that I would log on and explain how I
fixed it.

Here's the deal:

***DO THIS FIRST: Buy a "Hanes" manual for the car that you own. Look
in the trouble shooting section for "idle surging" or something like
that. Go through all the tests they have in there. The manual is
prolly 9 Bucks at Auto Zone. GET IT, SO USEFUL!

There are two valves that controll air flow to the engine and can
affect idle:

IAC Valve: (Idle Air Controll) This is a $220.00 valve that bolts
onto the throttle body. It consists of an electrical motor and a
plunger that moves in response to a computer on the vehical someplace.
There are great instructions in the "Hanes" manual on how to test this
part.

FITV: (Fast Idle Thermo Valve) This one costs $55.00. This little
puppy is not electrical dependant. It is a "thermo" switch. There is
a wax plunger that expands when warm and shrinks when cold. When the
engine is cold, the valve is compleatly open, increasing the air flow
to the engine to warm it up faster. As the engine warms, the plunger
expands and lets less air into the engine to slow down the idle.

*****THESE ARE NOT THE SAME THING!!!!!! One (AIC) is bolted to the
back of the trottle body close to the wall of the dashboard. The other
one (FITV) is bolted to the BOTTOM of the throttle body. Each one of
these valves can be removed and tested and CLEANED WITH THROTTLE BODY
SPRAY. Always clean before buying a new valve. This is what I did and
I fixed my problem for FREE!

I tested the IAC valve and found no problems. I took it off and
cleaned it. I then took off the Fast Idle Thermo Valve and took it
apart to clean it. I discovered that the entire inside is supposed to
be screwed into place on the assembly. Apparently the center/core of
the valve has slowly, over time, juggled out and the air was leaking
through regardless of what the little plunger was doing.

 

Well I have the Hanes manual already. I can try that. I wish the guy also mentioned having a power loss issue.

 

So if you have that service manual handy, I would start down this road now:

I would've started down that road back on page 2. Poor compression can cause a lot of issues, but it won't increase your idle speed. We've got a really nice compression tester and cylinder leakage tester at work. They're the least used *measurement* tools in the shop....

 

Well in my experience, idle controls typically will set a trouble code. Also all idle controls I have had experience with when they fail into open loop mode will cause a consistent high idle, and do not cause a loss of power, other than a slight drop due to running rich all of the time. The loss of power, and 1 dirty plug out of 4 were worth doing the compression test, and to overlook it is to miss an easy way of diagnosing several potential problems. For example what if the timing belt skipped a tooth (somebody mentioned to me via PM that this is can occur to these engines with as many miles as Papa's has). A compression test would give you an indication of that at least. As I mentioned I am not a Honda expert, but I do know a thing or two about engines, and feel the compression test was certainly worthwhile, as it did reveal some problem with cylinder #3. Exactly what that is will require further looking into.

He!! it could even be a bad fuel injector at this point too, but there is a process to follow to get to that point, you dont just ignore the basics.