What next for my gross polluter?

[skelly]

My car is about 5 times the acceptable limit for HC's. It has an audible and visibly detectable though subtle stutter to it while idling or under power. It has a brand new catalytic converter, perfectly sealing exhaust, and I replaced ALL ignition components (plugs, wires, coils and ignitor module) in an attempt to fix this problem. The engine does run significantly more smoothely now after doing all the major maintenance including replacement of all filters and all intake, exhaust and fuel gaskets/seals, but it is still imperfect and is failing emissions testing as a result. In fact there is a noticable degradation in performance as well, the turbo will barely spool up and make 5-7PSI.

My question is: what should I look at next? Because of the high HC's, I was told and understand that this is the result of uncombusted fuel making it through the system. Initially it was presumed that the uncombusted fuel was the result of poor ignition, but now that I know the ignition is in top shape, I have to presume that there could be another cause for poor combustion. A compression test reveals acceptable numbers across the cylinders within a few PSI of one another, so if I had to venture a guess, it was going to be that the injectors are sticking/leaking and dribbling non-atomized fuel, or otherwise leaking into the cylinders resulting in poor combustion - does this sound plausible?

I'm going to have to spend more money to get this puppy running correctly, but I'm not sure what to spend it on next. Should I pull the injectors and send them out for esting/cleaning/flow matching? Is the behavior I'm seeing something that could be caused by a bad O2 sensor? Is there any way to test the O2 sensor? I really don't like this shotgun approach to troubleshooting, just blasting money at miscellaneous parts until the problem goes away!

Also, though I haven't done a intake system pressure test, I'm 95% sure that there are no leaks in the induction system; I ran around it with carb cleaner and checked each joint while the engine was running with no resulting change. I also temporarily tapped a mechanical vac/boost gauge onto the intake manifold and it shows strong vacuum at idle, but the stock dash "gauge" bobbles around in the vacuum range without touching the throttle. Is the peculiar response of this gauge any further clue as to the cause of the problem? It seems to be timeg with the slight dip in RPM associated with the infrequent stutter.

The only other thing I can think to do at this point is to make a short video of the engine running and show how the gauges behave to see if anyone can recognize it.

I just want to drive my poor car. I've owned it several months now and haven't been able to drive it once without worrying whether I'm going to make it home...

Thanks in advance,
- SK

 

[Old Mitsu Tech]

I hope this helps. One thing that caught my attention was the fluctuating vacuum gauge. You didn't say if the replacement gauge fluctuated. If it does, then I would suspicion possibly a broken valve spring. As this could cause intermittent sealing problems and thus fluctuating vacuum. Another thing you could do is put a fuel pressure gauge in line and turn the key on. Check the fuel pressure and note it. Turn the key off and watch if it bleeds off. If it does then you have a leak in your fuel system. If you cannot see any thing external then I would think that possibly you could have a leaking injector. Finding which one may present a problem but maybe a plug read could help there. I would also assume that you have on hand a 4 or 5 gas analyzer. You can also pressure up the system and pulse the individual injectors and note the pressure drop over each injector to see if any one of them would differ. The pulser is available through Snap On or maybe even your local parts store. One other thing. According to several posts the 255 Walbro will run rich at idle on a stock pressure regulator. If you have one I would check into that. Not least is the pressure regulator diaphragm leaking. Pull the vacuum line with the engine idling and see if you can smell raw gasoline on it. This is a place to start. Let me know if it has helped. Old Mitsu Tech

 

[skelly]

Thanks for the response.

The idle vacuum is strong and solid according to a mechanical gauge. Only the air-meter driven, phoney-baloney electronic one on the dash fluctuates a bit when the engine stutters. The mechanical gauge remains at a steady 21inHg or something, doesnt budge until I touch the throttle.

I don't think a valve spring could be to blame herre because the engine runs very quiet and smooth aside from the stutter and the compression checks out fine. Broken valve train parts have a nasty habit of making your compression drop to nearly nothing.

I would like to check the fuel pressure, but I don't see any conventient location to tap a FP gauge into the system without cutting the hose between the rail and the filter. Not ready to do that just yet because it will become a failure point after that.

A plug read would indeed reveal if one is fouling moreso than the others, however it is worth noting that the engine's running operation doesn't change significantly if any one spark plug is pulled. This makes me think that no one particular cylinder is causing a problem, but rather ALL cylinders are seeing it. As a result I don't expect to see different readings on the plugs. Even so, I would check them, but I need to put some more miles on the thing while it's running like this in order to see any results as the plugs are all new and the fouling takes time.

I only have a stock fuel pump. Another thing I was thinking of trying actually was to re-wire the pump with a heavier gauge wire to make sure that it is getting all the current it needs to run reliably. But that's another one of those "shotgun" diagnostic things with no real underlying justification. I've already spent about $800 replacing crap on this car while taking that approach.

The fuel system does hold pressure after the power is off. This was verified when I broke the seal on the banjo bolt at the fuel filter when I changed that filter. There is no smeel of fuel, no leaks of any kind anywhere except a couple drips coming from the power steering banjo which is missing a crush washer.

Any other thoughts?

 

[Old Mitsu Tech]

Ok, now that we got that out of the way. The following can cause High HC's

1 Coolant Temp sensor. If the ecm sees very low temp it will modulate the injectors longer for the cold temp. You will need a scan tool to determine this
2 TPS If this goes way high you will have problems also. Scan tool here too
3 MAF Same here You should have a check engine light if any one of these goes out of parameters.
4 Yes the o2 sensor could be the culprit. You can check it with a good dvom with analog scale to check if it is moving from 1 volt which is rich to below .450 volt which would be considered lean. If it is not moving when you blip the throttle then the problem could be a lazy o2 sensor.

Try these. Old Mitsu Tech

 

[ddavisaf]

Ok, now that we got that out of the way. The following can cause High HC's

1 Coolant Temp sensor. If the ecm sees very low temp it will modulate the injectors longer for the cold temp. You will need a scan tool to determine this
2 TPS If this goes way high you will have problems also. Scan tool here too
3 MAF Same here You should have a check engine light if any one of these goes out of parameters.
4 Yes the o2 sensor could be the culprit. You can check it with a good dvom with analog scale to check if it is moving from 1 volt which is rich to below .450 volt which would be considered lean. If it is not moving when you blip the throttle then the problem could be a lazy o2 sensor.

Try these. Old Mitsu Tech

Excellent advice.
I would reorder these.
.1-.9:Boost leak test (do it, don't argue, trust me....)
1: o2
2: coolant temp
3: MAF
4: TPS

I want to know what kind of plugs you're using and what they're gapped at. This isn't in relation to your problem, but I just want to make sure everything is jif, just like the peanut butter and smooth....
I just made that up on the spot btw, I know thanks, I'll be here all week!!

Your problem sounds like you're losing unmetered air, or your have a faulty sensor.
Two more questions.
Does your CEL come on when you first start the car during the light bulb test?
Is your CEL on during normal driving?

I would like to check the fuel pressure, but I don't see any conventient location to tap a FP gauge into the system without cutting the hose between the rail and the filter. Not ready to do that just yet because it will become a failure point after that.

You did, stated here...

The fuel system does hold pressure after the power is off. This was verified when I broke the seal on the banjo bolt at the fuel filter when I changed that filter. There is no smeel of fuel, no leaks of any kind anywhere except a couple drips coming from the power steering banjo which is missing a crush washer.

Your fuel pressure system is peachy. Leave it alone.

 

[Strm Trpr]

I read in a post some where on this site about someone having very similar issues.
He also had a 90 dsm.
He swapped his 90 ECU for his 91 ECU all of his problems went away.

 

[skelly]

Excellent advice.
I would reorder these.
.1-.9:Boost leak test (do it, don't argue, trust me....)
1: o2
2: coolant temp
3: MAF
4: TPS

Okay, I'll look into that stuff next. It would be nice if it was as simple as a boost leak somewhere...

I want to know what kind of plugs you're using and what they're gapped at.

They're OEM spec NGK plugs and if memory serves they're set at .035; I can't recall off the top of my head exactly, but I gapped and re-checked them a couple times according to specifications in the book when I was convined that it was an ignition problem...

Your problem sounds like you're losing unmetered air, or your have a faulty sensor.

Quite possibly. I can't imagine where though. Hopefully a pressure test will reveal this. I disassembled, cleaned and inspected every single vacuum hose and intake system tube, fitting, etc and everything looks to be in good shape and went back together tightly. The ONLY thing I did not inspect was the intercooler itself to make sure it's not ruptured. I wasn't up to removing the wheel and puling the plastic shroud away from it at the time. The pressure test should reveal that pretty quickly though.

Does your CEL come on when you first start the car during the light bulb test?

Not sure, I'll have to check that - looking for a dead bulb would not be a bad idea, because...

Is your CEL on during normal driving?

No, there is no CEL while the engine is running. The only dash light I have is the one for the radiator/coolant; I believe it to be the result of the coolant reservoir's float thing having come disconnected from the microswitch that alerts for low coolant level.

So it sounds like I have a couple more things to look at before I start spending money again...

Thanks fellas!

- SK

 

[skelly]

I read in a post some where on this site about someone having very similar issues.
He also had a 90 dsm.
He swapped his 90 ECU for his 91 ECU all of his problems went away.

That's a somewhat inspiring story, but I think I'll reserve that for the last. I did have leaking capacitors on my ECU which I removed, cleaned the electrolytic from the board, made certain there were no carbonized shorts, and then replaced. The eletrolytic fluid had destroyed another surface mount capacitor which I removed and just left off. It was merely a noise filter cap and there were two to three others bridging the same ground planes, so I'm confident that the loss of the one will not have a significant impact on the operation of the PCM.

That being said, it is without a doubt that the 91 ECU is going to be a more evolved design which makes up for a variety of problems discovered in the earlier model, so it is not a bad prospect to consider if I'm up against a wall after the remaining diagnostics.

- SK

 

[ddavisaf]

They're OEM spec NGK plugs and if memory serves they're set at .035

The book is actually wrong, they should be gapped at .028-.032

Otherwise, you're

Preform a Boost leak test
Then we'll go from there.

 

[ddavisaf]

I must also say skelly that it is extremely refreshing to converse with you on this matter. Your attitude and knowledge is a step above.

 

[Strm Trpr]

I must also say skelly that it is extremely refreshing to converse with you on this matter. Your attitude and knowledge is a step above.

Indeed......

Do the boost leak test like ddavisaf mentioned, and don't forget a spray bottle with soapy water to help locate the leak.

 

[skelly]

:: Well it helps when you've been around a few cars and do all your own work to speak at least somewhat intelligently to the topic.

Thanks for the advice - I'll start poking around on it again soon!

 

[akdsmer]

Excellent advice.
I would reorder these.
.1-.9:Boost leak test (do it, don't argue, trust me....)
1: o2
2: coolant temp
3: MAF
4: TPS

Your fuel pressure system is peachy. Leave it alone.

Let me reorder these for you.

1) Give us all the tailpipe (gas) readings from your test. This makes all the difference to determining whether this is a lean, rich or mechanical/ignition issue.

2) Perform a RPM drop of your cylinders to find if it is a individual problem. I do this at the injectors, although it requires the removal of the retaining clips on the injectors. I also pull the throttle cables away so that it is easier. Poor cylinders create less of an rpm drop. If you find one cylinder that has less of a drop than the others then you have found the offending cylinder. This is important because you can focus on that cylinder. If you can't identify a single cylinder than you are back to a system problem.

At this point you can create a diagnostic strategy.

I'll give you an example. I rebuilt my motor. It had sat for 6 months while I remodeled my house. When I started it up, I could feel the misfire. Now I could start checking everything on my motor, but why not see which cylinder... 1) I pull the injectors one at a time. Number 4 cylinder drops RPM, but not as well as the other 3. 2) I pull the spark plug to make sure there is spark present. Yes. Because it's easy I swap the spark plug (they were new NGK's) to ensure I didn't have a bad plug. The RPM drop is still weak only on #4 cylinder. 3) I remove the injector from #4 and swap it with #3. Another RPM drop show that the misfire moved to #3. Guess what the problem was. (by the way, sometimes sticking injectors can be made operational by tapping on them (not too hard) with a hammer or such. It frees up the sticky pintle).

This took about 10 minutes to do. I would start there.

Another example...Misfire (high HC) during my I/M test. 1) I do an rpm drop. I find #1 cylinder with little rpm drop. 2) I swap the plug, the injector with no change. 3) I use my labscope and amp clamp to verify a good clean injector signal from the PCM and the amperage is good. (High amperage due to wrong/defective injectors can turn off the driver circuit. Can it do it to this vehicle, I don't know....I just check!) The command is good, amperage is normal. I then look for a vacuum leak. 4) I used propane and some carb cleaner (at hand) and check around the #1 cylinder. It sucks because the coil is right there, but be I spray it good and find the rpm picks up. 5) I remove the intake manifold and replace the gasket (torn gasket). Misfire gone. Outside of the labscope to check the signal, easy to diagnose.

AFTERWARDS we can start to deal with a system wide problem. To diagnose this we need to know what the other 3 (NOx doesn't matter) gases are: CO, CO2, and O2. We can then determine if it is a lean, rich or mechanical/ignition misfire. The only reason to look at the MAF, O2 and CTS (inputs is if the high HC is accompanied by either very low or very high CO. If you CO is around 1% you do not have a lean or rich problem causing the misire and looking at this data is a waste of time.

A note:

HC (Hydrocarbons) is raw, unburned fuel-this occurs due to a misfire. The feedback system controls the amount of fuel, but it's effect is mostly CO (Carbon Monoxide) by adding or removing fuel.

Your fuel pressure is UNKNOWN unless you know the system pressure and volume. If the fuel system pressure is only a 5-10 psi low, then the VOLUME of fuel delivered to the injectors is quite low. Whether this could be a cause of your problem is uncertain until the following has been addressed:

IF a PCM rich/lean command problem exists then the following order is used:

1) Fuel pressure- The computer doesn't sense this, although you can usually use fuel trim to gauge whether it's a problem.
2) The feedback sensor- O2 sensor.
3) The load sensor-MAF
4) The coolant temp sensor (only until roughly 95-98, after that CTS is almost useless for fuel control).
5) TPS-used as an anticipator of a load. has little effect until WOT
6) Idle switch-effects idle fuel control

Give us the rest of the readings and/or try to specify the cylinder that is acting up.

This comes from a guy who does this for a living..

 

[skelly]

Nice bit of info there - I've never used that approach to troubleshooting before (injector isolation). I also wouldn't have guessed that the cylinder with the least effect on the RPM once shut down would be the culprit, but it makes sense now that you mention it.

I have the emissions numbers from my last SMOG test right here:

900RPM, unloaded
HC: 709 (MAX=120; Gross Polluter @ 270)
CO: 0.27 (MAX=1.00; Gross Polluter @2.50)

2500RPM, unloaded
HC: 1038 (MAX=140; Gross Polluter @ 290)
CO: 0.80 (MAX=1.00; Gross Polluter @2.50)

There are numbers for CO2 and O2 as well, but they said those numbers aren't used for the pass/fail because of being unloaded.

Thanks for the in-depth information!
- SK

 

[ddavisaf]

I stand corrected.

I never would have thought about isolating the injectors. My brain hurts from all the new information I have just ingested. Thank you akdsmer

 

[akdsmer]

900RPM, unloaded
HC: 709 (MAX=120; Gross Polluter @ 270)
CO: 0.27 (MAX=1.00; Gross Polluter @2.50)

2500RPM, unloaded
HC: 1038 (MAX=140; Gross Polluter @ 290)
CO: 0.80 (MAX=1.00; Gross Polluter @2.50)

There are numbers for CO2 and O2 as well, but they said those numbers aren't used for the pass/fail because of being unloaded.

- SK[/QUOTE]

You can figure out the lambda (air/fuel ratio) by using CO, HC, O2 and CO2. This will accurately tell you if you are rich at 2500 (that's what the CO numbers are telling us). If you give up the rest of the values, I can help you out. Looking at what you provided, you are not in a rich misfire, nor are you in a lean misfire (especially at 2500). I'm betting on a vacuum leak along with a weak cat (cat not the source of the misfire, but damaged by the amount of heat it has developed due to the large amount of raw fuel available to convert).

 

[skelly]

I just had a breakthrough with this car today. It's been one problem after the next over the months, but I finally got rid of the stutter today. I have to get a timing light on it to validate the changes, but the steady running of the engine now is proof enough that it's running better.

So what happened?

Well, on a tip from a friend who said that his Honda Prelude once failed smog as a result of a bad knock sensor, I rushed off to get a new knock sensor. see the first time I removed the intake manifold, I was not aware that the knock sensor was installed in such a way that the wire would rip out of the sensor upon removal of the manifold without careful pre-extraction. In truth I didn't even know it was back there and all the working bits and pieces were under such a coating of dirty, oily GUNK, that it was impossible to tell.

So after kicking myself over damaging the sensor, I wasted about an hour of my life digging all the filler out of the sensor, reattaching the wire internally and filling it back up with high-temp silicone RTV. Since I had been running on this screwy, half-baked sensor repair, I figured that it was very likely that it was a botched job that wouldn't work.

So today I set about replacing the sensor. I removed the TB's intake tube, the battery and tray, reached down under the IM and felt around for the sensor, sized by big crescent wrench to the new sensor and plunged in to remove the old one.

As I was wrenching on it, I kept hearing this "tink tink tink" sound every time I moved. I looked for the source of the sound not expecting anything to be making a sound like that down in there and discovered my EGR valve hanging by a few threads from a single bolt!

I thought surely that was the problem all along - that I had neglected to torque down the EGR's mounting bolts, that one of them worked its way out while driving and the other was on its way. I removed the remaining bolt, got a pair of matching replacements at the hardware store, and fabricated a new gasket out of a scap of tin because the original had blown through via the escaping exhaust gases.

I was hyped, certain that this was the problem, even though it meant that I had wasted nearly $100 on the new knock sensor. I put everything back together, fired her up and the problem was STILL THERE!

I was frustrated at that point, but still not ready to give up because I have yet to get my O2 sensor wired in correctly - it's running without at the moment. So I just let the car run there for a bit, stumbling and stuttering, pondering about what the problem could be, going over a list in my head of all the things that I had already addressed. I placed my hand on the intake manifold and leaned on it so that I could feel the vibrations and motion of the engine while running. That's when it dawned on me: my original hypothesis had to be right.

From the bucking of the engine and the sound of the exhaust note in time with the stumble, the effects I was experiencing were definitely intermittent mis-fires. No doubt about it.

Now I will reiterate that I have replaced every single dang ignition component: plugs (several times, all perfectly gapped), wires, coils, and ignitor module with no result. I also set the timing with a timing light exactly as specified in the Haynes manual for this car. this was confirmed by two separate technicians during the smog testing process as well.

But today I said "screw it!" and loosened the two bolts strapping down that ignition timing deal near the TB while the engine was running. I rotated it counter-clockwise and the stumbling worsened, nearly stalling out. Quickly I reotated back the other way, up to and past the point where it had been set and wallah: problem solved. Perfect, smooth idle and revving with no sign uf a stutter/stumble.

So I have to put it on a timing light now and make sure it will still look right to a SMOG tech, but I'm much relieved to have been validated after all the time & money I've put into it. Now I have only to repair two minor things now and this car will be back on the road.

- SK