Civic VX Emissions Failure - VTEC Malfunction, or Hopelessly Epic Melodrama?

[badur]

Originally Posted by bobski

http://en.wikipedia.org/wiki/Catalyt...rter#Poisoning

It could have been a simple matter of the fuel trims getting reset.

The bold above was just meant to make the question stand out of the rest of my giant blocks of text, not meaning to yell

I suppose it's definitely possible that in its past, the engine has had Catalyst contaminating substances put into it. Could you maybe explain what fuel-trims are exactly, in simple terms? I understand vaguely that it is function of the ECU that determines fuel mixture... but nothing beyond that. It was definitely some permanent change that happened, because the car has been driven about 800 miles since the ECU has been changed, and the emissions still test with lower HCs).

Originally Posted by bobski

Looking at those numbers, I think you're getting incomplete combustion of some sort. You've got both excess hydrocarbons and oxygen in the exhaust stream. You're sure your cam and ignition timing is set right? The proper heat range plugs? I guess if the EGR valve was stuck open, it could cause combustion issues.

I really don't know about my Cam and Ignition timing. I was assuming they tested the ignition timing as part of the emissions test, and I don't know how to go about checking the cam timing. How would one go about doing that?

I know my spark plugs are the correct model and heat-rating (NGK ZFR4F-11) (see my third post).

[GasSavers_bobski]

Originally Posted by badur

Could you maybe explain what fuel-trims are exactly, in simple terms?

The ECU uses stored tables (like a spread sheet) to decide how much fuel to inject at any given moment. Going down the table, you have engine speeds (RPM) and across the table, you have manifold pressure, read by the MAP sensor. Once the ECU picks a base value from this table, it adjusts it based on input from other engine sensors - coolant temperature, air temperature, oxygen sensor readings and such. The fuel trim is just a set of adjustment values the ECU stores and tweaks over time, based on sensor observations. If the ECU has to constantly lean out the mixture due to the oxygen sensor reading, it could adjust the fuel trim to compensate. What use does that have? If you've ever watched the reading from a standard narrow-band oxygen sensor on a engine with a trimless ECU, it's constantly swinging back and forth between lean and rich. If the ECU only uses the oxygen reading to slightly tweak the trim value, it will smooth out that swinging effect, zeroing in on the middle, average value, resulting in the proper (stoichiometric) air/fuel mixture.

Originally Posted by badur

I really don't know about my Cam and Ignition timing. I was assuming they tested the ignition timing as part of the emissions test, and I don't know how to go about checking the cam timing. How would one go about doing that?

I thought they checked it as well, though I don't recall seeing any specific numbers. To check the ignition timing, you need a timing light. A high quality light will set you back anywhere from $70 to $200, but you should be able to find a no-frills light for $15-40. The light is just a strobe light with a sensor you clip over your no. 1 (closest to the timing/accessory belt end, furthest from the distributor) cylinder's spark plug wire, which triggers the strobe when the plug fires. Once the light is hooked up, you start the engine, let it warm up until the radiator fan kicks on once and then point the light at your crank pulley (the pulley that spins the alternator and A/C compressor drive belts). If you look closely, you should see 4 little notches in the edge of the pulley; One notch off on it's own, and 3 in a group. The 3 in a group are for ignition timing, and the one off on it's own is the top dead center (TDC) mark... We'll get to what the TDC mark is used for shortly. Really, you can find these notches while the engine is stopped, but each strobe flash should occur when the marks are near the top of the pulley where you can see them.

Molded into the timing belt cover near the pulley, you should see a notch-and-pointer type sight. If you look down the sight, the pointer should be aimed somewhere in that group of 3 timing marks. The center mark is the ideal timing setting (16 degrees or 18 degrees or whatever), and the two marks to either side are 2 degrees advanced or retarded, which is the acceptable range of adjustment. Advancing the timing (the mark further from TDC) will shift the power band up the RPM range - make the engine more efficient at high RPMs, while retarding it (the mark closer to TDC) will have the opposite effect. Since these tests are taking place at low RPMs, I would adjust it to, or at least near the retarded mark.
Adjustment is a matter of loosening the distributor mounting bolts and rotating the distributor while watching the pointer and timing marks. Keep in mind that this is while the engine is running (loosen the bolts just enough so you can turn the distributor), and the ignition is a high voltage system - grab the distributor by the plastic cap just to be safe.

Cam timing (also known as static timing) is the orientation of the camshaft relative to the crank. The crank and cam are connected by the timing belt, which has teeth unlike the alternator or A/C compressor belts. The crank and cam gears (or sprockets, or whatever the correct term is) are different sizes - in fact, the cam gear has exactly twice as many teeth on it compared to the crank gear. This makes the camshaft spin at exactly 1/2 the speed of the crank. Since these are 4-stroke engines, it takes two crank revolutions to complete a full combustion cycle (intake, compression, power and exhaust strokes). During those two crank revolutions, the cam only rotates once.
Cam timing problems occur when either the cam gear or crank gear slip a tooth or more, mis-aligning the two. It's very easy to similarly mis-align the crank and cam when changing the timing belt, if you don't intentionally set the two to TDC.
The distributor is driven by the far end of the cam shaft, so bad cam timing will throw the ignition timing way off (you can't turn the distributor far enough to get it right, or turning it all the way in one direction just barely gets there). If you don't see this when checking the ignition timing, your cam timing is probably fine.
If you really want to check it, turn the crank by hand counter-clockwise until the TDC mark lines up with the timing belt pointer. Now, remove the upper timing belt cover (two bolts on the cover itself, pull the spark plug wires out of the plug wells, unbolt and lift the valve cover and pull the belt cover away). You should be looking at the face of the cam gear. It should have an "UP" marking on it... If UP isn't UP, turn the crankshaft a full rotation counter-clockwise until the TDC mark lines up again. The cam gear should have various lines stamped into its face.
Here's an example of an '88-91 gear showing the marks used with the D15B2/D15B7 engines:

For the VX engine (D15Z1), you want to use those same lines on the face of the gear, but they should be in alignment with a pair of pointers molded into the back half of the belt cover (the plastic part behind the gear), rather than the valve cover mating surface.

[caccox]

Just to clarify, I was referring to ignition timing NOT cam timing. bobski did a great job explaining both. Your cam (or valve) timing should be fine. As it is not adjustable the only way you could have screwed up your cam timing is by reinstalling your timing belt off one or more teeth when you (or a previous owner) removed the head. Although I have never had the misfortune of a cam out of time, I expect you would be way down on power, get terrible mileage, and have way worse emissions numbers, and poor compression. If your cam timing was off you would know it.

So back to ignition timing: Yes they do check timing as part of the smog check in California. On OBDI cars, it is a manual inspection (just like the gas tank pressure test). The tech takes out his timing light and verifies the timing is set between the two marks on the crank pulley. However, Ive never seen them take more than 10 seconds to check it. In my experience, the marks are really hard to see if you dont first mark the pulley, so I think the tech is probably just going through the motions. What I do is bump the engine over until you can reach the marks from above or below. Then I very carefully mark the 14 degree and 18 degree lines (I looked up the spec, it is 14 - 18 BTDC, but the actuall numbers dont matter - you just use the marks on the pulley) with white out (try not to get the bottle dirty so your girlfriend, wife, mom, etc wont notice when you put it back). Then you can set it as described by bobski. (You should have a new cap, rotor, plug wires already if you are trying to pass smog so you wont get shocked - If you do get shocked due to a bad wire or something your arm will hurt for an hour). For emissions testing you want it as retarded as possible. Advanced timing is more efficient (but produces more emission - HC at least, I have a chart of all three emissions vs ignition timing, but not in front of me so I forget what happens to the other two - I always had HC problems, so that is why I remember it. As a general rule, smog check = retarded timing). Once you pass the smog check you set it as advanced as possible for better power and mileage. Either you set it to the advanced mark, or until the car knocks and pings under load and then back off.

On OBDII cars, both the ignition timing and gas tank pressure test are monitored by the ECU. It stores a code if either are off, so all the tech has to do is verify there are no codes stored in the ECU - hence the tech does not perform a timing test on new cars.

[GasSavers_bobski]

Originally Posted by caccox

Although I have never had the misfortune of a cam out of time, I expect you would be way down on power, get terrible mileage, and have way worse emissions numbers, and poor compression.

I haven't spent that much time familiarizing myself with the attributes of out-of-time engines... The few times it has happened, I've gone for a test drive and decided "this isn't right", pulled back into the garage and corrected the problem. What I have observed is that being off a tooth tends to make the power band peaky... For instance, you'll get very little power at low RPMs to the point of it being difficult to get the car rolling, but then it will take off like a rocket once you break 4500 RPM. I'm not sure how much messing with the ignition timing would change that or if it would make the car drivable.
Based on my experiences, I assume milage would take a hit, as would emissions. I don't think it would show up much on a compression test since scavenging effects (for which valve timing is crucial) don't occur without combustion events.

Originally Posted by caccox

You should have a new cap, rotor, plug wires already if you are trying to pass smog so you wont get shocked - If you do get shocked due to a bad wire or something your arm will hurt for an hour.

I'm more concerned about the distributor's ground path. The distributor wiring does not include a ground wire... Three pairs of sensor wires, an ignition trigger signal, power from the ignition switch and a tachometer signal, but no ground. The ground path for the distributor components (including the high voltage coil) is through the distributor body and cylinder head. If you loosen the bolts holding the distributor body to the cylinder head, it's possible (though unlikely) to interrupt that ground path. You probably wouldn't take a full discharge from the ignition system (your body becoming the ground path), but you might get zapped due to capacitive effects (your body stores the electrical charge, rather than conducting it), which I can tell you from experience is still something worth avoiding.

[carpediemracing]

FYI I was told my cat was bad on my VX, and I got one off some eBay merchant for a reasonable price. $185? Fit right in. If this is still a concern I can look up the merchant name.

I'm in the same boat with failing emissions (1500+ HC), then some hesitation which prevented a retest.

cdr

[badur]

A couple days ago I had pretty much given up hope, and was ready to sell my lovely car, and buy a Civic HX coupe that was already registered in California. There were still a couple of things to try though.

Firstly, I bought a Timing Light. I checked my ignition timing (thanks to bobski and caccox's excellent explanation of the process). It turns out, my engine is adjusted to about 14 degrees, which should result in the cleanest emissions. So this being adjusted improperly definitely wasn't causing me to fail.

Next I wanted to try reading the voltage coming out of the O2 sensor, to make sure it was functioning properly, as described here.

I also found a guy selling a CA version P07-L00 ECU on craigslist for $35. Swapping in the in the replacement P07-A00 ECU is the only thing that I had done that had any effect on the emissions readings (it dropped the HC @ 25 by about 25 points). I figured it was worth a try, even though caccox indicated that disabling lean-burn in this fashion would likely only affect NOx emissions output.

It turns out that disabling lean-burn dramatically affected the HC emissions as well (at least in my car ... who knows if there is some other problem that my other FED ECU is causing). Here is the emissions test results that I got today, with the CA model ECU. ( Yay!!! )


I also recorded some video of the O2 sensor output while driving from a cold start, with the CA and then the FED ECU hooked up. There are some alligator clips hooked up to the D14 and D16 wires. I just carved a bit of the casing off each wire, and taped it back up when I was finished. A better technique might be to stick a couple of needles into the socket and hook alligator clips to them.


Here is the video of the Federal model ECU's O2 sensor output:
http://www.youtube.com/watch?v=9S_CFX5XOKo

And here is the one of the CA model ECU's O2 output, with lean-burn disabled.
http://www.youtube.com/watch?v=8LaIC-y0_5E

The CA ecu spends a lot more time in negative voltage land, than in lean-burn positive voltage land, when accelerating.

I hope this is massive thread is informative or useful to someone! I know I've learned a lot reading through these forums, and I am happy to contribute back.

Thanks so much to everyone who helped me, and I'm glad there is a happy ending! Tomorrow I head to the DMV to get my car registered.

[badur]

Originally Posted by carpediemracing

FYI I was told my cat was bad on my VX, and I got one off some eBay merchant for a reasonable price. $185? Fit right in. If this is still a concern I can look up the merchant name.

I'm in the same boat with failing emissions (1500+ HC), then some hesitation which prevented a retest.

cdr

Thanks! 185 is cheaper than most places I came across selling the manifold-mount cat for the vx, but it looks like I don't need one any more.

1500 HC sounds insanely extreme. My HC at worst was only 119. Good luck getting your car to pass!

[Farlander]

Badur,
My first thought reading your post (thank you by the way) was that your oil consumption problem was clogging your LAF sensor, causing the fuel ratio to be way off. I have read that burning oil will quickly kill these air/fuel ratio sensors... anyone have any info on that?

I too have a VX and failed my first attempt at emissions today with 247ppmHC, limit 183,, and 1.91%CO, limit .73%. I have a check engine light on for code 48, LAF sensor, which I plan to replace soon (already have the new one) after I swap the coolant, thermostat, and change the oil to high mileage engine oil to try and swell the valve seals, since I don't want to kill my new LAF sensor by burning oil. I will also try running the test again with the sensor unplugged, which I assume would put the system into open loop mode as well, eliminating lean burn, thus saving me having to buy a new Cali model ECU... correct?

Interestingly, at 15mph, the car failed miserably, but at 25 mph it passed with flying colors.. I have all stock engine and trans (to my knowledge) with 175/65R14 tires.

[Farlander]

Ok, today I went in for my restist, four new NGK plugs and a new LAF sensor and I passed like the engine wasn't even running. Now I have this funky oscillation at idle though, revvvvv die, revvvvv die, revvvvv die...

[add|ct]

It sounds like it could be your IACV(Idle Air Control Valve), it probably just needs some cleaning and it's worth a shot since these tend to need cleaning once every couple years at times. There are how-to's on this online, someone has it linked in their sig, but otherwise check it out. It's not a hard job to do on the VX at all. Once you get an idea of how to remove-reinstall, the only thing you have to worry about is cleaning it out once removed.

Get some throttle body cleaner and go to town.

Link on how-to:

http://www.honda-tech.com/showthread.php?t=1575913