Hey everybody. Firstly, this is my first post here, and I just wanted to acknowledge how amazing this community is. I discovered the Civic VX from this site a year ago, and managed to find one that was in very good well maintained and surprisingly unmolested condition, with 182,000 miles on it. This 1993 Honda Civic VX was and is my first car.
Recently, I moved to California. For the last 3 weeks or so, I have been trying to get it to pass the emissions test here, in order to get it registered. Here is the emission test results:
I am not the most mechanically godlike person, however, I am good at learning about things. When my car first failed, I did a bunch of reading on various forums, and a whole bunch of learning about how my engine works (helped greatly by all of the awesome information on this site!)
The first thing I tried was to check and change the spark plugs. I bought a set of NGK ZFR4F-11, and replaced the ZFR5F-11 plugs that were in the car. The plugs were clean, no carbon buildup, and no excess fuel present. The tip ceramic was slightly pink. I took it back to get re-tested, and there was no change.
There is an intermittent check engine light that comes on, usually when the car is run at higher rpms for a period of time, or when it is under heavy load. For example, accelerating in 3rd gear at 3000+ rpm for 20 seconds or more going onto the freeway can trigger it. I read the CEL error code and it is a code 22, which indicates a failure of the oil pressure switch in the spool valve assembly. I troubleshot it to the best of my ability with the troubleshooting diagnostic procedures in the Honda Civic 92-95 Service Manual, not being able to test the oil pressure without "special tools".
Desperate and sad, I added 22ml of Acetone to the 3 gallons left in my gas tank, and took it in for another re-test. (I had read reports of using Acetone as an fuel additive reducing HC emissions by up to 60% and thought I would give it a try). There was no change in the test.
Finally, I decided that this problem was beyond my capabilities, and even in my desperate poverty, I took it to a well-regarded automotive smog shop that is a gold-certified smog repair facility (emissions testing laws in California are intense and complicated).
2 days later, I got a call back essentially saying "your car is confusing and we couldn't fix the problem, but we won't charge you money." I talked to the mechanic who worked on it, and it soon became apparent that I knew about as much about my engine as he did. He guessed that the problem was the VTEC engaging too soon, because of the extremely different %O2 readings between 15mph and 25mph. He said that the VTEC shouldn't engage until 3000 rpm. While I think the value is supposed to be somewhere around 2600 for the VX, and depends on a bunch of factors like gear, throttle, fuel-air mixture, and various other incomprehensible things, there is probably some truth to his diagnostic. They also said that the Catalytic converter was likely slightly depleted, because of the higher than average NOx readings, but that it should still be good, and it would not account for the extremely high HC readings at 25mph. They recommended that I take it to the dealer to get the VTEC mechanism troubleshot.
They said that they tested all of the obvious things, that is, ignition timing, O2 (LAF) sensor, tested for vacuum leaks, and so forth, and that everything checked out fine.
I have a vague memory of the guy I bought the car from telling me that he "adjusted the vtec down to 2300, which is where it is supposed to be". From what I have read, it seems that if you adjust the "VTEC Crossover Point" without calibrating fuel maps and timing (I don't really understand this well enough to talk about yet), it will seriously throw off the engine's tuning, and it can end up with the ECU telling the injectors to squirt way more fuel into the combustion chamber that can be burned, resulting in an overly rich mixture, which could result in a high HC failure in the emissions test.
The car itself runs fine. It gets between 36 to 44 miles per gallon (my best tank is 45 on freeway driving). There is a slight hesitation in acceleration that is noticeable sometimes at around 2200 rpm. It is very slight and ... non-violent, so I assumed this was related to the VTEC mechanism engaging. The car idles at about 1200 rpm when started cold, and eventually equalizes to about 900 rpm. Sometimes, after driving for a while, the idle will drop all the way to 400-500, but it never idles rough, or stalls. The engine burns about a quart of oil every 1000 miles.
With this evidence presented then, my question for you, kind compatriots of fuel economy questing, is:
How do I adjust the VTEC engagement point, and is this even possible? And more generally, what advice can you give me from your collectively vast VX VTEC-E knowledge that might help me pay less money to get my fabulous car to pass emissions?
Thank you greatly in advance, and I apologize for writing a novel on my first post! ;\
Interesting. Is it possible to contact the previous owner and ask how he adjusted it? There may be several ways to adjust
Thanks for the reply, theholycow. I have sent the previous owner an email, however I have not received a reply from him yet. I bought the car from him a year ago, so it might be a dubious endeavor getting a reply from him.
Do you know what the several ways that VTEC engagement might be adjusted are?
What might be causing my oil to be burning, thus producing high HC readings? I have heard that faulty head gaskets results in oil leaking into the combustion chamber, and that this results in blackish smoke coming from the exhaust when running. There is some smoke that is visible when accelerating violently, but under normal conditions the engine produces no visible exhaust. I assume a leaking head gasket would produce other more noticeable symptoms as well.
If the head gasket is leaky, more likely it will burn coolant, not oil. If it burns oil I would suspect the valve guides, seals, and/or piston rings are to blame. None of which would be easy to replace.
Usually the way the vtec engagement is changed is by swapping in an ECU that can be hacked though the use of a non-honda program, all of which would make the car fail emissions testing if the car was not tuned properly. I have not heard of anyone being able to hack the VX ECU so if the wrong ECU is installed, you need to swap in the correct one. I know there is some helpful link on this site to check to make sure you have the correct ECU installed by checking out the numbers on it. Other than that the previous owner could have installed a Vtec Controller which is basically a hack into the engine's wiring and in no way helps performance.
The VX ECU does engage vtec at a lower RPM than any other honda vtec system so your mechanic being surprised does not surprise me, since this is a strange motor for most mechanics to look at and try to understand without some advanced reading into how it works, or at least a glance into the VX shop manual.
On the never-ending quest for better gas mileage...
Definitely check which ECU you're running. It should be under a cover in passenger side footwell, just forward of the door. There should be a label on it with a bar code and a couple of numbers. One of them should be in the form of #####-###-###. The last six digits of that number can tell you which engine the ECU was meant for, which emissions specification (cali, federal, high altitude, ect.) and which revision it is.
Similarly, look at the front of the engine where it meets the transmission. There should be a raised square cast into the block there. It may be covered with grime, but there are engine model and serial numbers stamped into that square. The model should be D15Z1, but engine swaps are pretty common so check to be sure.
Remove the spark plugs and lay them down in order (so you know which plug came from which cylinder) and compare the color of the electrodes. They should have a light tan color with very little variation from cylinder to cylinder.
So I checked my ECU, and indeed it appears to be the correct Federal VX model. The number on my ECU is 37820-P07-A00-670-111097, which is consistent with the characteristic VX model listed here.
My engine definitely is a D15Z1, as it was one of the first things I checked before I bought the car! ;P It has D15Z1 stamped on it.
I have looked around in my engine, and although I'm no expert, I don't see any vtec-controller type modifications that look like they are tapping into the wiring that goes to the VTEC Spool Valve Assembly.
Page 5-7 on the USDM 92-95 Civic manual mentioned in my first post explains that the Variable Valve Timing and Lift Electronic Control System (This is what VTEC Stands for! I think) crossover point is controlled by the ECU based on a number of different variables, and should engage between 2500 and 3200 rpm, depending on these variables. The explanation is as follows:
So since the RPM that the HC failure in the test occurs is around 2300, which is under 2500, the engine should still be in lean-burn mode, unless something is messed up. The question is, what is messed up, and how to fix it? I wish I understood this better!
The sparkplugs that I took out of the engine looked pretty clean, and consistent from cylinder to cylinder, as far as I could tell. Here is a picture of them.
Same order, rotated slightly:
#4 on the right seems to be slightly more dirty... but is this really enough to indicate a problem?
Yeah, those look fine. A lot like the ones in my 'Z6 (I had them out a few days ago).
Well, I'm going to say fix the diagnostic code before diving into other stuff. The ECU can behave strangely when it's trying to compensate for a perceived failure.
Looking at the pressure switch diagnostic procedure, it has you check the wiring first, then go on to oil pressure/switch tests. Like you said, you need some extra hardware to do those tests, so I say skip them. Instead, look at the results of those tests: Either inspect the spool valve, replace the pressure switch or replace the ECU. I would inspect the valve first (p. 6-23) since it only requires maybe half an hour with a multimeter and a metric socket set. If the valve seems to be fine, replace the switch. If that doesn't fix the code, try a known-good ECU.
So I did some troubleshooting to the spool valve and oil-pressure switch mechanisms. Not having the equipment for the oil pressure tests, I tested all the electrical continuity stuff that I could, according to the VTEC Troubleshooting instructions on 6-18 to 6-22. Everything seems as it should be.
Following the directions on 6-23, I inspected the spool valve. The resistance between the 1P connector and chassis ground is about 16 ohms, which is right within the specified 14-30. I took apart the spool valve assembly, and checked the Spool Valve filter and the filter behind the oil pressure switch. It all looks clean, and appears to be functioning normally. The movement of the spool valve itself also appears to be working properly.
I decided to check the engine error code again, just for the hell of it, and realized that you are supposed to reset the ECU after reading the error code, which I hadn't done before. So, I look in the under-hood fuse box (for the first time), and I discover that the BACK UP fuse (7.5A) is blown. I replaced the BACK UP fuse with a spare 7.5A fuse from the other fuse panel, and drove the car for a minute, and it didn't blow again.
Could this blown fuse have been messing up the ECU's function? Or perhaps could it have been causing the problem in some other way? Maybe it's nothing, but it does seem a bit suspicious to me. Any thoughts on this new development would be appreciated!
Wow, that's strange. Yeah, if the back up fuse were blown, the ECU could have been resetting itself every time you turn the car off. Or Something... The ECU reset procedure calls for putting the fuse back in place when you're done. I'm not sure if the ECU would behave in a predictable manner if you fired it up with no power on the memory backup circuit at all. You had properly stored (persistent between resets) error codes, right? As in if you turned the ignition off and back on, the ECU would still display codes from a previous CEL event when you jumped the service connector?