drhawkinz

So I had a CEL code for P0420 (i think that's the right code) which said the cat converter wasn't outputting sufficiently. From what I read on the forum I figured I'd start with the o2 sensor and see what happens. So here's a side by side of the o2 sensor, obviously the old one is on the right:



My question is the white residue on the old sensor normal wear and tear or is that a sign that the cat is bad or going bad? The car drives better and certainly accelerates smoother than it did before so I take it the old sensor was bad. Also the new one is bigger than the old one but the car didn't throw any codes saying it was the wrong one so I'm assuming it's no big deal. Just thought it was odd.

Also if anyone is going to replace their o2 sensor you need a special socket. I didn't know this from the info I had read up here so just wanted to let you know if you're going to get a new o2 sensor you can rent an o2 socket set as well so make sure you get that. Looks like a regular socket just has a slit in it so it can fit over the wire.

ezone

Is the original sensor a DENSO or is it NTK?


I usually go straight for the cat when there is a P0420 code.
It is very rare to be caused by flaky O2 sensors.

bsmiley

I'm hoping the old one is one the left... the one on the right looks new... What year and configuration is your Civic? Which did you replace, the upstream or downstream sensor?
You bought a Denso, where did you get it (dealer)? How'd you know which replacement part to get?

I have an intermittent CEL for my O2 sensor. I believe the upstream one. From what I've read on here I should replace with NTK or Denso, whichever is the one in the car. I have an 01 LX. I can get either one for $33 (amazon). I was going to just live with the CEL once in a while, but you said it smoother so now I'm thinking I should replace it.

Did you rent the O2 socket from a local auto part store?

drhawkinz

I'm not sure. The only writing I can make out is OHC 501-H13 and a 2F12 on the part the socket fits onto. It seems to be the original one because it has a rubber boot on the cable to fit into a wire holder clamp.


yeah it's weird, the CEL was on for about a day and a half, I took it by autozone and they pulled the code. Then it was on the other day for about 2 days then went off. I think both times it had rained heavy the day before but trying to see what happens.

drhawkinz

2002 EX
Downstream
Autozone
The code was for the cat and this is the only sensor after the cat, well technically it's directly in the cat. But seemed it was the only one that made sense to cause a CEL.
Yes. Autozone does free rental so you pay $30 upfront but you get it back when you turn the tools back in.

ezone

Get a wire brush or wire wheel on that sucker and clean it off.

The name might even be on the thin upper shell.



Here's a pair I had out for whatever reason, one I hit on the wire wheel to clean it off

ezone

The code was for the cat, not any of the sensors.

drhawkinz

Found it, the original plug was a NTK. Does that matter?

Yeah but there's not an internal sensor for the cat right? I thought the o2 sensors measure the gas going in and coming out of the cat to determine if it works properly. From what I had read if the last sensor was bad it could say your cat was bad even though it was fine so it's a good place to start since it's cheaper. Usually cheaper I guess, not sure for other cars.

The car is getting ready to hit 150K. What's the longevity of cats? The back end of it looks rusted pretty bad but looks fine and if I bang on it it doesn't make a sound. I had heard if you bang on a bad cat you'll hear it rattle around inside.

Also would heavy rain or moisture affect it? It seems that both times the check engine light came on it was during or just after heavy down pour.

ezone

I guess you'll get to find out for yourself later on.
I figured that was probably why the new sensor looked so much different from your original.

When it comes to driveability and engine controls, I, along with many other professionals in this industry, have long recommended OE parts for many reasons, the most obvious among those is that they work exactly as the designers/engineers intended.







Just the O2 sensors.

O2 sensors measure Oxygen (O2) content in the exhaust stream, not fuel.

The primary sensor is (generally) used for engine fuel control, the secondary sensor is (generally) used to test the cat. The computer measures the cats' ability to store and release Oxygen, using those sensor readings.



A factory trainer once told us to think of an engine and all of the computer controlled systems that run that engine as merely a life support system for the catalytic converter. EPA regulations have made that cat the most important thing in the engineers' lives, and everything in the car revolves around getting that cat working as soon as possible, and keeping that cat working for as long as possible.
Emissions first, performance second.

Many people (some techs included) just don't understand how it all works.
Some people get to reading about how the cat is tested and figure since the computer uses the O2 sensors that they must be the problem. That's pretty rare these days, due to extensive on-board testing of the sensors....but not impossible.

The computer runs many self tests on the O2 sensors and must judge the sensors as "good" before it will run a test on the catalytic converter. If a sensor fails any of their tests and causes a code, testing of the catalytic converter is suspended.

There's actually about 15 other systems/sensors/areas that can cause the computer to suspend cat testing if they have caused fault codes.....Although a sensor that passes well enough to not set a code doesn't always mean that sensor is reading accurately.

(NOTE: I did not address false air/false O2 sensor readings from exhaust leakage here. It can happen.)




Flip side: Many can't justify spending big bucks on a factory catalytic converter so they get a cheapo converter that can't keep the computer happy (P0420 code keeps returning---but the cat is new so it CAN'T be bad, right?!?!), then buy sensors in the mistaken belief that they might solve the problem. And of course they buy the wrong (cheapo) sensors. It's a vicious cycle, circling the bowl.

Plus, you almost never hear that the ones who bought sensors first actually ended up having to buy a cat anyway. Nobody wants to admit that sort of thing.

About that long.

Actually, lifespan of a cat depends on many factors.
And several things can cause a cat to go bad prematurely.


Wow.

If the substrate or its suspension has come apart, then you might hear it rattle.

That's certainly not the only way a cat can fail though.

The cat CODE sets when the Oxygen storage capacity of the cat has diminished beyond a predetermined cutoff point.

This is usually due to the microscopic pores of the substrate being filled with contaminants.

I can't think of any logical reason for this.

The setting of a cat failure code in that car requires 2 failing test results in 2 consecutive trips in which the monitor runs. So it had to fail the self test during the trip immediately prior to your trip in the rain too.


-----------------------------------------------------------------------

My basic thing is to make sure the computer is in good control of fuel delivery, sensor readings look normal on a scanner datalist, and there are no exhaust leaks. Then I must trust the computer.

Occasionally, we may clear a cat code and let the car go, just to see how long it takes for the code to return. A cat in the early stages of failing a test may pass several more tests before becoming much more consistent at failing.

Figuring out what, if anything, killed a cat is usually another investigation entirely. Sometimes it's obvious (misfire problems, impact damage), sometimes not so obvious (fuel contaminants and additives, oil consumption, fresh oil flash-off, ethylene glycol poisoning, etc.).




Holy Carp that was long. I haven't even had any coffee yet.

drhawkinz

Guess I'll just have to ride it out and see if the CEL keeps coming on with the same code.

Hahaha. Thanks for the education and the help!

drhawkinz

check the mileage today and still getting about 34MPG so hopefully the cat is still in good working order for now.

bsmiley

I replaced my primary O2 sensor this week. I got the PN from the NGK's website http://ngksparkplugs.com/part_finder...t.asp?mode=nml
I got the sensor from Amazon for $23.
The engine is smoother and more responsive, but if you didn't drive it everyday you probably wouldn't notice.
The primary O2 sensor is used for a variety of things so I'm hoping that my mileage will jump up 2-3 miles a gallon. I'll post my findings after a few tanks of gas.

Reverb2005

Changed my O2 sensor a few months back and my comparison was identical to yours, other than the white residue. 'Technically' the NTK and Denso sensor work on an identical principle so there shouldn't be any problems with the new sensor. I haven't had any problems with my O2 sensor after changing.

I'm not a mechanic but I am a chemist and that white residue is a almost definitely a metal oxide or ceramic. White residues in extremely hot areas are not organic (related to combustion) and are typically oxidized metal compounds like aluminum oxide, magnesium oxide or titanium oxide, like what you would find in a cat, along with ceramic. Professionally, as a chemist, I would say that white residue is probably not a good sign since these oxides are most likely not coming from the gasoline or oil (or any combustion source) and probably from a metal source, like the cat, though not conclusively. Anyway, doesn't bode well for the cat .

Anyway, that's just my hypothesis.

drhawkinz

Thanks for the insight. The CEL is still intermittent although it's usually on more than off. Car still seems to drive good and maintains good mileage but replacing the CAT is on my to do list!

drhawkinz

The other morning it was extremely cold, around 20 degrees or so. I went out and cranked the car to let it warm up. After it had been running for 10-15 min I get in it to go work and see the CEL flashing. I tapped the gas and it went away, then I drove it around the neighborhood and it seemed fine. Drove it work and it was fine for a few days. CEL stayed on of course but wasn't flashing. Then the other day the CEL would start flashing while I was sitting at stop lights and I noticed the car had a little bit of a shake to it. I could rev the engine up to 1500 RPMs and the light would quite flashing, car would smooth out. So I figured the CAT was completely dead so I parked the car and ordered a new CAT. I got a Walker 82681 which is CARB compliant. Found it on amazon for $350, didn't want to spend the $700 for the Honda replacement.

So I replaced the cat today with the new one and the car runs better while driving but I still get the same thing while the car is at idle. The car shakes and CEL will flash. I know the flashing CEL means a misfire, I had driven to autozone last week before parking it and it pulled the P0420 for the cat and then a random cylinder misfire and then a misfire for each cylinder. I was assuming the misfire was from the clogged CAT but I don't think that's the case. Now I'm not sure where to start to figure what's causing the misfire. I replaced the plugs over the summer when I changed the timing belt. They are NGK Iridium which I think are OK to use from what I've read up here. Any suggestions on what approach I should take now?

ezone

Flashing CEL is a misfire.
A cat problem cannot normally cause misfire unless the substrate came apart and sent debris bouncing around. That's certainly not the typical cat failure.

Misfire can RUIN a cat in a hurry. Even a brand new cat. If it's running bad enough, it might take only a few minutes to ruin it.


Compression test
Valve adjust
Recheck compression
IF still a compression problem, then cylinder leakage test to figure out where the loss is going

Figure out why it's misfiring. I can't tell what it is from here.

drhawkinz

i noticed there was green clump of turd on the negative battery terminal. So took the leads off the battery and cleaned them good and put it back together. The check engine light was off for a bit then came back on solid. Took it back to autozone and they pulled a code for misfire in cylinder #2.

Rented a compression gauge and the compression is as follows: cyl#1 = 180 , cyl#2 = 80 , cyl#3 = 170 , and cyl#4 = 170. Next weekend I plan to do the valve adjustment procedure and then recheck the compression.

I just find it odd that the compression is that low compared to the rest. Would valves being out of spec cause that much of a drop in compression?

ezone

If a valve cannot close, what effect will that have on the compression test numbers?

drhawkinz

I guess I was thinking the car would start showing problems sooner.

I did a valve adjustment today. First time ... and it sucked. Seems like there should be a more scientific measurement than how much "drag" you feel on the feeler tool/strip. But it was a good learning experience although now the car seems a little more chatty. So I guess I need to do it again and tighten up some of the valves that I thought were OK.

I did a compression check again and as I suspected it didn't help the compression on cylinder #2.

Cylinder #1 was still strong with 180 psi, Cyl #2 at 90 PSI, Cyl #3 slightly improved to 175PSI and Cyl #4 was just under 180 PSI.

From what I read, I can put some oil into the cylinder and see if it raises compression. If it raises compression then it's the something with the valves, if it doesn't raise compression then it's the rings. I'm assuming I don't need to do this to every cylinder right, just the bad one?

When I looked through the spark plug hole I can see the top of cylinder caked with sludge. Hypothetically, if a valve was stuck open due to carbon deposits, would you be able to tell that from the valve adjustment procedure? I would expect the valve to not have any wiggle when at TDC for that valve but I'm just speculating.

Thanks for all your help man! I really do appreciate it!!

ezone

It takes some experience to recognize that feel.

At least on Hondas you can take your own sweet time....Try it on an engine that has to have the valves adjusted HOT, before the engine cools down! If you are slow and it cools off, you have to put it all back together and get it hot again to finish the valve adjustment.


Loose valves just make noise.
Tight valves don't make noise, they burn.

Nah, don't bother. See below.

What's the next step?

I suspect it's gonna be a burned valve.
Oil will artificially boost the compression readings by creating additional seal between the rings and cylinder walls.
Too much is too much though. Too much oil can give sky high readings and contaminate the tester.
Should only need about a tablespoon of oil per cylinder.

If you want to do it, don't do just the bad one. The test needs to be done for all cylinders so you have an idea of how the bad one should have acted. All with good rings should see a nearly identical % increase.

Then expect a lot of smoke when you get it fired up again.

Your spark plug should have looked dark also. Low compression leads to poor combustion, which leads to carbon deposit buildup.
Maybe, maybe not. Did you notice anything that would lead you to believe this?

You would have had to closely measure all clearances prior to adjustment. Carbon holding a valve open would have resulted in excessive clearance.
Wiggle? Huh?

drhawkinz

oh wow that sucks!! what kind of engine requires that kind of maintenance

is that something a DIYer could fix or does the engine need some machine work? i'll do some research and see what i can find on the procedure.

yep, noticed that

no, was just wondering if you could tell from a valve adjustment if maybe a problem was with the valves or cylinder rings.

yeah, since you have .008" gap between the rocker and valve it should have that small amount wiggle when moving the rocker. But like you said, if the valve is held open by a carbon deposit then you'd have a lot of wiggle room. I was thinking it was the other way and you wouldn't have any.

ezone

All the old Mazdas had to be done this way back when I worked with that car line. My truck is one of those.


Depends. Anyone with ambition and tools can get the head pulled, but I'd send the valve work to a machine shop.

No. This is what the cylinder leakdown test is for, to locate the source of the leakage.
Oh ok. The ROCKER ARM will wiggle up and down, not the valve.
If a valve can wiggle, you got big problems LOL.

drhawkinz

Wow. Hopefully it doesn't need an adjustment very often. I can't imagine the heat from the engine would cause that much expansion. Does it have to be done warm because of the valves/cylinders or because the way the rocker meets the valves?

Yeah I think I can manage that without panicking too much. What valve work is involved? I was thinking it'd be a discard and replace with new kind of procedure.

Haha, sorry, should have been more specific. Still working on my mechanic vernacular. This is a first for me for getting this in-depth into engine work.

bsmiley

For your research in prep of the head removal...
https://www.civicforums.com/forums/1...elt-diy-6.html

ezone

Their service interval was every 15,000 miles back then on most of the cars. And it got a fresh valve cover gasket whenever I did them.

I let my truck go 30k on this.


Back in that same era, Honda used to recommend valve adjustments every 30k IIRC. Maybe even 15k.
Oh, but heat does have great effect. I've seen it. A few thousandths of an inch may not sound like a lot, until it makes the engine run like crap or burns a valve.



That's just the way Mazda wanted it done. That's what their specs were.

Honda spec'd these to be set on a cold engine.


To make it sound overly simple:
Inspect sealing ability, vacuum check and/or liquid check
Disassemble
Inspect
grind valve if serviceable or replace with new
grind valve seat
lap
reinspect sealing ability
reassemble
final inspection

I'm overly technical, I have a real hard time just assuming anyone knows what they are talking about, especially when incorrect terms are used.

drhawkinz

Thanks! I ran across that when I was searching, gives good info for taking the head off. Oh wow, guess that's why it's not in the Haynes manual .... hahahahha

hahaha, I know what you mean. I do TV installs and people don't understand the difference between the TV being on channel 3 and the cable box being on channel 3. It's rather annoying.

--------------------------------------------------------

I have a few questions about the leak down test. I got a tester gauge from CarQuest today. I've watched a few videos and read a few how-to's to get everything planned out.

I'm assuming i'll get best results while the engine is warm. So my question for this regards taking the valve cover and timing belt cover off. Which of these would be better:

Option one:
take the bolts out of the timing belt cover. Drive the car to warm it up. Then take the valve cover off and timing cover off to make sure the cylinder is at TDC and then do leak down test.

OR

Option two:
take the timing cover off while the engine is cool, tighten the valve cover back down and drive the car so it's warm with the timing cover off. Then do leak down test.

I know if the "UP" on the camshaft pulley is at 3 o'clock then cylinder 2 should be at TDC. So if I can see that then I shouldn't have to take the valve cover off but wasn't sure if it would possibly mess anything up driving with the timing cover off.

And it seems in some how-to's I should just remove the one spark plug, in others I should I take all 4 out and then do the test. So which one is the appropriate method? I was thinking taking all 4 out because if air is seeping out of the head gasket into an adjacent cylinder then I can hear the air coming out of the spark plug hole.

I watched this guys video: and he has an air tube hooked up to his crankcase. Do these cars have such a tube? I read that if air is getting past the piston rings then I can hear air coming out of the dipstick tube and/or oil cap. Where should I listen to hear if air is getting past the rings?

johndeerebones

TDC the test cylinder, set the tester to it's test point. Green is supposed to be good. If there is excessive leakage you won't get it to it's test point. Make sure it is at TDC, if you're unsure yes you can pull the VC and loosen the rockers to make sure the valves are closed. If the intake valve leaks you will hear it out the intake, exhaust valve out the exhaust. Rings out the block. With that much difference a burned or stuck valve is likely. Usually a ring problem will be across the board (such as it ate dirt etc). This is not ALWAYS true though. As an aside while the valve cover is off this is an excellent time to check valve adjustment too. As an aside while cranking to do a compression check make sure you open the throttle. Not an insult alot of techs forget that and they should know better.


You've been there so more for others than the OP, but the upstream sensor controls the fuel curves as a general rule and the downsstream is for nothing but CAT monitoring. As ezone said, engine controls are ALL for emmissions (thanks EPA). With that said though general public and some techs are all to eager to point at electrical components. The fact is that the majority of problems are mechanical in nature. Also, O2 sensors measure heat. Engine runs lean -it is hot, rich (fuel or oil either one)-cold. Either one kills a CAT and sets a CEL. Fouling as in rich will ruin a CAT faster, this is why it is not smart in the least for anyway to drive a misfiring vehicle around, CATs are expensive, not to mention on most exhaust systems there are atleast two now days, v configuration engines have atleast 4. OBD 2 started in 1998, that was working on 20 years ago so you can imagine how "smart" the gov't mandates vehicles currently.

ezone

As low as the compression test was, it probably won't matter.
Only close. The word UP is not a timing mark, it's more like a generalization of the area.

You must physically verify the piston in question is truly at TDC, because there are no marks provided by the factory.

This is where I stick a screwdriver in the plug hole and watch as I roll the crank by hand.


If the piston is not within a degree or so of TDC, air pressure can cause the piston to be pushed down. Moreso when using straight shop airline pressure LOL.

Hope nothing falls in the open timing cover.





Removing all the plugs also makes it a helluva lot easier to spin the crank by hand (with a wrench), you aren't trying to overcome compression this way.

However, when chasing valve leakage it may be necessary to reinstall plugs on the other cylinders. If for example your bad cylinder has an intake valve leaking, the air flowing through that valve into the intake manifold is free to enter any other cylinder that has open intake valves.

When this happens, you usually can't be sure if the leakage is from the intake valves or exhaust valves, so the affected cylinders must be "corked" so you can determine the source by the usual method.

If you run into this, it's more likely to be from air coming into that cylinder through open valves. These don't normally blow between cylinders.

It's easy to stick a plug back in the hole to cork the airflow if necessary.


DNWTFV.


Using an actual leakage tester is not so important in your case, we already know there is excessive leakage on that one cylinder.

YOUR quest is to find where that leakage is going.

During this type of test you must be able to sense sometimes miniscule amounts of air movement.
All air movement (leakage) needs to be concentrated to a localized area..... so you can feel air movement with your nosehairs. Seriously.

Any leakage into the crankcase will flow upward through the oil drainback holes in the head. If the valve cover is removed, you'll never be able to tell. With the valve cover installed, any air movement MUST exit through the vent tube, since that's the only open hole. That makes it much easier to sense air movement, it's all concentrated in that hose.

Same for the intake. If the throttle blade is wide open you may not be able to really tell if there is little air movement coming through that great big opening. Throttle closed, even taped off with duct tape.....and ONE vacuum hose unhooked so any air movement must go through that hose. You can feel it with your nosehairs that way.

On the exhaust...... Much more difficult to tell when the leakage is small. Last one I had like this....I attached one blue Nitrile glove to the tailpipe with a rubber band. That made it much more apparent when there was a tiny amount of airflow.


























OP needs to be aware that there will always be SOME amount of leakage past the rings, and this alone cannot be used to condemn anything.

Besides, his wet comp test results did not support a ring problem IMO.



I hate to nitpick, but I'm gonna.
Downstream sensor does have influence under certain conditions.
Read code. Throw part. Ugh!
*drags knuckles back to the toolbox*


Not true. They DO need to be heated to do their job though.

That reminds me...Ever seen a cat ruined by running out of gas? I'm pretty sure I've diagnosed it.

Some manufacturers and/or models were OBD2 compliant in 95, but it was mandated across the board for 1996.
Looking back now, some of the ways they met regulations were pretty primitive. Auto and emissions technology has progressed a lot in the last 2 decades.

The same technology made a large percentage of the people in the field obsolete. But then, so did fuel injection. So did electronic ignition. And so on, and so on.

drhawkinz

Good idea. I had to use a coat hanger since I didn't have anything long enough and small enough to fit through the plug whole. I guess it's probably best to do this during a valve adjustment as well.


Yeah I could feel the air coming back out of the other cylinders so I put the plugs back in and could feel the air coming back out of the air intake.

??? Do Not Watch This F#@$% Video ? haha



Yeah when I bought the leakage tester the clerk had said a local engine shop had come in looking for it but they never came back to buy it. Is this typically used for more high performance engines to help tone and tweak the valves for maximum performance?

I noticed afterwards that the air compressor had a regulator on it. I guess I could of just used that and set it to 30-40 PSI and then do the same thing since I'm just listening for the sound and not concerned about what % of loss I'm having.

Quote:

Originally Posted by ezone

YOUR quest is to find where that leakage is going.

Seems pretty apparent it's the intake valves. Which also seemed the hardest to get loose when I did a valve adjustment.
[youtube]https://www.youtube.com/watch?v=eZhJI9mmrSM[/youtube]Now it's time to start ripping this POS apart. Thanks for all your input and advice! I really do appreciate!!

ezone

Ayup, that appears to be one heckuva leak you have there.



You mean the clearances were too tight, or were you referring to the locknuts being tight?


"Tight valves burn" refers to lack of clearance.

'Did' not. It's a Farkism, akin to DNRTFA (did not read the ******* article).