PDA

View Full Version : No "Check Engine Light" but suspect bad cat or rear O2 sensor?



jnshk
05-13-2012, 09:52 AM
My car is a 1997 OBD2 California Emissions model.

Currently, there are no DTCs and the MIL is not on, however after considerable driving there are three incomplete emissions readiness self-tests: Catalytic Converter, Oxygen Sensor, and EGR. I can only have a maximum of two incomplete to pass annual emissions inspection. I would like to get all of these items taken care of but currently I am focusing on the cat/O2 and intend to address the EGR later. (I did remove and clean the EGR valve, so perhaps that will take care of that self-test, but regardless I will want to address the cat/O2 situation.)

I have been monitoring O2 voltage through my ScanGaugeII and the rear O2 sensor exhibits the following behavior:

- When at operating temperature and cruising, the rear O2 sensor measures 10% voltage or less, fluctuating between 5% and 10%, roughly in synch with the front sensor (generally reading between 10%-90%).

- When given full throttle and going rich, rear sensor tends to read 85%-95% voltage while front sensor reads 100% voltage.

- When decelerating in gear (overrun fuel cut) and going lean, both sensors read 0%.

Obviously, neither item is far enough out of spec to cause a check engine light, but since I'm not sure exactly what sort of voltage the rear sensor should be producing under normal circumstances, I'm not certain whether the sensor itself is just going bad or whether these readings are indicative of a bad cat. Given the age of the equipment, it could easily be either one, but I'd prefer not to replace either one unless I'm sure it is necessary.

Can someone who is well-versed in these things shed some light for me? Thanks!

Follow-up question: If the cat is bad and needs replaced, will a Federal Emissions cat work just fine or is the California OBD2 computer specifically calibrated to expect a California Emissions cat? Is the Cali cat simply a local requirement (which would not apply since I am in Texas) or does the computer actually require the stricter emissions equipment to complete its self-tests?

revlimiter
05-13-2012, 12:11 PM
Clean your EGR ports. Seriously. That can cause the NA OBDII computer to not run checks on the cat and O2 sensor.

LOL
05-13-2012, 05:35 PM
First off, do you know the drive pattern requirements for each of the three incomplete monitors? If not I would advise to look in the FSM for each and try to replicate them in order to run its respective monitor.

jnshk
05-14-2012, 12:14 AM
Rev - I cleaned the EGR valve itself the other day and also ran a pipe-cleaner brush into the EGR ports on the manifold. They were very dusty. The valve itself was still rather "cruddy" looking inside after I was done, so I may just replace that anyway. This evening I used an air compressor to blow air through the EGR valve via the boost sensor vacuum tube. I blew it into the exhaust side initially and then manually opened the valve to allow air to blow into the intake manifold side as well. Hopefully this will have cleaned out the ports sufficiently?

LOL - I have indeed performed the recommended Mazda drive cycle multiple times, and in recent days I've focused on driving patterns that should clear the EGR and Catalyst tests. So far to no avail, but perhaps after cleaning out the EGR valve and ports things may resolve after a few more drives.



What has me confused/annoyed is that all indications are that both the cat/O2 system and the EGR system are essentially functioning as designed. The EGR opens with application of throttle, and if the valve is opened manually at idle the engine stumbles and wants to die. There is no obstruction between the EGR valve and the EGR Boost Sensor. The only thing that I can think of is that something in one or both systems is slightly beyond spec, which is why the tests won't clear but why it isn't extreme enough to cause a check engine light. Perhaps the EGR Boost Sensor itself is bad, but would that not trigger a check engine light? Anyone know what kind of voltages the ECU looks for on the rear O2 sensor under normal circumstances?


Also (for when the time comes later, if not now) whether or not the ECU itself expects the efficiency of a California catalytic converter or whether a Federal emissions converter would suffice?

LOL
05-14-2012, 12:18 AM
Interesting, is there other requirements needed for the monitor to run? For example an idle signal or any other ECM input? Reason I ask this is because I had the same issue with two monitors not running on a 2006 Tundra, turns out one of the requirements was an idle signal and at idle the tps contacts that relay idle signal to the ECM were open and never showed a signal therefore not allowing the monitor to run.

jnshk
05-14-2012, 12:23 AM
Interesting, is there other requirements needed for the monitor to run? For example an idle signal or any other ECM input? Reason I ask this is because I had the same issue with two monitors not running on a 2006 Tundra, turns out one of the requirements was an idle signal and at idle the tps contacts that relay idle signal to the ECM were open and never showed a signal therefore not allowing the monitor to run.

That's a very good question--and one that I wish I knew the answer to. If I had a flowchart of what requirements needed to be met and when (as well as any dependencies) for each self-test, that would help me considerably in tracking down this problem.

revlimiter
05-14-2012, 08:40 AM
The drive cycles are a pain. And that early OBDII ECU is a bigger pain. I hated it. I too had these same problems when Sharka was a 97. I couldn't get those last 3 systems to set ready. I drove over 1000 miles and then lost the car in an accident before solving the problem. I vowed to never have an OBDII NA again.

Every time I've come across a thread like this on m.net, I've filed away the solutions in my brain. Most of the time it's a dirty EGR sensor or passage not letting the system work correctly. Sometimes it's a bad seal around the EGR. Seems to always be the EGR. I doubt the cat is changing anything. A healthy cat is a healthy cat, regardless of CA or 49 state status.

If I remember right, Miatas like to run these tests at light load between 45 and 55mph in 5th gear. If you can cruise like that for a while (a long while), you might get it to finally work.

jnshk
05-14-2012, 07:01 PM
Yeah, thankfully that portion of the drive cycle is easy for me to handle. Most of my daily driving around here is between 45-65mph steady state cruising. I'm leaning toward EGR as the most likely cause as well, since it looked so cruddy inside. I have my doubts that it is always sealing adequately. I am also slightly suspicious of the EGR boost sensor, as I don't think it would trigger the MIL if it fails, but rather it would be a silent nuisance. If I had more time, I'd try to isolate everything and go one at a time, but since I need to get this done ASAP, I've ordered a new EGR valve as well as the EGR boost sensor ($$$!). Also getting a new rear O2 sensor for good measure.

Hopefully this will resolve any potential issues. ::fingers crossed::

revlimiter
05-14-2012, 07:38 PM
When I had my coldside SC, I was still on the stock ECU with a ton of bandaids. I could get the computer to set the ready state just fine, and my boost sensor was hanging out in the open. It just needed to see a pressure different than the manifold vacuum. Ambient pressure took care of that one.

=/

jnshk
05-26-2012, 05:18 PM
Figured I should follow up on this with the solution:

"EGR Boost Sensor" (also known as a barometric pressure sensor or map sensor)
Standard Motor Parts p/n: AS129 (NAPA p/n: CRB 2-16479)
Delphi p/n: PS10007
Borg Warner p/n: EC1723

I replaced the rear O2 sensor first, and as expected the readings were identical to those of the original sensor. I replaced the EGR valve itself (which was functioning normally as expected as well, but I do feel was justified based on the condition of the old valve, even after cleaning it up). I cleaned the EGR ports. None of these things affected operation or available sensor readings, and therefore the self-tests would still not clear.

Last night I replaced the EGR boost sensor (which I had some difficulty obtaining, as most places are currently out of stock), and this morning all the tests cleared within ten minutes on my way to work.

Apparently the only function of the EGR boost solenoid is to report to OBD2 computer for emissions readiness status. It does not appear to affect the operation of the car and a failed sensor does not trigger any sort of DTC/MIL/CEL. So if it fails, there is no outward indication nor any consequences unless you have computerized emissions testing.

Cliff notes: If you can't get EGR, cat, and O2 sensor OBD2 tests to clear --> and if car is running well with no symptoms or CELs --> and if rear O2 sensor readings are within expected ranges (fluctuating at 10% or less while driving and seeing up to 90% when full throttle rich) --> and if EGR valve appears to be functioning correctly ==> replace the EGR boost sensor first.

kung fu jesus
05-27-2012, 08:16 AM
Nice work and documentation. Thanks for taking the time to post your results and solutions!

For the record, a Fed cat works fine in place of a California cat. My wife's NB is a CA model and I replaced the exhaust manifold cat with a 49 (Fed) state cat for cost reasons and it passed CA emissions fine.