AAGH! CEL back after 90 miles!

Having received a shiny new driver's side cat under warranty to finish my previously reported CEL saga , I picked up the car yesterday morning and went merrily about my business. Until last night, when the CEL came on again. Same circumstances, about 10 minutes after leaving the freeway in Redondo. I'm going to have a serious chat with my car tonight, to let it know that Troublesome Cars Get Traded. I may leave some 350Z/M3/RX-8/S2000 brochures on the front seat to encourage better behaviour... Seriously though, I am wondering if my driving style has anything to do with it: on the open road I'm as gung-ho as the next guy, but in town I tend to surf the torque curve at 2000-3000rpm. I've noticed that a lot of other rennlisters claim to hit 4000+rpm in each gear, but I must confess that I had put it down to a) hyperbole b) showing off by making lots of noise c) some people being new to driving a stickshift. I humbly apologise (especially for c), and request my fellow rennlisters' input: is it unhealthy to run a 996 in the 2000-3000rpm range?

no

it wont cause a cel..... its probably another sensor.. .are you on your original MAF?

So far as I know, it's the original MAF. I picked up the car in May with 51,000 miles, and it's now on 64,000. It did get a new RMS just after I bought it.

Have you checked the code? If so, what code is it?

I haven't had a chance to check the code yet. The last one was P0420, meaning a sick driver's side cat. When AMS in Carson looked at the O2 values on the other side, they were a little off, but that cat wasn't dead yet. I probably won't have time to check the code this weekend, but I've arranged to take the car for another trip to my dealer on Tuesday. Meanwhile, the car shows no sign of listening to the telling-off I've given it, but maybe it's thinking about it

When I had a bad cat on the driver's side my dealer replaced both cats.

Has it behaved itself since then? My confidence in my car as a daily driver has been dented a little.

I've had no problems since both cats were replaced.

cats dont fail too often but they do fail

I wonder if running the load factor you ran the engine at for the amount of time that you spent in the open road challenge burnt up the catalytic converters? I also wonder if changing the DME control unit programming record to the european fuel scheduling would prevent this?

clean the MAF... mine went at about the same mileage.

It's a fair question, but I'd be pretty disappointed if a German car couldn't handle 45 minutes at speed - when I lived in Europe I took most of my cars through Germany at one time or another, and even my unreliable MG coped just fine with a few hours at 120+.

the only way the engine controller has of keeping tabs on converters is by the O2 sensors and a bad (even intermittent) O2 sensor can appear as a converter problem. To be better able to id the culprit, O2 sensor or converter, you need to monitor the O2 sensor voltage levels. What has worked for me in the past is to get an OBD-2 code reader/data viewer and monitor O2 sensor voltage levels -- ideally you want a reader that can display at least all 4 O2 sensor readings simultaneously. I find this very informative to be able to compare the sensors on the two separate banks. A mis-behaving sensor, though its behavior is probably intermittent and will remain so for many more miles -- stands out when viewed this way. Unlikely hard driving has caused the converters to burn out. (150,000+ miles on my '02 Boxster's original converters, and while I've never driven it on the track, it has seen some extended high speed driving out in the great southwest. I might add that one O2 sensor (#1 sensor on the passenger side bank) in that time has needed replacement -- at around the 80,000 mile mark -- and I believe one or both the sensors on the driver's side is getting close -- every once in a while engine light comes on and its usually "converter performing below efficiency" or something to that effect but when I look at the O2 sensor voltages, they indicate the sensors are not responding with their usual vigor, which is usually how an O2 sensor goes wrong -- aging causes its response time to get slower and voltage levels less. Other check engine light error codes do flag the sensors so I'm going to replace the suspected ones before I replace the converters!) The converters can burn out if the engine's running rich, but the engine controller should catch this condition and the check engine light will flash. Only if you continue driving then could the converters be damaged. Sincerely, Macster.

ok............ your o2 sensors moniter cat efficency by comparing the pre cat o2 sesnors to the post or after cat o2 sensors. A faulty 02 sensor will usually throw a fault for o2 sensor ageing (or delay) if it becomes slow responding, if it reads out of range it will also throw a fault for the o2 sensor, its very unlikely that a bad o2 sensor will have just a fault for cat efficeny. Running rich will damage the cats but the check engine light will not flash it will come on and burn steady with o2 sensor adaptaion faults bank 1 and 2. only current fuel dumping misfires will be registerd with a flashing cel, and if you need a flashing light to tell you somethings wrong, you really need to pay more attention the the sound and feeling a misfiring car

Cruising for extensive time in your car will not damage your cats. Assuming your O2 sensors are working within specs, you are running lean while cruising. I frequently cruise at 3000+ rpm for 18+ hours straight and have never experienced problems with cat damage. As a matter of fact, the Motronic's computer requires numerous different engine speeds below 3000 rpm to even calculate a faulty oxygen storage capacity of the CAT.

I have had my CEL on and off for 22,000 miles, I dont pay attention to it. If its hitting on all 6, runs smooth, has good oil pressure, doesnt run hot, and gets good gas mileage, I dont pay attention to CELs. If a sensor gets out of its parameter 1 time, its triggers it. I only care about the vitals, which the CEL cant tell me.

Thanks for all the input so far. The idea about dead O2 sensors is a good one, but I think the original problem was really with the left cat. The reason is that the guys at AMS hooked up their laptop before I went back to Porsche, and they showed me that the post-cat sensor voltage was swinging up and down on Cat 1, whereas on Cat 2 it was less volatile (although they said it was still swinging a fair amount and Cat 2 might be dying too). I'll hand it over to the dealer on Tuesday in the hope that I'll get a shiny new passenger-side catalyst out of it, and if it turns out to be a different issue then I'll reconsider in light of all the advice here. The car certainly feels fine, and I can't detect any loss of power, gas mileage or smoothness, though I think the rate of oil consumption may be very slightly higher.

Latest update: The code is P0430, which I think is associated with the passenger side cat. The car is back in the dealers' hands, and today's courtesy car is a Subaru, which is certainly an improvement on the Fords I've been given before.

NNH:*but in town I tend to surf the torque curve at 2000-3000rpm. I've noticed that a lot of other rennlisters claim to hit 4000+rpm in each gear, but I must confess that I had put it down to c) some people being new to driving a stickshift.* I figure they put the 4000 mark on the tach where they did on account that's more or less where they wanted you to keep the revs if don't have the pedel down? I gotta admit the 996 is a dog at 30/kmh (19/mph) which is common for the *old towne* areas; doesn't feel right in either 1st or 2nd... I've been driving a shift for a while now... about 40 years: my father's Pontiac Tempest (3 on the tree), my grand father's '58 pickup (3 on the tree), MG-A model (3 on the floor), Ford tractor (had a bunch of gears), etc...

Good thing I got my apology in early!

The ECU uses the pre-cat sensors to know it has control of the fuel mixture. It does this by varying the fuel injection pulse widths by a very small amount which results in the fuel mixture swinging from just a bit rich to just a bit lean. How much is a bit? If the ideal air/fuel ratio is 14.72:1 the ECU will vary this by say 14.74:1 to 14.70:1 or even less -- I think it less but I don't have the exact numbers in front of me -- with the result the average is right on the ideal ratio which results in the converters getting an exhaust gas they can process sufficiently so the car and its emissions falls well within the allowable limits. The ECU knows the converters are performing up to efficiency by monitoring the post-cat O2 sensors. The ECU will (eventually) throw a fault code if it detects the O2 sensor is slow to respond, but this can take the ECU some time. The ECU is continuously monitoring O2 activity but the results are not examined and acted upon on a second by second basis. They are examined over a span of time. If one was to query the ECU for its readiness monitor test results he will often find the results for all sensors/subsystems monitored (and especially O2 sensors) are not available (they haven't failed, the results just aren't in yet) because the driving cycle was not long enough, not close enough to the necessary behavior, to allow the ECU to finish the continuous monitor testing it, well, continuously performs. The common assumption is the O2 sensor's either good or bad. If one is lucky, the O2 sensor fails outright and that's that. But most of the time they fail over time. What happens is the O2 sensor starts out working ok. Then it misbehaves. If it misbehaves bad enough (and most don't -- at least at first, though their performance does degrade more and more over time) or if it misbehaves long enough (sometimes) the ECU will log a DTC. In some cases, if the driving cycle (the trip) is short, or doesn't encompass the necessary variation, the ECU doesn't have enough "evidence" to flag the O2 sensor bad (emit an error code that could perhaps even point to the sensor) it may still emit an error. The error may be a bit of unintentional misdirection. Remember, the auto maker's on the hook for converters in its emissions warranty. The O2 sensors are not covered. The manufacturer is going to err on the side of caution and potentially flag the converter before it flags the O2 sensor. What the manufacturer wants to prevent is any damage to the converter that it may be obligated to pay for. If the car is brought in and the converter is found to be ok and the O2 sensor bad, well, you pay for the sensor and the manufacturer is not out the cost of the converter. The error code may implicate the converter when further diagnosis (perhaps even coming down to switching O2 sensors around to see if the symptom follows the sensor or stays with the converter) is done the culprit is found to be the O2 sensor. The CEL will flash when the engine controller detects a condition that puts the converters at risk, or when the ECU determines it is unable to detect this condition, that the ECU is no longer in control and no longer able to monitor its control over the engine the engine. (Theoretically, the engine could continue to run fine the engine controller deriving the proper ignition advance and fuel injection timings from maps (hard-coded info) in the engine controller's non-volatile memory) but the engine controller is really concerned about ensuring the engine's within the emissions envelop and that nothing the auto maker is obligated to pay for if it fails is at risk. In the case of a overly rich running engine -- say one fuel injector is leaking -- initially, the engine controller will try to lean the mixture out. It will lean (shorten the fuel injector pulse width) to all cylinders on the affected bank (determined by the O2 sensor inputs). If it is able to bring the mixture's richness to a level the O2 sensors indicate is within the safety envelop of the converter, all is well. (Well, all is not well, really, because one cylinder on that affected bank is still probably running really rich, while the other 2 are probably really lean. There is a limit to how much the engine controller can go to adjusting the mixture from the ideal (determined when the engine was new) and if the engine controller exceeds this limit, it will light the check engine light and log one or more error codes. The long term fuel trims are where the engine controller stores this adjustment value. Typically, if the adjustment value exceeds +/- 10% the engine controller will turn on the CEL.) The engine may misfire -- which will be detected by the engine controller and it can react in various ways. Misfire implies incomplete burning and excessive fuel making it to the converters. The engine controller will try some quick "fixes" but failing these to have the desired effect it will revert to limp mode in an effort to prevent damage to the converters. I'm not advocating anyone continue to drive his car with the CEL on and not have any concern until or unless the engine begins misfiring. I was just pointing out that had the converters been at risk of burning up, chances are very good he would have seen a flashing check engine light. In the original poster's case, my money is on one (possibly more than one) O2 sensor being intermittently bad (though this still requires replacement) and not a converter being bad. Sincerely, Macster.

Thanks for your input Macster. I'll report back as soon as I hear from the dealer.

This is a new code and indicates that the second CAT need replacing.

The dealer just called me to say that the other cat has indeed also died, so I now get another cat courtesy of Porsche and California's 7/70 emissions laws

Picked up the car yesterday morning. I've now put 150 miles on it, and so far (touch wood) it's behaving itself. Maybe the stern lecture I gave it, worked!

be sure, especially if one's on the hook to pay for any work that is done. Cat failure is rare. Usually, if they don't get banged around and the engine remains in good tune they last and last. You just know, though, the auto makers will strive to take some of that margin away... Sincerely, Macster

cats fail its not common but not uncommon.............cats fail way more often the o2 sensors on porsches

I would love see your evidenece to back that claim up.

few years of working on them, ive replaced probaly 4 o2 sensors and 50 cats

We are talkiing 996 CAT's here, not other cars. Are you trying to tell us that you have replaced 50 996 CAT's? I find that hard to believe since they are an emission warranty item that would be replaced free by the dealer.

When I had both cats replaced on my 02 the dealer indicated that he had replaced the cats on 2 other cars in the not too distant past. That's 6 cats. It may be that there was a bad supply of substrate for the cats. If one considers that the metal substrate is rolled to 0.0015", thinner than your hair, then a 10,000 lb. master coil would yield a lot of cats. Fecralloy and Apha-4 used to be common substrates for cats and both are difficult to make at best. I've been away from those products for a while so I'm not sure if they are still the prime candidates. Either of those alloys will withstand 1000C, 1832F, with no problems. To get beyond that temp the cat would have to be subjected to an overly rich condition. The typical failure mode for a metal cat would be small shards of metal separating from the monolith and exiting with the exhaust. However, if you have two cats in tandem the downstream cat might get plugged by an upstream failure. That is what you don't want to happen as you would get increased back pressure on the engine. A good reason not to ignore a CEL.

all but maybe 2-3 fell under 8yr/80k and were replaced by dealer <<<<<

So, what you really mean is that most of the other's were damaged by something that would not be covered by warranty or replaced by choice and not neccessity? Otherwise, any failure of a CAT not due to modifications would be fully covered since all but the earliest 996's under the mileage limit are still covered by Porsche.

No the ones that wernt under warrenty had over 80k miles on them, 80k is the limit on emissions warrenty. 8 yrs or 80k miles emissions warrenty on cats and dme

Why don't you check Firehawk's public profile and see what he does and where he works. Ciao, Adrian.

Im in with firehawk. Definetly replace more cats then lamda probes. And the dealer doesnt care, they dont pay for it, Porsche does.