Boxster 986 O2 sensors Misfire and running rough.

[DrRobin]

Hi,

 

I had my 2001 2.7L Boxster serviced in June and all was fine, it has done 93,000 miles.  A week or two later noticed an increased noise from the exhaust and on closer inspection there was a small hole in the CAT on Bank 1 (drivers side in the UK).

 

I got two CATs and had my local garage fit them (now July), no problems there and then filled up on the way home at a garage I don't normally go to as the tank was low.  The next few days I had random misfires, but only after deceleration and at low revs (less than 2,000 rpm) when feeding the power back in.  The misfire was random (sometimes cylinder 1, then 6, then 2, then 4).  I thought may be dirty fuel.

 

My OBD-II code reader showed ageing on all four O2 sensors, I guess they were just old and swapping them over into the new CATs finished them off.

 

After a long wait to get new sensors (now August), I decided to fit them myself, the old ones were difficult to get out of the new CATs, so I took both CATs off, fitted the new O2 sensors and all faults cleared with none coming back.  I took it for a test drive, no misfire and back to normal.

 

Roll on another week and the misfire was back, but not all of the time.  I tried a fuel cleaner and a couple of tanks of good fuel from my regular garage, but still got the odd misfire.  I then noticed that the battery was quite low and had been for some time, so recharged up to full on a recovery charge, the result no more misfire, although the engine is not running as clean as it was.

 

There is no problem at idle or when pulling, it's at small throttle openings or slow speed, it's not a misfire, but it isn't running very cleanly, so since O2 sensors was the last thing I have gone back to those.  I initially thought that the output looked normal, the two sensors pre-CAT move from rich to lean in nice smooth graphs and the post CAT sensors are fairly smooth, so I tried a re-learn and it was then that I noticed that Bank 1 Sensor 2 (Post CAT) is sitting at 0.2V soon after starting whilst Bank 2 Sensor 2 is 0.50V.  Once the engine warms up Bank 1 Sensor 2 does come up to around 0.5V when the car is driven, but soon drops to a fairly low voltage when idling, this shows the sensor is working but may be the heater isn't, is that what you would think?

 

You can see the big difference in the attached screenshot from Torque Pro.

 

I know Sensor 1 is used to set the mixture and thought that Sensor 2 is used to test the CAT, but is Sensor 2 also used to set the long term fuel trim and could a faulty heater in one sensor post CAT cause rough running?

 

Robin

[RFM]

Have you already tried cleaning the throttle valve ? At the same time clean the airflow meter with a spray developed for this purpose, CRC brand for example. Al this is inexpensive and takes little time, i would start with that.

[DrRobin]

I haven’t checked either yet, but will have a look next chance I get.

[DrRobin]

I took the throttle body off, there was some black deposits around the opening (see photo), but overall it was fairly clean.  There was a small amount of [clean] oil in the black plastic tube after the throttle body and small amounts in the plenum and second plastic tube with the valve in it (see photo).  The oil wasn't excessive and I changed the AOS last year, then there was a lot of oil residue in there.  A short test run and it is may be slightly better, but still a little rough at low revs/throttle openings.

 

I don't get any fault codes and don't have a misfire (since charging the battery) so I now suspect coils or plugs.  Coils were changed in 2008 at 52k miles, plugs I can't find when they were last done, I am sure they have been changed, but can't find a record.  I can't find the exact service interval for plugs, I see it quoted as 24k or 40k miles, so think I will have a look next chance I get.

[DrRobin]

With the engine at idle, I tried removing the oil filler cap.  There was a slight vacuum holding it in, but it removed okay.  As soon as it was removed the engine note changed and it sounded more un-even.  I think this means that the crankcase vacuum and AOS are both okay?

 

Just another question, the air intakes, straight after the throttle body there is a diverter (T shaped tube) that allows air to both plenums, but then at the front there is another connecting pipe with a vacuum operated butterfly valve in it.  I assume that this is some sort of balancer pipe, but when does the valve operate, at small throttle openings when the inlet vacuum is high?  Can anyone let me know what function this provides please?  I don't think there is any problem with this, just curious.

Edited October 19, 2021 by DrRobin

[mikefocke]

Since you asked, this was posted by a user many years ago and I snaged it.

 

The O2 sensors do just that, sense the amount of O2 in the exhaust gas relative to the amount of O2 in ambient air. Perfect combustion of a perfect mixture of air and fuel (around 14.7/1 air/fuel ratio) leaves behind only CO2 and water as products of combustion. All the oxygen gets consumed in the combustion and combines with all the carbons and hydrogens. If there is not enough fuel (lean mixture), then all the fuel gets burned leaving some oxygen left over. Conversely, if there is too much fuel (rich mixture), then all the oxygen gets burned leaving behind extra hydrocarbons (fuel). Now an oxygen sensor outputs a voltage between 0 and about 1 V depending on the difference between the amount of oxygen in the exhaust and the amount of oxygen in normal air. If there is a lot of oxygen in the exhaust (lean mixture condition), the sensor outputs close to 0 volts. Conversely, if there is no oxygen in the mixture (rich condition), then the output is close to 1 V. These O2 sensor voltages are read by the computer. This is the feedback loop that tells the computer how the engine is performing with regard to air/fuel mixture. It's impossible for the computer to hold the exact perfect air/fuel mixture constantly, so the way mixture control is designed is for the computer to continually adjust the mixture from very slightly rich to very slightly lean and back again using feedback from the pre-cat O2 sensors. This means that the pre-cat O2 sensor signal will oscillate back and forth from high to low to high to low voltage as the computer adjusts the mixture. In a normal running engine at idle the signal goes from low to high voltage and vice versa about every 1 second, with a transit time from low to high (or vice versa) being about 200-300 milliseconds. This transit time is important because as an O2 sensor ages, the transit time gets longer, and eventually it can get too long such that the computer will call it a malfunction and signal a check engine light and fault code for a slow responding O2 sensor. O2 sensors need to respond to mixture changes quickly so that the computer can keep up with the proper mixture adjustments.

So the bottom line is that the pre-cat O2 sensors should oscillate between about 0.2 to 0.8 volts regularly (about every 1 second at idle) in a healthy engine.

The post-cat O2 sensors are identical to the pre-cat O2 sensors (same voltage outputs). They are there only to monitor the performance of the catalytic converters. So, as discussed, the pre-cat sensor signals are oscillating between 0.2-0.8 volts. Once the exhaust gasses pass through the catalytic converter, most (all, in theory) excess fuel (hydrocarbons) will be combusted thus reducing hydrocarbon emissions. The cat uses oxygen in the exhaust to combust the fuel. So what you end up with in the exhaust after passing through the cat is a gas mixture that is reduced in hydrocarbons and reduced in oxygen relative to the mixture entering the cat. The post-cat exhaust gas mixture should be CONSTANTLY low in oxygen if the cat is doing its job of burning excess fuel. Therefore, the post-cat O2 sensor signal should be a constant lower voltage reading (not oscillating). So, if the post-cat O2 sensor is seen to oscillate just like the pre-cat O2 sensor, that means that the post-cat sensor is seeing the same gas mixture as the pre-cat sensor meaning that the catalytic converter isn't doing its job of burning excess fuel. The computer monitors the post-cat sensor and compares it to the pre-cat sensor. If the signals are similar, it assumes the cat is bad.

[DrRobin]

It has turned out to be coils, all 6 had at least one hairline crack.  I obtained new Beru coils and sin new spark plugs, fitted yesterday and all is now good.

 

i had replacement coils at 55k miles and now again at 93k miles, not sure what causes them to fail, but they don’t seem to last that long.