Notices
Boxster & Boxster S (986) Forum 1996-2004
Sponsored by:
Sponsored by:

Genius needed - unusual performance problem

Thread Tools
 
Search this Thread
 
Old 04-27-2018, 03:50 PM
  #1  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default Genius needed - unusual performance problem

Grab a cuppa, I'll give you a basic run down of the problem history.

1. AOS failed and was replaced and ran fine for a week
2. Had sudden lack of performance and bogging
3. Cleaned the throttle body, was covered in oil due to the failed AOS. Not much change in performance.
4. Days later, had P0491 & P0492 codes (SAI ineffective in both banks)
5. Replaced MAF, no change
6. Days later, P0491 & P0492 codes cleared themselves, car mysteriously ran better (but only what feels like 90% of what it should be)
7. Days later, back to being bad, though no SAI codes
8. Next day, car run's great.
9. Gradually, SAI pump starts to whine and make bad noises on cold starts (explains the codes). Car still running good.
10. Days later, car down on power again, and starts bogging when pulling off as if I wasn't giving enough throttle. I stalled it a few times when pulling away
11. Next day, performance improves drastically
12. Days later, stuttering acceleration and stuttering deceleration when on light throttle (either applying or lifting off throttle), as if it can't make up it's mind if the throttle is being pressed or not, and as if I was feathering the throttle, however a steadily pressed throttle was fine, this only happened between 0-10% of throttle (I've checked TPS, readouts seem fine and don't fluctuate).
13. Next day, car performance improves drastically. Again.

What in the name of ****.

Now the SAI codes must be separate to the performance issue as the pump turns off after a cold start, and also wouldn't affect performance like this. The car has also just passed an MOT with an emissions test yesterday, so it's not the CAT's.

Suspected vacuum leak then, but I've checked and re-sat all pipework and tested with carb cleaner everywhere, all seems fine (will run a smoke test in due course).

Then a few days ago, the stuttery acceleration/deceleration re-appeared on a drive home from work. I drove to a quiet industrial estate to troubleshoot it, drove at 10mph in first gear and applied full throttle. The car felt like I stabbed the throttle fully, lifted off, then jabbed it again (within the space of half a second), as if there was a sudden and instant drop in power for a split second.

Today the car has felt okay, but this is where the plot thickens.

Using the android Torque app on my head unit, I checked my fuel trim and O2 sensor readouts. If I had a vacuum leak, the fuel trim would be very high to compensate for the leak, correct? Here's mine. Short term (instant) on top, long term on the bottom:

Name:  X0Qs02Jl.jpg
Views: 68
Size:  31.7 KB

Completely normal. Normal operating readouts oscillate constantly between around 5% and -5%. If I were to have a vacuum leak, the long term and/or short term would be way into the positives.

Now look at my O2 readings on a cold start after the SAI has turned off. Top two are pre-cat, bottom two are post cat:

Name:  eotvmoul.jpg
Views: 68
Size:  44.1 KB

Post cat O2 readings are still fine - bang in the middle. No signs of running lean.

However, I then drove the car for a while to warm it up, stopped, and checked it again:

Name:  poXZKMil.jpg
Views: 69
Size:  39.7 KB

Suddenly the post cat results are way, way too lean. The two lines should be in the middle.

Now take a look at this - these readings are whilst driving steadily at 30mph on the same quiet industrial estate:

Name:  wC1wBvkl.jpg
Views: 67
Size:  26.9 KB

Name:  1cWBnIWl.jpg
Views: 67
Size:  41.0 KB

Then when I stopped the car:

Name:  QaTRG0cl.jpg
Views: 71
Size:  42.9 KB

They're all over the place!

Reading up more and, apparently, bad coil packs and misfires can cause a lean result (though I've never had a misfire code). You'd think that they'd cause a rich result due to the unburnt fuel, but as O2 sensors only detect air and not fuel, they'd pick up the unused O2 in the cylinders and would apparently cause a lean result. But mine's constantly lean when warm, and not spiking, so that doesn't seem to make sense.

And finally, after that information, I checked my coil packs. Almost all of them have cracked casings, and this part is totally missing off one:

Name:  EYypjHLl.jpg
Views: 70
Size:  21.8 KB

So… could it be coil packs? Or does the O2 read outs not make sense for it to be ignition related? Or does it?

I'm at an end to how much effort I'm putting into fixing this problem. That car's been a bitch and has been so unreliable with every common fault (bar the IMS) going wrong on it. They're all fixed, and this is the last issue that I have, but I’m literally on the verge of selling the car as this has been happening for months.

Help me Obi-Wan Kenobi, you're my only hope
Old 04-28-2018, 03:28 PM
  #2  
Byprodriver
Rennlist Member
 
Byprodriver's Avatar
 
Join Date: Feb 2010
Location: So.CA
Posts: 3,454
Received 173 Likes on 135 Posts
Default

While you are in there pull the spark plugs to see how they look, probably want to replace too. Be Certain to fully connect coil lead wires when reassembling, easy to miss this step. If a coil etc is misfiring you will get a 300 trouble code.
Old 04-28-2018, 03:59 PM
  #3  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

I've read a few places where bad coils affect performance and don't throw any CEL's. I'm so lost.

Update - here's the fuel trim readings on this morning's cold start:

Name:  LA4kd9Ll.jpg
Views: 68
Size:  40.5 KB

But after about 5 minutes they return to normal, and stay normal for the rest of the day (for as long as it's still warm).

Also, the post cat sensors very rarely but sometimes mimic the pre-cat sensors, but only for a few seconds at a time.

Name:  9t8AUNvl.jpg
Views: 71
Size:  49.2 KB

Now that's normally a sign of bad cat's but as I've said it happens very rarely, and emissions were fine. Happens on both banks.

I'm going to do a smoke test soon to rule out a vacuum leak. Getting a little tired, I'll be honest.
Old 04-28-2018, 07:54 PM
  #4  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

So thinking about it, the lean fuel trims in my last post is due to the SAI pump not working. Lack of air because of that, so it reduces the fuel to match, then puts it back up to normal once the car is warm. So that could explain that.

Anyway. Should the pre-cat sensors always have that saw toothed readout? When I'm driving it's all over the place but it goes back to normal once idling. I'm wondering if they were damaged with the oil when the air oil separator failed, and now they're lazy. Possibility maybe? I've got two spare, may throw them on.
Old 04-29-2018, 01:46 PM
  #5  
Macster
Race Director
 
Macster's Avatar
 
Join Date: Sep 2006
Location: Centerton, AR
Posts: 19,034
Likes: 0
Received 254 Likes on 224 Posts
Default

What MY? How many miles?

Are there any active error codes? That is is the CEL on and what codes are present? Be sure to pull the freeze frame data and code as this will be the error code that triggered the CEL. Be sure to pull any pending error codes and any permanent error codes. (These are codes that only the ECU can clear after so many warm up cycles with error absent.)

Not sure about those O2 sensor readings. Narrow band sensors only read from 0 to 1V. Under normal operation the #1 O2 sensor voltage swings from 0.1V (or a bit less) to 0.7V (or a bit more). At hot idle the swings occur roughly every second.

Under normal operation the #2 O2 sensor voltage remains in the 0.6V to 0.7V range which indicates there is very little oxygen in the exhaust gases even though the ECU knows it allowed some excess oxygen via a "lean" mixture as indicated by the low voltage readings from the #1 sensor. The "high" voltage reading indicates the excess oxygen is being consumed in the converter and all is well with the converter.

However odd the sensor readings are I suspect you need to consider the secondary air injection errors.

Just a WAG, but in replacing the AOS a secondary air injection hose got dislodged or due to age cracked or split when disturbed. Check the air hose to/from the secondary air injection pump.
Old 04-30-2018, 05:19 AM
  #6  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

Originally Posted by Macster
What MY? How many miles?

Are there any active error codes? That is is the CEL on and what codes are present? Be sure to pull the freeze frame data and code as this will be the error code that triggered the CEL. Be sure to pull any pending error codes and any permanent error codes. (These are codes that only the ECU can clear after so many warm up cycles with error absent.)

Not sure about those O2 sensor readings. Narrow band sensors only read from 0 to 1V. Under normal operation the #1 O2 sensor voltage swings from 0.1V (or a bit less) to 0.7V (or a bit more). At hot idle the swings occur roughly every second.

Under normal operation the #2 O2 sensor voltage remains in the 0.6V to 0.7V range which indicates there is very little oxygen in the exhaust gases even though the ECU knows it allowed some excess oxygen via a "lean" mixture as indicated by the low voltage readings from the #1 sensor. The "high" voltage reading indicates the excess oxygen is being consumed in the converter and all is well with the converter.

However odd the sensor readings are I suspect you need to consider the secondary air injection errors.

Just a WAG, but in replacing the AOS a secondary air injection hose got dislodged or due to age cracked or split when disturbed. Check the air hose to/from the secondary air injection pump.
There's no CEL's believe it or not, just error codes that have been stored that pop up.

Surprisingly, I've just had a new one. P2096, showing that the post cat sensor has caused the fuel trim to be too lean. After some research, this is also a faulty sensor code.

Could that be the whole issue? Bad sensors? Maybe when the AOS failed it's fouled the sensors?
Old 04-30-2018, 06:49 AM
  #7  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

Okay now I'm confused. It seems that on the 996, the P2096 is a bad O2 sensor, and on the Boxster the P2096 is the fuel trim being detected as too lean. Why on earth are they two separate meanings with the same code - I had my hopes up that this would be a case of a bad sensor and that's that.
Old 04-30-2018, 11:59 AM
  #8  
Macster
Race Director
 
Macster's Avatar
 
Join Date: Sep 2006
Location: Centerton, AR
Posts: 19,034
Likes: 0
Received 254 Likes on 224 Posts
Default

Originally Posted by geraintthomas
There's no CEL's believe it or not, just error codes that have been stored that pop up.

Surprisingly, I've just had a new one. P2096, showing that the post cat sensor has caused the fuel trim to be too lean. After some research, this is also a faulty sensor code.

Could that be the whole issue? Bad sensors? Maybe when the AOS failed it's fouled the sensors?
Been through 3 AOS failures with my Boxster. The sensors can get fouled with oil. After the 1st AOS failure the SM told me that it can take some driving (IIRC he said up to 50 miles) to rid the engine/intake of the oil, the effects of the oil that the engine ingested from the failed AOS. The driving wants to be such the engine gets fully up to temperature and the overall engine gets hot. This helps any residual oil in the intake work its way into the engine.

After the 1st AOS failure there was an O2 sensor error. The trouble is this has been so long ago I can't remember the error code. Even the details are fuzzy in that I can't remember the exact circumstances, if the O2 error went away after some driving of the car after the AOS failure only to return some time later. I do remember having some sensors replaced but I do not believe it was that soon after but long enough past the AOS that the AOS failure was not really to blame. The sensors (or just a sensor) acted up because it just happened to go bad.

Since then 2 more AOS's have failed and there has not been any issue with the O2 sensors. The last time I had the car in for a new AOS the CEL came on as I was driving the car off the dealer lot. I do not recall the error code but I cleared it and just drove the car normally and the CEL never came back.

If the engine were running ok and you were just getting the CEL I'd be inclined to recommend you run a bottle of Techron through the engine. But with the engine running poorly I'm not a fan of running the engine any more than necessary, just enough to identify the cause of the problem in order to fix it.

Generally in my (now gone) Boxster OBD2 references, and in my 996 Turbo OBD2 references, the guidelines called for if there was an error code pointing to a bad sensor to replace the sensor or sensors indicated by the error code (or codes) the clear the codes and road test the car.

I think you need to consider doing this. My advice would be unless both sensors are new and one is just manifesting a case of infant death syndrome you replace sensors in pairs. Say the #2 sensor on bank #1 is triggering a bad sensor error code. I'd replace that one and its counterpart on the other side to ensure both senors are new and have the same behavior so both banks get fueled only based on their actual different fueling needs -- if any -- and not due to an aged sensor that hasn't gotten bad enough to trigger a code yet, but is still performing below par.

Clear the codes, even if none are active to reset all the adaptations back to their defaults, then road test the car. If the engine manifests any untoward behavior you should not continue to drive the car as there is something else wrong.
Old 04-30-2018, 01:24 PM
  #9  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

Thanks for the advice, much appreciate it. Going to replace them and clear the codes.

Except for what just happened.

A friend of mine who vapes decided to try and blow his vape into one of the vacuum lines to see if we could see where the leak is. The line we chose is the one going from the triple-pointed join in the middle (item "i"), running to the left of the engine into a valve (item "J"). The line joining these two is much longer on mine though:



We disconnected that line from item "i", kept it attached to the valve (item "J") and he blew into it. Except he couldn't blow anything into it. I took the pipe off and it wasn't blocked, but for as long as it was connected to the changeover valve (item "J"), he couldn't get any air into it at all. Normal?

The other thing was that I started the engine and heard the pump at it's usual weird high-pitch whine that's not normal. At this stage I disconnected the hose and it blew A LOT of air out, and the pitch dropped drastically. Basically, there's no air getting into the pipe and it's putting stress onto the pump.

So I have two problems. The first was the fact that no air could be blown into the vacuum line whilst attached to the changeover valve (the valve that links to the resonance flap in the manifold), and that the SAI is blocked. I don't know which part is blocked though. Is item "D" on that diagram the cause?

And then I snapped said line.

Name:  cRPpscEl.jpg
Views: 68
Size:  38.8 KB

Yup! Me being an idiot.

Weird thing is, the engine hasn't been affected in the slightest... I'm driving it home in a second so we'll see.

So now I've got multiple questions:
- Is the changeover valve (item "J") faulty as we couldn't blow into it? Or maybe it's normal that we shouldn't have been able to blow anything down those small lines whilst attached (engine off and running)?
- What's blocking the SAI's pipes? Item "D"?

One step forward, two steps back.
Old 04-30-2018, 02:07 PM
  #10  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

Confirmed, drove home and theres no performance difference with that line disconnected. Drives exactly the same
Old 05-01-2018, 12:08 PM
  #11  
Macster
Race Director
 
Macster's Avatar
 
Join Date: Sep 2006
Location: Centerton, AR
Posts: 19,034
Likes: 0
Received 254 Likes on 224 Posts
Default

Originally Posted by geraintthomas
So now I've got multiple questions:
- Is the changeover valve (item "J") faulty as we couldn't blow into it? Or maybe it's normal that we shouldn't have been able to blow anything down those small lines whilst attached (engine off and running)?
- What's blocking the SAI's pipes? Item "D"?

One step forward, two steps back.
Well, without my references -- which I gave away after I got rid of my Boxster -- I'm not really able to offer much more in the way of help. Fortunately over the years I had no secondary air injection issues with my Boxster (although it might have been a problem with this that was causing the P1128 error code that forced me to dispose of the car when 2 dealers were unable to diagnose the cause and I could not smog and thus could not register the vehicle).

As you can see there are a number of valves and all have to work properly so the air is routed to the proper location. Hoses likewise have to be in good condition and of course properly connected. The secondary air injection system is rather complicated and I'm not an expert on it by any stretch of the imagination.

A google search though I'm sure will turn up a number of hits of others who have dealt with secondary air injection issues with their Boxster and you can probably get some good ideas from reading of other's experiences.
Old 05-02-2018, 12:32 AM
  #12  
Geof3
Advanced
 
Geof3's Avatar
 
Join Date: Dec 2016
Posts: 88
Likes: 0
Received 0 Likes on 0 Posts
Default

Can you simply leave the SAI system off, you are already there, and get a reflash at a dealership to accommodate? Pretty popular here in states that don't require emissions testing.
Old 05-02-2018, 04:37 AM
  #13  
geraintthomas
Cruisin'
Thread Starter
 
geraintthomas's Avatar
 
Join Date: Apr 2018
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Default

But the problem isn't the pump, it's the lines/system, which is also affecting performance. Disabling the SAI won't help unfortunately, I need to find the problem rather than mask it.
Old 05-04-2018, 03:55 AM
  #14  
Geof3
Advanced
 
Geof3's Avatar
 
Join Date: Dec 2016
Posts: 88
Likes: 0
Received 0 Likes on 0 Posts
Default

Originally Posted by geraintthomas
But the problem isn't the pump, it's the lines/system, which is also affecting performance. Disabling the SAI won't help unfortunately, I need to find the problem rather than mask it.
The SAI system is simply an emissions system that isn’t actually needed. However, the car looks for it. The ECU can be programmed with a different tune ROW for us, to eliminate the system completely. Fairly common mod as the system can be problematic as you are experiencing. It is however, illegal in some states in the US to tamper with the emissions controls, but in states where this is not an issue, it is no big deal.

Old 05-04-2018, 12:30 PM
  #15  
Macster
Race Director
 
Macster's Avatar
 
Join Date: Sep 2006
Location: Centerton, AR
Posts: 19,034
Likes: 0
Received 254 Likes on 224 Posts
Default

Originally Posted by Geof3


The SAI system is simply an emissions system that isn’t actually needed. However, the car looks for it. The ECU can be programmed with a different tune ROW for us, to eliminate the system completely. Fairly common mod as the system can be problematic as you are experiencing. It is however, illegal in some states in the US to tamper with the emissions controls, but in states where this is not an issue, it is no big deal.

The SAI is part of the vehicles emissions system. It is needed for the car to remain emissions compliant.

It is a violation of federal law to tamper with, remove, modify, render inoperational any of the vehicle's emissions systems. Furthermore vehicle service providers are under pressure to report this: "According to EPA guidelines, service providers can be found liable for emissions tampering not only for altering an engine, emission and associated electrical system, but also for not reporting — to the EPA, vehicle owner or both — when evidence of tampering is found in during engine-related service work."

The OP needs to diagnose the source of the problem and address it.


Quick Reply: Genius needed - unusual performance problem



All times are GMT -3. The time now is 09:37 PM.