410 replies to this topic

[LukeFranky]

Installed a new fuel filter and a replacement BBoD then reset the ECU.

Car boosted well. No surging. Running pretty smoothly.

Had my wife use my phone to scan for fault codes as I drove. We repeatedly saw a pending too lean code at the top of really long hills on the highway. That too lean code never once become an engine light or a confirmed code even after over 50km of test driving. It often cleared itself cruising down the other side of a hill. I only cleared the codes manually once during the test. The same pending code came back after that too.


Whatever the ECU is seeing it wasn't enough for it to trip up the engine light or confirm the code. This was decent test drive which would normally give me an engine light pretty quickly so I think some kind of progress has been made.

I'm not out of the woods yet. I'm going to swap the non-stock secondary for the stock one this weekend and do yet another smoke test to make sure I don't introduce any new leaks.

Still waiting on the VAG cable.

[LukeFranky]

Side note: Could I install some sort of filter in the vacuum lines coming from the IC to stop my new BBoD clogging up? Bit like the filters before the differential pressure sensor?

[Jimbo]

You could but they'de only get clogged again. A catch can would be the best way to minimise oil in the system.

At least your making some progress, though something is still making the ECU unhappy. I reckon you should ditch the tune.

[LukeFranky]

You could but they'de only get clogged again.

- I know, but that's exactly my thinking. It would be easier (and safer) to replace a filter than to clean out the solenoids.

 

A catch can would be the best way to minimise oil in the system. 

- I have a lot of blow by oil. Thinking of setting up a catch can actually.

 

 

I also forgot to mention, it is still misfiring at idle. I think this is a different problem which has something to do with low voltages at idle. After sitting idle for a 10-20 seconds the dash lights will start to dim and that's when the cam shaft sensor plays up and I can feel it miss. Reading the voltages on my scanner correlates the dip in voltages. Gets worse with the lights on.

[allpaw4]

Sounds like you have an intermittent earth fault there somewhere Luke, for lights to be dimming means bad things. Have a look at all firewall bungs and all wiring across the motor. Something is dragging the voltage down, but it probably isn't after a fuse, otherwise it would be blowing it. I'm thinking the main alternator cable. Also be worth checking the alternator itself. Unlikely, but worth checking.

[RX25SE]

Yep. Despite the ECU monitoring battery voltage, I have seen low voltage do some weird stuff to idle quality.

 

 

First thing I would be checking is charging voltage at the battery, then a voltage drop test from alternator B+ to battery positive post.

[LukeFranky]

Did some basic tests. The voltage at the alternator and battery matches pretty closely what the ecu reports on the scanner.

Strangely the low voltage problem didn't manifest during my test. It never dipped below 12 volts even with all my accesories and lights on. Didn't misfire either.

I'll have to try again.

[LukeFranky]

Found a GFB crankshaft pulley on my car. I have heard aftermarket pulleys can drive accessories like the alternator a bit slower. Could this cause voltage issues at idle?

[Robbks]

Did some basic tests. The voltage at the alternator and battery matches pretty closely what the ecu reports on the scanner.

Strangely the low voltage problem didn't manifest during my test. It never dipped below 12 volts even with all my accesories and lights on. Didn't misfire either.
 

Dropping below 13.8V at idle shows there is a problem.

My legacy would run like rubbish, wouldn't shift properly, ran lean until it was warm.

Dropped a new battery in it.
starting and cold-running problems fixed.

Charge the battery overnight (disconnected from vehicle),
let it sit for an hour
measure voltage.
then connect your multimeter across the terminals and put a load on it (something lie a 25watt brake light globe.)
see how fast the voltage drops.

[LukeFranky]

I'll try those suggestions.

Just had the battery tested at Autobarn quickly after it was sluggish to start today. Tested 'Good but charge'. He tested the alternator, 14.1 volts at the time.

Got home, car was misfiring at idle so grabbed the multimeter. 12.1 volts at the battery at idle with the lights on.

[RX25SE]

If the battery is more than fours years old, replace it; regardless of test results.

[LukeFranky]

Put in the stock turbo. Made no difference.

Replaced the battery with a decent new one. Dropping to 12.8V at idle with A/C, headlights and fans on. 14.0V - 14.3V when the revs get up. Hmm.

I also noticed the car is getting hotter than normal. Sits above halfway. Still in the normal range. Could just be the warmer weather.

[allpaw4]

Mine usually sits just below half. Not by much, but enough your eye can see the angle.

Something that might be worth replacing is the rectifier on the back of the alternator. New battery seems to have helped it, but certainly not cured it. Get a temporary replacement from someone who has a gen 3 as a check to see if it works before you shell out for another one.

Might be fine, but might not too. That's the problem with intermittent faults. They are a bastard to catch. Also, have you checked the state of the brushes and slip rings in the alternator itself?

[LukeFranky]

I haven't checked anything on the alternator yet. Not sure where I'd find a spare rectifier to try.

Car surged a little bit on a steep hill yesterday. Not exactly like before but there was definitely a pulse in the boost a few times.

Would a failing fuel pump drag a lot of extra current at idle?

[RX25SE]

Would a failing fuel pump drag a lot of extra current at idle?

 

Probably less current as the pump would be doing less work. I'm guessing a normal pump draws somewhere between 4 to 6 amps and if there was a major blockage it may rise to nearly 10 amps.

 

As above, intermittent charging issues are often faulty voltage regs.

[LukeFranky]

Probably less current as the pump would be doing less work.

- Thanks. Thought it was unlikely, was just thinking aloud.

 

Only reason I suspect a fuel pump issue is the car did improve with a new fuel filter (and BBoD). It takes a lot of load (steeper hills) on the engine to bring up the engine light now. I only have to worry about replacing the main fuel filter don't I? There's not any other filters that could be blocked?

 

Cam shaft sensor plays up every time I first start the car now because of low voltages. Drive around for a bit, come to a stop and it is OK on idle until the voltage drops. Lights on makes it worse.

 

I haven't had much to do with alternators before. Is there anything on it I could clean or cheaply replace before forking out lots of cash? From what I do know, if it is faulty voltage regulator then I need a new alternator?

[LukeFranky]

Tested the exhaust for the third or fourth time now, except this time it was hot and the engine valves weren't in a position letting the smoke flow into the intake.

Found a leak. Guess where?
The secondary turbo exhaust actuator arm. I can upload a photo if you need. Basically the actuator arm goes to a bolt/pin that goes into the inner workings of the valve in the exhaust. It is leaking around that bolt/pin. Seems like an obvious place to leak, not sure how you would seal it with the heat it is exposed to.

Maybe it leaks from the day it leaves the factory and isn't actually an issue? Mind you, the wastegate bolt/pin doesn't leak. Anyone else smoke tested their B4 exhaust?

Before you ask, I didn't swap that part of the exhaust along with the turbo. I think it is the original on there, I can't remember. I do have another one I can bolt in place. Thinking of pressure testing the spare somehow before even bothing to pull the turbo again to swap it.

[LukeFranky]

Data logs!

 

Earlier time stamp is up a steep hill that brings on the lean error code and then back down the other side (for comparison).

 

Second file is of me accelerating to highway speeds then just cruising for a bit (in case it helps).

 

Anyone care to see if they can see a problem?

Attached Files

•  Documents.zip   20.49KB   10 downloads

[natho]

I'm not an expert on this, so hopefully someone else can validate my findings... two things stand out to me - Air/Fuel Correction #1 (%) is quite often at 25%, which I believe is the maximum positive (adding fuel) correction. the lambda values, if i've converted them correctly, indicate extremely lean AFRs, e.g. 1.3 = approx 19:1.

 

Still sounds like a vacuum leak, faulty AFM or front O2 sensor. Could also be a fuel supply problem. Without going back over the thread I can't remember what you've already tried.

 

You haven't logged the knock correction parameters which might be useful, to see if it's been pinging. If the mixtures are lean this could be a problem too. It would also be useful for you to run learning view (google it, download from romraider forums) and post up a screenshot of the data this gathers. it shows learnt knock correction parameters as well as fuel trims.