EGR valve inspection/cleaning on 16v/1,4l (BBY) - a complicated task?

[sepulchrave]

You're assuming that every valve behaves the same way, in practice solenoids cannot be manufactured that accurately and every valve will be slightly different so the PWM voltage value applied will need to calibrated against the measured position potentiometer value in order that the ECU can request a specific amount of valve lift according to engine operating conditions, these values must be stored somewhere, over time the valve spring will soften, the flow will be reduced by deposits and the solenoid will perform worse than when new due to thermal cycling so the system must be dynamic and the values stored must be modified accordingly. Therefore if you change any of these variables by fitting a new valve or cleaning out the deposits then the system must be told that it needs to start from scratch again.

[Breezy_Pete]

Command a certain pot value, then dynamically tweak the PWM ratio until the pot position is achieved? Simple feedback-regulated system? 

[sepulchrave]

I know for a fact that the MAP sensor is used to correlate requested with actual flow, this is only done once the coolant temperature is above 80C, i.e. after a few miles of running. It must be a parametric comparison using stored values.

[gewstarr]

First, big thanks for all your input, it's of extremely great help.

Now, it's been a couple of days, and I've driven the car more or less madly. Around 75 mph on the highway on fourth gear, for example. Some engine-torture, some would probably say. I was actually just trying to reproduce that CEL light on my dashboard, but it hasn't come back on.

However, something weird, occasionally the engine starts coughing. Seldom on high speeds, but often when I'm around 35~40 mph and say third gear, with a steady foot on the gas pedal. It's almost that feeling of going empty on fuel and stalling, but only happens for a split second and then it's back to normal. No dashboard light, though. Apart from this, things have been normal. I have driven at least 70 miles since EGR cleaning, this without any adaption.

Oh, VCDS-Lite has always been of great help to me, and I believe that great deeds (and great software) shall be rewarded, so yesterday evening I sat down and Paypal:ed $ 99 to Ross-Tech for a license.

Tonight I got an email with the license file. I went out, hooked my netbook up, and ran the procedure. Well, except for the last part, clicking 'Switch back to basic settings' were grated out after taking a read on 074, so all I could do was read the data and then leave. However, it tells me that adaption was okey, I think?

Afterwards I started the engine. No difference in driving, as I could feel anyways.

I can't help wondering why the throttle is occasionally acting up, like described above. Ideas?

[rum4mo]

Yes, VCDS will always, in my experience, report when "perform basic functions" or adaption has failed to provide "good" results.

 

Your throttle problem, could you say that it feels like a form of "hunting" but only on light engine loading, if so, I also detected that after doing this work on a BBY engine, I was concerned that I had messed up the TB calibration by poking about with the air vane slightly, eventually that feeling disappeared no doubt as the ECU built up a new sensible table of look up values.

 

I've forgotten, did you say that you have cleaned up the MAP sensor now as well?

 

One other thing that you could do now is, carry out that same adaption task on the TB - I did that at the time I cleaned up the TB - but did it a couple of times later to see if it helped with the "wooly" feeling on light throttle.

Edited 3 November, 2015 by rum4mo

[Breezy_Pete]

If you haven't been able to switch to basic settings, I don't think you've done any adaptation. Maybe there's a delay in your software being suitably registered or something, due to time differences?

[rum4mo]

I for one do not know the limitations of VCDS-Lite, is this sort of thing possible with it?

[sepulchrave]

I for one do not know the limitations of VCDS-Lite, is this sort of thing possible with it?

 

Yes.

[felicia16v]

You need to get the engine hot,then adapt the egr valve and the the throttle body. Ign on engine not running. Also first go to adaption and save channel 0 to reset to factory defaults.

Make sure coolant temp is over 80° .

[gewstarr]

Again, thanks for all your input.

Now, situation is slightly changed for the worse. Actually, it's terrible. It started a couple of days ago, trying to pull that adaption. Oh by the way @felicia16v, according to Ross-Tech coolant should be 5-95° C when doing throttle adaption.

Anyways. I finally got to the 'Switch to basic settings' menu, but it failed. I started the engine afterwards, only to notice how the throttle was all messed up. I pressed it all the way but the rev only slowly raised from idle up to like 1300 rpm, and that was it. I though to myself; damn, now I've messed things up big time. I cleaned fault codes and restarted the car. Suddenly, everything was back to normal. Puh. What a great release. Me and my lady drove home under normal conditions.

The same evening I was going for a drive. Car started normally but after less than 100 m from the parking I felt that I lost one canister. The machine was shaking, EML started blinking a couple of seconds later. Went back to the parking lot and checked codes. Cylinder 1, misfire. I called it a night. I was tired.

Now, yesterday evening I went back out to investigate. Things looked bad, from what I could see. So, I tried replacing both spark plug and spark coil on first cylinder, but the fault still remains on this very (first) cylinder. Now, judging from the operation I've just been through, I unfortunately suspect that exhaust hasn't run smooth, after all, causing one (#1) internal cylinder exhaust valve to 'burn', as they call it. If this is the case, I understand that there's no shortcut; you will need to remove the engine head to get it replaced.

Gosh. Guys, seriously, please tell me that I'm jumping to mad conclusions over here, I could really need it right now. I have a compression testing tool lying somewhere in the basement, I'll dig it up and check data next thing I do.

Any ideas?

Edited 7 November, 2015 by gewstarr

[rum4mo]

Well one thing that you could do is to swop coils between cylinders, really just to prove that the new coil works, maybe same thing with new plug. I thought, a long time since I've done that job, that if the adaption failed, the new values were not saved so you would still have the original values. I seem to remember having had "adaption failed" happen to me, but I persisted and eventually I must have managed to get a good adaption and save these new values.

 

Does a SAAB 96 or 95 still look like a bad alternative (sorry - don't bother answering that!!!).

[sepulchrave]

Relax, your exhaust valves are fine, it takes many years to damage them, it certainly can't happen overnight.

 

Adaptation failure suggests a faulty EGR valve, Your misfire isn't related to this problem, it's most likely a faulty coil, spark plug or the wiring to the coil.

 

Keep fighting, you will win.

[felicia16v]

you need to do some more diagnosis before condemning the valves. 

Is the spark plug wet when you remove it? it could be the fuel injector thats gone wrong.

[gewstarr]

Gosh. Now I've more or less excluded the possibility of spark plug / coil. It's not even misfire fault codes (after resetting) anymore, but fuel bank (1). Engine shakes like a alcoholic on detox, though. I've unbolted the EGR and tried to see how the engine acts with it being tight versus not. This is the result, ultimately the car stalled, as you can see in the end of the video.

 

[gewstarr]

More on this. I unscrewed the two 5 mm Allen bolts holding the fuel injector rail. I had read in Haynes about first depressurize the system, but this seemed to be a complicated task involving reach for the fuel filter, so I skipped it and just pulled the rail. First time I mess with these little fellows. Used a slotted screwdriver to undo the clips holding the nozzles. After undoing one there was only a few drops coming, but after undoing the second it started pouring gasoline from both holes, some sort of beer bong mechanism, I guess. I was quick to put one nozzle back, so it stopped. Ran in with the other (#1) and had it in ultra sonic cleaner for three minutes, only water. Went back out, swapped 1 and 2 and reassembled the rail. O-rings were fine by the way.

Now, I managed to undo the two T15 screws holding the MAP sensor. I must say these were not in the smoothest of places, really tricky reaching them without dropping them in the engine bay. I have mad sausage fingers. Anyways, sensor came off, it looked pretty okey on first inspection, but I sprayed it with brake cleaner anyways and tried pushing some steel wool inside the small entity in it. Think I managed to bend some pig in the middle during this operation, but I bent it back to the center before I plugged it back in. Still getting misfire and different fault codes. Different cylinders, and also fuel trim, bank 1.

Now, lying here wondering, I suddenly got an idea. You know the big black plastic engine cover containing the air filter? During all these attempts to get the engine running smooth again, deleting fault codes, running with/without EGR/MAP connected/disconnected, I have never bothered putting it back between attempts. I've pulled the vacuum hose from it and it's currently lying in the trunk. I know it's not good bypassing the air filter, but this couldn't be the cause of this terrible idle, could it? It would definitely be an embarrassing end to this story. Think I'll head out right away, erasing codes one last time, and actually put the hose and engine cover back on before starting the car. I'll report back!

[Breezy_Pete]

No, the missing air filter won't explain it I don't think.

 

Did you not find your compression tester? I thought that was a very sensible thing to try, as it's completely free and quite easy.

[gewstarr]

You're right, problem remains even if I have the air filter / cover on.

Anyways, these were the fault codes I had in memory before that last erase.

VCDS-Lite Version: Release 1.2
Sunday, 08 November 2015, 21:29:58:37164
Control Module Part Number: 036 906 034 CB
  Component and/or Version:  MARELLI 4MV G      4831
           Software Coding: 00071
            Work Shop Code: WSC 13765
                      VCID: 54DB6E7FB7E3
           Additional Info: TMBHC26Y833807906     SKZ7Z0C1924276
7 Faults Found:
16684 -  Random/Multiple Cylinder Misfire Detected
        P0300 - 35-10 -   - - Intermittent
16686 -  Cylinder 2: Misfire Detected
        P0302 - 35-10 -   - - Intermittent
16687 -  Cylinder 3: Misfire Detected
        P0303 - 35-10 -   - - Intermittent
17961 -  Barometric / Manifold Pressure Signals: Implausible Correlation
        P1553 - 35-10 -   - - Intermittent
17912 -  Intake Air System: Leak Detected
        P1504 - 35-10 -   - - Intermittent
17564 -  Manifold Pressure Sensor (G71): Open/Short to Ground
        P1156 - 35-10 -   - - Intermittent
16497 -  Intake Air Temp. Sensor (G42): Signal too High
        P0113 - 35-10 -   - - Intermittent

Quite a few, heh.

 

After erasing and restarting car it actually ran without immediately firing away the CEL. Idle was pretty steady in terms of RPM, but still shaky, like misfiring-shaky. I pressed the throttle a couple of times but still no CEL. Then I got into 1:st gear and released clutch. Now I felt that it was definitely still something wrong. Lack of power, indeed so. After only a few meters the CEL turned back on. I circled the lot back into my parking space, and re-read codes.

VCDS-Lite Version: Release 1.2
Sunday, 08 November 2015, 22:01:42:37164
Control Module Part Number: 036 906 034 CB
  Component and/or Version:  MARELLI 4MV G      4831
           Software Coding: 00071
            Work Shop Code: WSC 13765
                      VCID: 54DB6E7FB7E3
           Additional Info: TMBHC26Y833807906     SKZ7Z0C1924276
3 Faults Found:
16684 -  Random/Multiple Cylinder Misfire Detected
        P0300 - 35-10 -   - - Intermittent
16686 -  Cylinder 2: Misfire Detected
        P0302 - 35-10 -   - - Intermittent
16687 -  Cylinder 3: Misfire Detected
        P0303 - 35-10 -   - - Intermittent

Sad. One thing strange I noticed, though. When I turned the car key from position 0 to 1, there was a really weird sound coming from somewhere in the rear. I'm guessing it's the petrol pump. I turned the key back to 0 and then back to 1, same weird sound. It was like a slurping sound. Weird. Don't think the fuel filter has been replaced on this car for at least 5 years, but that shouldn't cause this sort of sudden problem, would it? I mean, the car started fine that evening when I was about to take it for a drive, then less than 100 meters from the parking lot it suddenly started shaking, and ever since that moment it has never acted healthy. Apart from fuel filter, could the fuel pump itself be on it's way out? Any easy way to inspect it, or do you need to get in under the car, somewhere near the fuel filter? Is it possible to do testing and/or adaption on it via VCDS on Fabia Mk1's? I just spotted this article on Ross-Tech, describing some sort of process. There was also a Youtube video.

 

https://www.youtube.com/watch?v=VZYJjMNfQ-I

 

 

Gosh, I'm aware that this has traveled far away from my original post, and it's probably only unfortunate circumstances / bad timing, i.e. not even related to the EGR cleaning, but I really hope to find an happy ending.

 

Yeah, @Wino, I'll do the test tomorrow. I remember doing the test not long after me and my lady bought it. All cylinders showed approximately 13 kg/cm² which I believe is pretty much normal. Hopefully it'll be the same now. What could compression (or lack of such) tell me? Could I use results to rule out cable harness / wire loom, for instance?

Edited 8 November, 2015 by gewstarr

[sepulchrave]

The fuel pump is in the tank and not easily accessible. The filter will be fine, there is a MASSIVE amount of nonsense talked about fuel filters, any loss of pressure through a dirty filter would be very gradual and very hard to notice indeed.

[Breezy_Pete]

If you take out the rear seat, access to the fuel pump is through a panel there, from the top of the tank.

 

As far as I know a petrol fuel pump is either on or off. I don't think there's any control that VCDS could have anything to do with. 

 

Before your misfires became more general, rather than just cyl 1, I thought that a compression test would be useful to rule out a mechanical problem in cylinder 1.  I think now that the problem seems to have moved/spread, it is unlikely to be a mechanical cause. 

 

If the fuel pump wasn't managing to deliver enough fuel to the injectors, it might result in lean misfiring. 

 

Did you tighten up the EGR connection again after your video experiment?

 

What is that thick red wire going across your engine, by the way?

[felicia16v]

To be honest, you need to start at the beginning as I recon you now have an air leak after messing with the throttle body and egr system.

start by pulling the throttle body back off, and the egr spacer plate, make certain everything is spotless especially where the gaskets sit and clean the actual throttle body inside around the throttle flap, this will be covered in carbon deposits and oil from the egr and breather system, make sure everything is clean and refit carefully, then before starting the engine, do adaption 00 save, then basic settings 60 to adapt the throttle body, sometimes can take a few goes at this

and basic settings for the egr valve (cant remember if its 70 or 74).

also check with the engine running for leaks on the brake servo pipework they usually split where the joins are.

[gewstarr]

Hey guys, due to work I had to put the project on hold for a day, but at least today I have some progress as for what troubleshooting goes.

I first swapped coils and plugs a bit randomly. Still getting a pretty consistent fault on cylinder 1 and 2. Then I removed all plugs and coils, and went for that compression tester I mentioned earlier. I'm not 100% sure about the correct way of doing this, but I screwed the thing on the first cylinder, then went in and held ignition in three cycles, approx 6 seconds each. Then I read results, pressed the release button so it got back to 0 bar, unscrewed, and went on to the next cylinder. After all four were done, results turned out like this.

#1)

#2)

#3)

#4)

Around 17 bar (~250 psi) on all cylinders. Is this to be considered normal for a 2003 BBY engine? If my head gasket were blown, would it be likely for compression to be lower, or more uneven? It would be a great relief if this pretty much rules out the possibility of mechanical cylinder failure. Of course, I'm no mechanic, so I wouldn't know.

I'm thinking about the MAP sensor as well, but it's weird that it seems to be mostly cyl. #1 & #2 that are affected. If the air pressure sensor was bad, wouldn't the misfire even out between all four cylinders? The same fact pretty much brings my theory about the fuel pump into weak stuff. One thing I find weird is the suddenness of the failure. Car started okey, got of the parking lot driveway, then bam, and she has not been the same since.

When I reattached the MAP sensor after some quick cleaning, I only used one of the two T15 screws holding it to the manifold, since the lower one sat bad and I wasn't sure if it was worth going the mile since I could be removing the sensor again anytime. So, it could be that it's not 100% vacuum there right now, yet almost. Feels unlikely to be the cause though, since situation evolved before I had removed the sensor.

Ideas? [emoji1]

Edited 11 November, 2015 by gewstarr

[sepulchrave]

Your compression is healthy enough across all cylinders so it isn't that.

 

It's electrical.

 

The MAP sensor will also be fine.

 

Did you bleed the fuel rail when you reassembled it after cleaning the injectors?

[gewstarr]

You mean pressing the Schrader valve under the blue little cap on the right of the fuel injector rail? Nah, I skipped that step. However, been running/troubleshooting the car on idle for so long after so any air pocket(s) should have vanished, I mean been passed through the system, back to the tank or wherever air lands. Even if not, could this cause this engine? My gut feeling tells me this isn't the source of the problem.

Actually, I'm starting to move back against the idea of a broken EGR valve. After all, this all started a couple of days after my intense ultrasonic cleaning of this piece. I've dug up dozens of articles on vacuum leaks and most agree that often the first or first plus second cylinder is affected, which is exactly my case.

In fact, I'm guessing my EGR is stuck open. It failed to operate after starting that car that day, and it took ~50 meters of driving (just of the parking) before the surplus air made impact, starting to cause misfire, ECL, etc. This is a theory only. But hey, earlier during troubleshooting I had the car on it's rough idle, and the unplugged the EGR live, and there was absolutely no audible/notable difference in engine performance. Plugged, unplugged. Plugged, unplugged, engine still running exactly the same. If you have a healthy intake system (and EGR valve), is it supposed to be like this, or will you more or less note an instant difference? If so, that's even more (substantial) evidence that my EGR had failed.

Actually, this morning I decided to try something wild and crazy. I took a paring knife from the kitchen drawer and went out with it. I unscrewed the copper pipe from the inlet/throttle housing just enough to fit the knife blade in there. Plan was for it to be a wall, i.e. simulating a fully closed valve. Started the engine. It behaved somewhat difference. I had at least ½ smooth running for ~1 second, then rough running for the next second, and then loop back to a second of smoother running, and so on. Looked in the engine bay. It was whistling from where I had placed the paring knife blade. Obviously my troubleshooting solution was far from airtight.

Is this something to spin further upon? Someone with a developing recipe? I thought about removing the valve itself and put 12v/GND on pin 1/5, but I'd rather not use a direct power source such as the car battery. Anyone with intel on approx how many ampere I'll need to see valve movement?

Skickat från min GT-I9506 via Tapatalk

[Breezy_Pete]

I'm not sure, but I think EGR may not be active at idle on petrol engines, so unplugging the valve then wouldn't make a difference.

 

I think just 1 or 2 amps will be plenty to move the valve. What you should probably do though, is use VCDS.  It has an 'actuator test'* mode where you cycle through operating things like the fuel pump, EVAP purge valve, EGR valve listening for them to work at each stage.  I confirmed a faulty EGR valve on our 1.2 by doing this actuator test a couple of times.  Absence of a heavy 'click' noise at the time movement is requested suggests a fault situation.

 

You could also measure the dc resistance of the solenoid between pins 1 and 5, if you can get probes into the connector. I can't remember now what is typical but maybe 6-12 Ohms territory?

 

* may not be the exact name, but it's something similar.

Edit: Looking at one of your photos up there ^, maybe it's "Output tests".

Edited 12 November, 2015 by Wino

[sepulchrave]

You mean pressing the Schrader valve under the blue little cap on the right of the fuel injector rail? Nah, I skipped that step. 

 

Oh, well that's what it's for, that and for attaching a fuel pressure gauge for diagnostics.