Jump to content


Photo

Weird Boost Fluctuations/Surging

turbo wastegate surging

  • Please log in to reply
23 replies to this topic

#1 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 09:43 AM

Got another minor gremlin,

 

For a long time my car has had a weird boost fluctuation at part throttle and low boost. It happens at any speed but needs just enough load in the primary turbo rev range to be coming onto boost.

 

The issue can be easily reproduced during highway driving up a long hill. If I gradually accelerate or hold speed it will fluctuate up and down 2 or 3 psi every second or so. It's not a dramatic effect but is noticeable.

 

Throttle position (about 20%) and boost level (0-5psi) need to be just right. That's why I notice it on hills as it's not often you drive steadily in that boost range. Any extra or less throttle causes it to stop. If I continue to hold the throttle steady it will just keep pulsing until I need to slow down, as it tends to pick up speed.

 

I had hoped a retune would solve it, but alas it seems to be mechanical or sensor related. On the logs I can see the secondary waste gate opening and closing in sequence as it happens. I assume the secondary waste gate refers to the ECV and correct me if I'm wrong but this model uses the ECV for boost control. I know if I block off the positive ECV port the fluctuations don't happen. The ECV is now ported but it did do this on a stock one too.

 
I suspect some kind of lag between the ECV opening and a sensor seeing the loss of boost. As a result the ECV is opening and closing too slowly. Another idea is the ECV is opening too much initially and losing lots of pressure, when the ECU sees this it shuts the ECV and gains too much pressure, so the cycle repeats.
 
I've attached some logs with one 'normal' log of it not doing it so you can see the difference.
 
Any ideas?

Attached Files

  • Attached File  Logs.zip   10.12KB   3 downloads

The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#2 TSG

TSG

    Cured

  • Members
  • PipPipPip
  • 2,569 posts
  • Gender:Male
  • Location:Brisbane

Posted 28 September 2016 - 09:57 AM

I think others have experienced surging boost at highway speeds. I did once or twice. No idea how to fix it. But if you want to tighten up the ECV, you can try the 'ADJUSTING ECV OPERATION ROD' instructions here.

 

http://www.subyclub....-to-twin-turbo/

 

My thread on the ECV issue I had with detailed adjustment instructions:

http://www.subyclub....hl=+tsg +rattle



#3 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 10:13 AM

Thanks TSG.

 

I should have mentioned. On the old ECV I tried tightening it. In fact I tried it tightened all the way up and I don't think there was much of a change.


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#4 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 11:52 AM

Is this at about 2.5-3k ?  I get the same thing. Almost a sort of boost stutter.

 

I think you're interpretation of the 'secondary wastegate' is correct (ECV).

 

Another thought is that its cycling in and out of closed loop AF ? Boost up -> open loop -> boost down -> closed loop -> rinse and repeat ?



#5 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 12:02 PM

Ignore above - just looked at the log

 

I notice that the fuel pump duty is also cycling from 30% to 60% when the problem occurs.

 

Is your fuel pump up to scratch ?  Maybe its having problems at 30% DC and the fuel rail pressure is dropping.. get your tuner to flip it to the higher duty cycle earlier and see if that helps.



#6 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 01:03 PM

Ignore above - just looked at the log

 

I notice that the fuel pump duty is also cycling from 30% to 60% when the problem occurs.

 

Is your fuel pump up to scratch ?  Maybe its having problems at 30% DC and the fuel rail pressure is dropping.. get your tuner to flip it to the higher duty cycle earlier and see if that helps.

- I did think about that, but surely it would play up under other circumstances as well? It is tied closely to that specific amount of boost and I'm 99% sure it has something to do with the positive port on the ECV. When it is blocked off it behaves as expected... Oh, and if it was the pump I'd expect to see the same thing on part throttle in the secondary boost range? (Not that I ever drive at 20% throttle in secondary mode :sarcastic:)

 

I have read that some ECUs have a delay between Open and Close loop so your other theory makes sense.


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#7 Jimbo

Jimbo

    Livin' the Legacy

  • Members
  • PipPipPip
  • 2,771 posts
  • Gender:Male
  • Location:B'town

Posted 28 September 2016 - 01:08 PM

Sounds like my problem. Try to hold a steady speed off boost and it shudders until you apply boost and its fine.

I thought along the lines of switching between closed/open loop
Sounds like my problem. Try to hold a steady speed off boost and it shudders until you apply boost and its fine.

I thought along the lines of switching between closed/open loop

'04 Liberty GT Premium "The Sprog Chariot"


#8 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 01:21 PM

Sounds like my problem. Try to hold a steady speed off boost and it shudders until you apply boost and its fine.

I thought along the lines of switching between closed/open loop
Sounds like my problem. Try to hold a steady speed off boost and it shudders until you apply boost and its fine.

I thought along the lines of switching between closed/open loop

- I see what you did there...

 

Mine doesn't shudder as such. It's a smooth fluctuation as if I'm gently on and off the throttle. Rhythmic almost.


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#9 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 02:23 PM

yeh -- pretty slow cycling, as you say.  That is certainly consistent with the ECV response time.

 

The logs don't show OL/CL cycling.. unless I misinterpreted them.



#10 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 02:50 PM

yeh -- pretty slow cycling, as you say.  That is certainly consistent with the ECV response time.

 

The logs don't show OL/CL cycling.. unless I misinterpreted them.

- I forgot that OL-CL weren't cycling.

 

Is the ECV slow to respond in general?

 

I find it interesting that I can see a difference between 25.88% and 26.27% on the ECV duty, but there isn't any percentages between 0% and 25.88%.

 

I suspect the fuel pump switching between 33% and 66% duty might be symptom rather than a cause. I can imagine the ECU would be altering the amount of fuel going in when it sees the boost fluctuations.


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#11 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 03:21 PM

Oh - I didn't look at scale --  25.9% to 26.3% should be fork all of nothing.



#12 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 03:36 PM

Oh - I didn't look at scale --  25.9% to 26.3% should be fork all of nothing.

- I think I didn't explain my point well enough. Between 25.9% to 26.3% is not the fluctuation. The fluctuation is between 0% and 25%. I was just noting that it is interesting that I can see such a trivial difference in some of those readings but no readings fall between 0% and 25%, e.g. 10% open.


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#13 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 04:22 PM

oh. Ok.

 

Probably just has 'off' and 'on/more-on/more-more-on'.

 

Sort of like the primary DC never getting above 90% - to avoid overheating the solenoid.



#14 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 28 September 2016 - 04:29 PM

I'm going to put a spare restrictor pill into the top ECV port. It might make things worse by slowing the ECV response time even more, but on the off chance it might respond more smoothly and predictably. Can't hurt any more than blocking the port entirely.


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#15 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 05:41 PM

That would probably at least confirm your suspicions as to the cause - if it slows the cycling.

 

I think the behaviour I get is too fast to be the same thing.



#16 Jimbo

Jimbo

    Livin' the Legacy

  • Members
  • PipPipPip
  • 2,771 posts
  • Gender:Male
  • Location:B'town

Posted 28 September 2016 - 06:45 PM

I don't think my experience is relating to the ECV as my changeover is instant and faultless.

What's actually happening when the O2 sensor is switching between CL to OL?

Does the Fuel Pump Controller also switch between load percentages at this time? So in CL it's always at 30% DC and then when OL switches it's 60% onward to 100% depending on load?

My understanding is CL is basically running the car-off boost and OL handles the rest.


I'm thinking ECV is unrelated in my case, as my issue is best described as a relatively violent on/off or shuddering between off-boost and on-boost. Possibly rail pressure drop-off, Dunc?

If it was ECV, it would be a gentle surging as you say, the ECV couldn't open and close quickly enough to give harsh shuddering.


Luke maybe in your case you can successfully do the ECV +ive port bypass mod. I know D-E's don't like it much, though. 

Basically, block off the +ive pressure port line coming from the BBOD, then put a tee-piece into the line that goes from the secondary turbo/SBPV bridge. Run that new line to the ECV +ive port.

The +ive ECV port is then fed by the secondary turbo itself, rather than all the solenoid trickery.
 


'04 Liberty GT Premium "The Sprog Chariot"


#17 duncanm

duncanm

    wha' who ?

  • Members
  • PipPipPip
  • 2,002 posts
  • Gender:Male
  • Location:Sydney

Posted 28 September 2016 - 08:45 PM

Jimbo,

 

I dug up some xml's, and the FP ranges are based on RPM and injector duty cycle.



#18 LukeFranky

LukeFranky

    Regular Member

  • Members
  • PipPipPip
  • 587 posts
  • Gender:Male
  • Location:Armidale, NSW

Posted 05 October 2016 - 11:59 AM

The pill didn't help, as to be expected it felt like it increased the delay between cycles. I'm going to try tightening the ECV again.

 

What does a Rev D hate about the ECV bypass mod? What side effects are there?


The B4 Bomber - 2002 MY03 Subaru Liberty B4, ADM, Manual, EJ208. 


#19 Guest_KONG_*

Guest_KONG_*
  • Guests

Posted 05 October 2016 - 01:01 PM

The pill didn't help, as to be expected it felt like it increased the delay between cycles. I'm going to try tightening the ECV again.
 
What does a Rev D hate about the ECV bypass mod? What side effects are there?


Boost is erratic and surges.
Mine hit 25psi.

#20 Guest_KONG_*

Guest_KONG_*
  • Guests

Posted 05 October 2016 - 01:17 PM

I experienced the same gentle surging for a while.
Fairly certain the fix involved shimming and tightening of the BOV spring and running mid to full atmo on a GFB BOV.





Also tagged with one or more of these keywords: turbo, wastegate, surging

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users