SaabCentral Forums banner

Hissing during boost?

3K views 23 replies 9 participants last post by  Solid Snake 
#1 ·
Do you guys hear this on your cars or do I have a leak? The car does not boost into the red, even with the wastegate actuator hose disconnected. Are there common failure points, maybe the recirc valve? The manual says I should be boosting at over 15 PSI (B234R w/TD04). Thanks.
 
#2 ·
With the wastegate hose disconnected you should hit the fuel cut at well over 1 BAR, ie in the red. I'm stage 3 tuned and if I disconnect mine I fly into the fuel cut at 1.54 BAR easy. If you can't hit the fuel cut with it off, I reckon you've got a massive leak. The hissing would suggest that also ;)...
 
#3 ·
ECU will revert to base boost if there is a leak upstream or dwnstream of the turbo in most cases. Therefore, assuming your plugs are good and the DI is not failing the most common cause of base boost is an air leak.My Aero just had the clamps on the throttle body rubber elbow left loose after a service and that caused base boost. Vacuum hoses with leaks will also do that as will a leaking bypass (hooter) valve which I assume you mean when you use "recirc". Technically, "recirc" is an air conditioning function.
 
#4 ·
That valve is known as the recirculation valve on many other cars too (vs a vented blow-off valve).

Anyway, what do you mean by upstream of the turbo? The only pressurized area is between the compressor outlet and intake manifold, correct?

Should I block off the outlet of this hooter valve and see if the hissing continues?

Also, about my boost gauge... atmospheric pressure is a good bit into the yellow at 12:00. It seems the white portion of the gauge ends while it's still in vacuum. Where do your boost gauges sit with the engine stopped?
 
#6 ·
Thanks... that's interesting. So the manual's description of the gauge must not be accurate. If it was, the needle would be right on the white/yellow border when stopped.

There seems to be no difference in the car's behavior with the bypass valve connected or disconnected. It seems suspicious.
 
#8 ·
Solid Snake said:
Thanks... that's interesting. So the manual's description of the gauge must not be accurate. If it was, the needle would be right on the white/yellow border when stopped.

There seems to be no difference in the car's behavior with the bypass valve connected or disconnected. It seems suspicious.
Right, I typed that earlier but when I went out took a closer look. I lied-it sits at about 350 degrees (with 360 being vertical, obviously), with the engine off, which when I looked closely is near enough exactly on the boundry between white and yellow, so where Saab says it should be. Just while sat at my computer I thought the yellow started earlier than it does haha. So my mistake, sorry.
 
#9 ·
Sorry to inject my own question into the thread, but should I assume that any hissing noise is indicative of a leak somewhere? What noise, if any, is normal under boost? My car seems to boost okay still, though I have no quantitative measurements to back it up...
 
#10 ·
Well I'm not an authority on this by any means, just keeping the thread ticking over, but I've owned 3, driven 4 and been in about 6 9000s and I've never heard any hissing under boost. The only hissing you should hear will be a slight hiss of the BOV when you change gear, when you're not on boost. My advice would be to get some qualitative measurement to back it up-ie a boost gauge. People kept telling me, and I wasn't sure if it'd be that useful, but now I couldn't live without it. Any turbo car should come with one from the factory.
 
#11 ·
1) turbo with no hose to the WG = big pressure quickly - there will be no ECU control - if you don,t get this you have a leak or a knackered turbo or the WG actuator pin has come off
2) they don,t cut back to base boost if there is a leak - they just leak
3) they can hiss on partial load throttle = recirc valve is just cracking open
4) any hiss on full throttle is a leak
5) first port of call is the small hose going to the recirc valve, if you are plumbed as per std - then there is a T piece from this hose that goes to the saab plasebo gauge - if the gauge does not swing to the left from the engine off position on start up or after lifting off ( i.e. shows vacuum) = good chance this hose is burst - if it,s burst then you have no pressure assist close on the recirc valve = its open under boost = big leak - try running a hose direct from the inlet spigot to the recirc valve
6) if it,s your IC look for the long yellow worm :) - the plastic end caps can come away from the matrix and hence the big seal pop,s out - easy fixed - glue the seal back in ( repair any broken section with an o-ring splice) - clip the end tank back on and secure it with copious amounts of tie wraps
 
#12 ·
there is a post on here somewhere about all possible causes of base boost... i think this thread should be merged with that one to avoid confusion as to what base boost may be, and also to help rule out areas where you may have leaks... very informative, and should probably be stickied somewhere...
 
#13 ·
My 97 displayed base boost when the throttle body rubber elbow clamps were left loose after a recent service.


My mechanic says the engine will revert to base boost if there is a leak downstream of the turbo. I assume this means a big enough leak for the ECU to "realize" that all is not well.

If this did not happen then a leak of pressurized air might allow or cause the turbo to overspeed with disastrous results. I realize that with a constant leak from start up then the engine will not devlop enough boost to drive the turbo to produce more boost. Still, my engine was definitely at base boost as it would boost to that point on the gauge and no more until I reclamped the rubber elbow.
 
#14 ·
Thanks, that helps a lot. The hiss was pretty endearing till I saw this thread. Did not seem like base boost but it did hiss on full throttle. It's out of service now because it's getting a new radiator, starter and a few other bits looked at. So far some suspect items:
  • Brass fitting on throttle body (going to idle control valve) broke off while the hose was being removed
  • Missing PCV check valve
YoungSaaber said:
Well I'm not an authority on this by any means, just keeping the thread ticking over, but I've owned 3, driven 4 and been in about 6 9000s and I've never heard any hissing under boost. The only hissing you should hear will be a slight hiss of the BOV when you change gear, when you're not on boost. My advice would be to get some qualitative measurement to back it up-ie a boost gauge. People kept telling me, and I wasn't sure if it'd be that useful, but now I couldn't live without it. Any turbo car should come with one from the factory.
Norman Lovie said:
2) they don,t cut back to base boost if there is a leak - they just leak
4) any hiss on full throttle is a leak
 
#15 ·
Yep, it was the fitting on the underside of the throttle body.

I took it out for a spin, stomped it in first gear, the front tires let out an ungodly bark and my neck snapped back :D

It started boosting into the red for a while, hitting FCO, then backed off into the yellow.

Where does that tube on the underside of the TB go?

I'm currently running water injection, so if that thing fills with water, will anything break?
 
#17 ·
OK, water injection is back on at 1.5GPH and i'm running 93 octane fuel. The motor boosts fully into the red. Acceleration is brutal but I have to make an effort to not hit fuel cutoff.

Does the APC unit watch manifold pressure or does it only react to knock? Is it possible there's no knock whatsoever so it continues boosting into fuel cutoff?

Also, where on the car can I tap for a knock signal and what is the signal type?
 
#18 ·
the Trionic monitors manifold pressure and regulates boost accordingly.
to get a knock signal, you can drive a LED from the ECU.
Pin 8 on the Trionic ECU switches to ground when the engine detects knock, so by having a 12v LED (or lower voltage LED with required resistor inline) wired in to a 12v supply and pin 8, it will flash when knock is detected.
you cant really run anything else from that pin on the ecu, as it doesnt have the current handling capacity to do much else than drive a LED.
 
#19 · (Edited)
the Trionic monitors manifold pressure and regulates boost accordingly.
to get a knock signal, you can drive a LED from the ECU.
Pin 8 on the Trionic ECU switches to ground when the engine detects knock, so by having a 12v LED (or lower voltage LED with required resistor inline) wired in to a 12v supply and pin 8, it will flash when knock is detected.
you cant really run anything else from that pin on the ecu, as it doesnt have the current handling capacity to do much else than drive a LED.
Is this pin unused? Where does it go and what was it originally intended to do? How does it compare to the DI's knock wire?

If I want to test this, what's a relatively harmless way to induce knock?
 
#20 ·
Solid Snake said:
I took it out for a spin, stomped it in first gear, the front tires let out an ungodly bark and my neck snapped back :D

It started boosting into the red for a while, hitting FCO, then backed off into the yellow.
Hmmm... So if when i floor it in 1st gear i cant get my wheels to break loose and it only boost into the yellow, not near the red, i could have a leak?

In second and third it gets almost into the red.
 
#21 ·
Hmmm... So if when i floor it in 1st gear i cant get my wheels to break loose and it only boost into the yellow, not near the red, i could have a leak?

In second and third it gets almost into the red.
I've heard the APC limits torque in 1st and 2nd, but if you want to test for a leak, pull your wastegate hose and see if it gets into the red. You'll have no knock protection during this test though.

On mine, it goes PAST the red at around 60% throttle. It'll put down stripes in 1st every time and in 2nd, torque steer is very strong and unless i'm going perfectly straight, it'll squeal the tires again. It feels like I'm not letting the APC system train itself by not flooring it. It tends to bounce from the middle to the end of the red. I'm just controlling boost with my foot right now. If I floor it, it will hit FCO a few times, then back off into the yellow.

Is it safe to say the APC system doesn't detect any knock but is having a hard time limiting boost to the proper 1.08 bar?
 
#22 ·
Solid Snake said:
Is this pin unused? Where does it go and what was it originally intended to do? How does it compare to the DI's knock wire?
this is an unused pin on the ECU wiring harness, you will need to add the wire to the ECU connector to be able to use pin 8.
I do no know how this compares to the DI knock output.

safe way to induce knock? no real thing as "safe" knock, except say, if you filled up with regular unleaded and took it for a bit of a WOT run.. that will definitely induce some knock.
 
#23 ·
once you fit the knock led ( which doesn,t detect it just reacts to the DI signaling the ECU - when knock is detected the ECU grounds pin 8 and hence you get the led to light up)

when ECU gets enough knock counts ( programmable) it goes through a range of measures including boost cut via the APC and ignition retard and swingsa into a richer knock fuelling map
Your apc tries to keep the boost in line with the boost request map - it doesn,t just keep bumping up boost until knock. If your getting boost overshoots to the point of fuel cut - the issue is usually in the apc / actuator hardware - but can also be tamed by changing the ECU APC software control parameters.
Knock etc.. is - in itself not so bad, but left to progress to detonation then things can go bad very quickly.
most engines knock to some degree
from my experience the knock led is a very usefull toy - why - because a lot of knock signals are false - typically due to a dodgy plug - so when you start seeing flash,s = time to change the plugs ( a lot more often than folks consider!) if you don,t - the ECU doesn,t know the plug is bad and it will cut back your performance
Next issue with knock is you tune to the open loop condition - no knock fine - but when the engine is running in closed loop ( lambda) you can get knock even though its at a lower performance - why because in closed loop it leans off the mixture - add some juicy ignition advance etc.. from the open loop tune and hayho knock
I would be very surprised if you have to do anything to induce the led to light up
 
#24 ·
I found a 7% grade a mile long to train the APC on so now it doesn't hit fuel cut, though it does boost to the end of the red. I changed the water pump circuit to inject according to MAP rather than TPS. It starts at 0.6 GPH at 2psi of boost and hits the full 1.5gph at about 10psi. There's no bucking or hesitation.

I'm curious if the water injection would trigger a false knock signal since the DI uses the ionized gap to detect combustion/detonation. Since the water is conductive, it might increase the current transfer, right?
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top