STI 7 : important information about engine failure
#1
Scooby Regular
Thread Starter
iTrader: (1)
Join Date: Mar 2002
Location: Norn Iron
Posts: 2,786
Likes: 0
Received 0 Likes
on
0 Posts
No offense, but your results sound very strange. If for example your car was running 1 bar, and having a lambda value of just 0.99, then I am pretty sure you would have melted your pistons already.
It is too weak for a full throttle mixture, let alone a car running 1 bar.
I also find it impossible to believe that a jump from 1 bar to 1.1 would cause the reading to drop from lambda=0.99 ( 14.5:1 AFR approx ) to lambda=0.77 ( 11.4:1 AFR approx ), although this would be a safe full throttle mixture reading.
What wideband system are you using to take the measurements with ??
I also assume than an STI7 has a cat in the up-pipe, which would also cause any sensor placed after it ( where you claim to have your sensor ) to give innacurate readings, unless I am mistaken.
[Edited by ustolemyname??stevieturbo - 4/3/2003 7:06:26 PM]
It is too weak for a full throttle mixture, let alone a car running 1 bar.
I also find it impossible to believe that a jump from 1 bar to 1.1 would cause the reading to drop from lambda=0.99 ( 14.5:1 AFR approx ) to lambda=0.77 ( 11.4:1 AFR approx ), although this would be a safe full throttle mixture reading.
What wideband system are you using to take the measurements with ??
I also assume than an STI7 has a cat in the up-pipe, which would also cause any sensor placed after it ( where you claim to have your sensor ) to give innacurate readings, unless I am mistaken.
[Edited by ustolemyname??stevieturbo - 4/3/2003 7:06:26 PM]
#2
I'm glad to share with you the result of some runs reading a wide band oxygen lambda directly plugged directly after the turbo on my STI7.
You will be surprised with the result we have read.
First of all, for those who are not specialists (I'm not too :-I ):
– the 1.00 value of oxygen is the one we can read when the motor runs very poor, it corresponds to a 14,7 air/fuel value.
If we read for example a 0.75 value from the Lambda, the mix is rich.
The most important thing we have read is this one:
– When we run with constant throttle at the pressure of 1 bar, the lambda gives a value of 0.99 ! So the mix is absolutely too poor and at this point of charge it's very dangerous for the pistons.
As someones said that the failures appeared running at constant throttle, if it was at the pressure of 1 bar, my info could help to understand why.
Over 1 bar, no problem, see the following point...
Second important thing:
– Apart the case above, when we run full throttle at any gear or rpm, between 1 bar and 1.2 bar, the value goes to 0.77, so it's suddenly too rich !... and NOT too por as supposed.
As the rumor about the STI motor failures were also supposed to come from the fuel pump debit, we have seen that it's not the case as at higher rpm, the values are very rich; it indicates that the pump delivers the lot of fuel ordered by the ECU.
It seems that Subaru wants to have a rich mixture to avoid problems at high boost BUT if the constant runs @ 1 bar are so poor, they don't avoid the pistons slaps risks. Are they crazy ?
So, those were the mains things I wanted to share with you, hope those will help !
P.S. Those readings come from my Continental (Swiss) STI7, as you have a different ECU ref. #, I can't assure that your cars run with the same mixes, but it's probably the case as we have the same problems.
You will be surprised with the result we have read.
First of all, for those who are not specialists (I'm not too :-I ):
– the 1.00 value of oxygen is the one we can read when the motor runs very poor, it corresponds to a 14,7 air/fuel value.
If we read for example a 0.75 value from the Lambda, the mix is rich.
The most important thing we have read is this one:
– When we run with constant throttle at the pressure of 1 bar, the lambda gives a value of 0.99 ! So the mix is absolutely too poor and at this point of charge it's very dangerous for the pistons.
As someones said that the failures appeared running at constant throttle, if it was at the pressure of 1 bar, my info could help to understand why.
Over 1 bar, no problem, see the following point...
Second important thing:
– Apart the case above, when we run full throttle at any gear or rpm, between 1 bar and 1.2 bar, the value goes to 0.77, so it's suddenly too rich !... and NOT too por as supposed.
As the rumor about the STI motor failures were also supposed to come from the fuel pump debit, we have seen that it's not the case as at higher rpm, the values are very rich; it indicates that the pump delivers the lot of fuel ordered by the ECU.
It seems that Subaru wants to have a rich mixture to avoid problems at high boost BUT if the constant runs @ 1 bar are so poor, they don't avoid the pistons slaps risks. Are they crazy ?
So, those were the mains things I wanted to share with you, hope those will help !
P.S. Those readings come from my Continental (Swiss) STI7, as you have a different ECU ref. #, I can't assure that your cars run with the same mixes, but it's probably the case as we have the same problems.
#3
A standard STi7 doesn't have a cat in the uppipe - it only the WRX that has the delights of one.
Assuming that the information is correct, running at the stoichiometrically correct ratio at 1 bar doesn't automaticly mean that you will melt the pistons, but you will run high EGT's. Is this for wide open throttle when the ECU is in open loop mode, or part throttle when the ECU is running in closed loop mode and hence the Lambda sensor will shoot for a target reading?
Presumably Subaru checked this during the mapping process. The strange thing however is that their normal practice is to run rich at full throttle, partly for the fuel cooling effect and partly because max power is normally generated with air fuel ratios richer than stoiciometrically correct. Hence the question above about part or full throttle.
Duncan
[Edited by BugEyed - 4/3/2003 7:28:06 PM]
Assuming that the information is correct, running at the stoichiometrically correct ratio at 1 bar doesn't automaticly mean that you will melt the pistons, but you will run high EGT's. Is this for wide open throttle when the ECU is in open loop mode, or part throttle when the ECU is running in closed loop mode and hence the Lambda sensor will shoot for a target reading?
Presumably Subaru checked this during the mapping process. The strange thing however is that their normal practice is to run rich at full throttle, partly for the fuel cooling effect and partly because max power is normally generated with air fuel ratios richer than stoiciometrically correct. Hence the question above about part or full throttle.
Duncan
[Edited by BugEyed - 4/3/2003 7:28:06 PM]
#4
First, the STI has no cat uppipe , the '01 and '02 WRX, yes.
I agree, it's incredible but it's true. Thoses results are the sames than another STI7 with 1000 km, so a quite new one !
I had a discussion about the STI problems with a Ecutek agent and he told me about the poor mix @ 1 bar, saying me to be very careful. As I said it was absolutely unbelievable for me, he proposed to me to run my car with the Lambda...
During the measures, one time I was driving and one time as passenger, the Motec screen was giving the sames ratios. I don't know the brand of the Lambda itself but it seems that the full package was purchased from Motec.
I agree, it's incredible but it's true. Thoses results are the sames than another STI7 with 1000 km, so a quite new one !
I had a discussion about the STI problems with a Ecutek agent and he told me about the poor mix @ 1 bar, saying me to be very careful. As I said it was absolutely unbelievable for me, he proposed to me to run my car with the Lambda...
During the measures, one time I was driving and one time as passenger, the Motec screen was giving the sames ratios. I don't know the brand of the Lambda itself but it seems that the full package was purchased from Motec.
#5
I'm alsmost certain I've seen similar behaviour using delta dash - but just ignored it because of the narrow band lambda in the car.
btw stevie, don't be so agressive...the guy posted specifically because what he found is different from the current views on the subject...It's called sharing new information
btw stevie, don't be so agressive...the guy posted specifically because what he found is different from the current views on the subject...It's called sharing new information
#6
Thanks Bruce, that was exactly my meaning.
Let me be surprised about the few reactions about this post... big affair when the subject is a rumor (as the fuel pump) and yet, when there is a new and serious way, big silence !.. and big amount of visits certainly
Let me be surprised about the few reactions about this post... big affair when the subject is a rumor (as the fuel pump) and yet, when there is a new and serious way, big silence !.. and big amount of visits certainly
#7
Cruising will be at vacuum, not 1 bar unless we are talking insanely high speeds to hold for any length of time.
I dont deny you have seen these results on whatever car you tested, but the commonly accepted cruise followed by failure has occurde at cruises where the manifold pressure is in vaccum, and also the older cars have an entirely different ecu and map.
How many sti 7s have been dying in the same wayas our old shapes?
I dont deny you have seen these results on whatever car you tested, but the commonly accepted cruise followed by failure has occurde at cruises where the manifold pressure is in vaccum, and also the older cars have an entirely different ecu and map.
How many sti 7s have been dying in the same wayas our old shapes?
Trending Topics
#8
OK Adam, but it's possible to run, for example, in 4th gear @ 1 bar when you are waiting to pass another car. During this time, your motor runs effectively with a big amount of EGT, the pistons don't like that.
How to be absolutely sure about the real and exact conditions of the engines failures ? When this sort of thing appears, the driver is not concentrate to register the conditions of a failure.. not ?
How to be absolutely sure about the real and exact conditions of the engines failures ? When this sort of thing appears, the driver is not concentrate to register the conditions of a failure.. not ?
#9
Scooby Regular
Join Date: Sep 2000
Location: London
Posts: 7,039
Likes: 0
Received 0 Likes
on
0 Posts
Can you tell us what RPM we are talking about? I can't think where I'd use 1bar of Boost on my STi with a small throttle opening, and even if I could, I would imagine it wouldn't be for long - in the waiting to overtake example, I'd expect to see very little boost until I started the move, at which point I'd floor the throttle?
Perhaps this is an issue in the transition from Closed to Open Loop?
Interesting spot and thanks for bringing it up!!
Ta,
Matt
Perhaps this is an issue in the transition from Closed to Open Loop?
Interesting spot and thanks for bringing it up!!
Ta,
Matt
#11
On Delta Dash,I'm sure I've seen BIG lambda reading differences between open and closed loop control...but just ignored it, will have a closer look.
Matt, the ECU ignores Lambda at high revs (and probably just goes to some state where it is sure it's running rich) so I'm sceptical about what Delta Dash will show up (if the ECU ignores the reading, you can't really trust it yourself, can you?).
Although you'll probably mostly run above or below 1 bar, if these observations can be verified (by someone in the UK - scoobynet seems to suffer from the Not-Invented-Here syndrome ) it could be very helpful (especially for guys running dawes device /EBC who are exposed to low octane fuel) - It seems better/safer to set hold boost to slightly above 1bar.
Matt, the ECU ignores Lambda at high revs (and probably just goes to some state where it is sure it's running rich) so I'm sceptical about what Delta Dash will show up (if the ECU ignores the reading, you can't really trust it yourself, can you?).
Although you'll probably mostly run above or below 1 bar, if these observations can be verified (by someone in the UK - scoobynet seems to suffer from the Not-Invented-Here syndrome ) it could be very helpful (especially for guys running dawes device /EBC who are exposed to low octane fuel) - It seems better/safer to set hold boost to slightly above 1bar.
#12
As it was not possible to ask for registering the runs with a notebook connected to the Motec, we were only able to see the datas on the digits of the Motec...
I assume that in some conditions from full throttle to some less throttle when you are, for example, in 5th or 6th gear you can go to 1 bar and stay at this pressure. I repeat that this is not the case when you accelerate @ full throttle until 1.2 bar without the 1 bar step.
When you are, as us, in hills open roads conditions for the test, you have to watch also for the others cars etc. Trying to repeat each time the same thing is giving a lot of stress, you can do it, but for a short lap of time... on track, it would be much easier :-))
The same day, I have also seen the main 3D maps of the Subaru STI's ECU, it's really a pity in many cases, and you could be affraid by the one regarding the timing.
Before the runs with my car, I have tested an Ecutek modified STI7 and really, i'ts a missile ! 350 DIN Horsepower and 488 Nm @ 3550 rpm. The motor runs smoothly and the power comes "without steps" all over the rpm... and the car runs with it original fuel pump ;–)
I assume that in some conditions from full throttle to some less throttle when you are, for example, in 5th or 6th gear you can go to 1 bar and stay at this pressure. I repeat that this is not the case when you accelerate @ full throttle until 1.2 bar without the 1 bar step.
When you are, as us, in hills open roads conditions for the test, you have to watch also for the others cars etc. Trying to repeat each time the same thing is giving a lot of stress, you can do it, but for a short lap of time... on track, it would be much easier :-))
The same day, I have also seen the main 3D maps of the Subaru STI's ECU, it's really a pity in many cases, and you could be affraid by the one regarding the timing.
Before the runs with my car, I have tested an Ecutek modified STI7 and really, i'ts a missile ! 350 DIN Horsepower and 488 Nm @ 3550 rpm. The motor runs smoothly and the power comes "without steps" all over the rpm... and the car runs with it original fuel pump ;–)
#13
These STi-7's are some very delicate cars.....unreal how many different theories are out there.
Some replace the fuel pump when having a remmap done.....some don't
Some knock after a remmap.....some don't
Some say the standard map is ok.....some say it is what is causing some of the engine rebuilds
I know it is very hard to pinpoint exactly what is going on; however, as a consumer / end-user that has limited technical knowledge I would just like to know for example if I get a remmap and decatted zorst that my car will be safe if correct fuel is used and ambient conditions don't change significantly.... this doesn't seem to be the case as a K/L looks to be the next step but then the pros are even saying now that K/L is not really the be all and end all and one should have some det cans tested on it....where does it end ? Are cars always this hard to tune etc ?? I would have though that with people runnning over 400 bhp on STI engines that running 330 - 340 bhp on a STI-7 would be fairly straight forward and safe with a remmap....
Trust me I am not complaining, I would just love to know that my car was safe after pumping thousands into it, taking all feasible precuations and not asking for ridiculous bhp.
Anyway, for those out there that keep on finding new info keep posting as it is a big help to all and thanks again to all the experts for their continued input
Cheers,
Ray
Some replace the fuel pump when having a remmap done.....some don't
Some knock after a remmap.....some don't
Some say the standard map is ok.....some say it is what is causing some of the engine rebuilds
I know it is very hard to pinpoint exactly what is going on; however, as a consumer / end-user that has limited technical knowledge I would just like to know for example if I get a remmap and decatted zorst that my car will be safe if correct fuel is used and ambient conditions don't change significantly.... this doesn't seem to be the case as a K/L looks to be the next step but then the pros are even saying now that K/L is not really the be all and end all and one should have some det cans tested on it....where does it end ? Are cars always this hard to tune etc ?? I would have though that with people runnning over 400 bhp on STI engines that running 330 - 340 bhp on a STI-7 would be fairly straight forward and safe with a remmap....
Trust me I am not complaining, I would just love to know that my car was safe after pumping thousands into it, taking all feasible precuations and not asking for ridiculous bhp.
Anyway, for those out there that keep on finding new info keep posting as it is a big help to all and thanks again to all the experts for their continued input
Cheers,
Ray
#14
Having multiple discussions with Scooby50STI, before and after his wide band lambda tests (on two different STI7 cars, same results).
The results are :
- WOT from 2000RPM up to 7000RPM, no problems seen : lean (14.7:1) before boost reaches 0.5 bar (close-loop), then rich (down to 10:1) up to 7000RPM (open-loop). That whatever the gear ratio is.
- at part throttle (lets say 50%) : the same perfect behaviour, except when boost pressure reaches 1 bar where the fuel mixture becomes suddenly lean (14.7:1 !!!) for few seconds before comming back rich again !!!
- at part throttle, trying to keep boost pressure close to 1 bar : fuel mixture stays lean (14.7:1 !!!), whatever the gear ration and the RPM are.
My opinion :
1 - there is no fuel pump problem as we can saw that during WOT, even at high rev, the fuel mixture stays rich.
2 - Do not drive at part throttle and 1 bar (or close to 1 bar) of boost pressure.
3 - when accelerating, always perform a WOT and not a small amount of throttle increase.
No doubts that 1 bar and 14.7:1 fuel mixture will lead to an engine break very quickly. Anybody driving its car in an "ON/OFF" manner will have no problems. Anyone driving very smoothly should have problems!!!
For informations, I know a third STI7 which has been tuned by Fochesato a year ago (remapping using its own software). At that time, nobody was talking of any potential fuel pump failure.
He performed its remapping, tuned it using a wide band lambda sensor and knock sensor, and the car performs 345 HP and 50 Nm at 1.4 bar (full decat 3" exhaust).
It is now 30000km that the car runs with this remapping, without any problems. The guy driving the car use to drive on race tracks and pushes his car at the limits, especially on tracks such as French Dijon and Magnycourt where top speed can be up to 230/240 km/h. The car still runs fine.
For all these reasons, I'm confident in the stock fuel pump. The remapping must only adjust fuel mixture and timming regarding to the 1 bar/part throttle response.
If anyone can confirm, or re-do the same kind of test, it will be very useful and helpful.
Xavier
The results are :
- WOT from 2000RPM up to 7000RPM, no problems seen : lean (14.7:1) before boost reaches 0.5 bar (close-loop), then rich (down to 10:1) up to 7000RPM (open-loop). That whatever the gear ratio is.
- at part throttle (lets say 50%) : the same perfect behaviour, except when boost pressure reaches 1 bar where the fuel mixture becomes suddenly lean (14.7:1 !!!) for few seconds before comming back rich again !!!
- at part throttle, trying to keep boost pressure close to 1 bar : fuel mixture stays lean (14.7:1 !!!), whatever the gear ration and the RPM are.
My opinion :
1 - there is no fuel pump problem as we can saw that during WOT, even at high rev, the fuel mixture stays rich.
2 - Do not drive at part throttle and 1 bar (or close to 1 bar) of boost pressure.
3 - when accelerating, always perform a WOT and not a small amount of throttle increase.
No doubts that 1 bar and 14.7:1 fuel mixture will lead to an engine break very quickly. Anybody driving its car in an "ON/OFF" manner will have no problems. Anyone driving very smoothly should have problems!!!
For informations, I know a third STI7 which has been tuned by Fochesato a year ago (remapping using its own software). At that time, nobody was talking of any potential fuel pump failure.
He performed its remapping, tuned it using a wide band lambda sensor and knock sensor, and the car performs 345 HP and 50 Nm at 1.4 bar (full decat 3" exhaust).
It is now 30000km that the car runs with this remapping, without any problems. The guy driving the car use to drive on race tracks and pushes his car at the limits, especially on tracks such as French Dijon and Magnycourt where top speed can be up to 230/240 km/h. The car still runs fine.
For all these reasons, I'm confident in the stock fuel pump. The remapping must only adjust fuel mixture and timming regarding to the 1 bar/part throttle response.
If anyone can confirm, or re-do the same kind of test, it will be very useful and helpful.
Xavier
#19
Excellent work guys.
Thoughts from the pros please: Pat, Pete, Bob, John, etc
Cheers,
Ray
Pat and Pete, I assume that one of your remmaps takes care of the part throttle and 1 bar lean situatiion ?
Thoughts from the pros please: Pat, Pete, Bob, John, etc
Cheers,
Ray
Pat and Pete, I assume that one of your remmaps takes care of the part throttle and 1 bar lean situatiion ?
#20
I am also confused about the fuel pump issue!!! Having used a Standard Australian Spec STi on the track I found fuel consumption to shoot up to close to 1L per 1KM. How is this possible with an inadequate fuel pump?
#21
The fuel pump on the STi 7 varies/switches between 6-12v iirc depending on load, I think it would be worthwhile to check or log what its actually doing at those points in the map.
#23
Interesting.As others have mentioned its difficult to see where you would cruise at part throttle at 1 bar.Would be useful to get some input from the Ecutek tuners ie Bob/Pat/John etc
#24
I have repeated this afternoon the way to run @ 1 bar continuously.
Not so difficult to do, but easier when I drive on a road going to be slighly climbing : stay in 6th gear, the motor does'nt suffer but is going to run @ 1 bar quickly and if you stay at the same throttle position even for some seconds... it's lean time !
That was only one of the ways to stay @ 1 bar.
Not so difficult to do, but easier when I drive on a road going to be slighly climbing : stay in 6th gear, the motor does'nt suffer but is going to run @ 1 bar quickly and if you stay at the same throttle position even for some seconds... it's lean time !
That was only one of the ways to stay @ 1 bar.
#25
If the pros could answer ?
I've had some e-mail exchanges with john Banks, but it was before discovering this lean stage @ 1 bar.
One thing I'm now defenetly sure is that the fuel pump is not causing any problem, nor is the electronic switch driving it.
I've had some e-mail exchanges with john Banks, but it was before discovering this lean stage @ 1 bar.
One thing I'm now defenetly sure is that the fuel pump is not causing any problem, nor is the electronic switch driving it.
#26
Its more than possible for this to happen on a maf based car, I always richen the fuel map up to compensate for this, I am a great believer in good part throttle boost response as, quite often, the car is quicker like this under overtaking circumstances than at wot. Its really the main down side, the same load point can be hit at part throttle as at wot.
So yes I agree that the problem exists, but its not new, for me its just part of the remapping process, adjust the wastegate duty cycles so that you need to use more throttle to achieve the boost, car is not so much fun though and its easily corrected when mapped.
bob
So yes I agree that the problem exists, but its not new, for me its just part of the remapping process, adjust the wastegate duty cycles so that you need to use more throttle to achieve the boost, car is not so much fun though and its easily corrected when mapped.
bob
#27
Scooby Regular
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
I will look at an STi when I next can and see if I can reproduce it. Does it not also show on the factory pseudowideband?
I completely rewrite the boost tables as they are a rather funny in standard form, but also like it sounds Bob is doing go for more rather than less boost on part throttle, with mixtures to match.
I completely rewrite the boost tables as they are a rather funny in standard form, but also like it sounds Bob is doing go for more rather than less boost on part throttle, with mixtures to match.
#28
100% certain there is not a fuel pump or fuel switch problem ??? How can this be, all the pros say change them on here when upping the power on the STI-7's and higher, not to mention that the PPP now comes with an uprated fuel pump.....I would think Prodrive know a thing or two...
Cheers,
Ray
Cheers,
Ray
#29
Hi RT,
I will have to do a Delta dash run and let you know whats up.....I just think it is going to be hard to hold 1 bar as you say....
What paramaters should I have hecked on the DD when doing this test ?
Oh, yea I have a JDM STI-7 with full SS decat zorst turbo back, 1.25mm restrictor soon to install a new pump and ECU plug-n-play map
Cheers,
Ray
I will have to do a Delta dash run and let you know whats up.....I just think it is going to be hard to hold 1 bar as you say....
What paramaters should I have hecked on the DD when doing this test ?
Oh, yea I have a JDM STI-7 with full SS decat zorst turbo back, 1.25mm restrictor soon to install a new pump and ECU plug-n-play map
Cheers,
Ray
#30
RT, I think your car is a Japan's one. And an RA type.
Maybe the problem we pointed out appears on Euro cars only, due to the specific MAP required to pass the Euro 5 vehicule emission.
The test you've done between 0.95 bar and 1.05 bar should have give you a lean lambda value as it did for us on two different stock cars.
Can someone having an STI7 Euro type can do the test again ?
And the fact that two remapped cars providing 350 HP each perform well since more than 25000km makes me confident in the stock fuel pump.
Xavier
Maybe the problem we pointed out appears on Euro cars only, due to the specific MAP required to pass the Euro 5 vehicule emission.
The test you've done between 0.95 bar and 1.05 bar should have give you a lean lambda value as it did for us on two different stock cars.
Can someone having an STI7 Euro type can do the test again ?
And the fact that two remapped cars providing 350 HP each perform well since more than 25000km makes me confident in the stock fuel pump.
Xavier