Question to ECU specialists (STi7 & 8)
11 March 2004, 04:32 PM
#1
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Question to ECU specialists (STi7 & 8)
My concern is the erratic knock detection between 5600 and 6200 RPM on STi7 (AG820-2592 ECU).
Following tests I've done on many STi7 with AG820-2591 and 2592 ECU, at reasonnable boost level (1.25 bar) I can see the ECU retarding timming between 5600 and 6200 RPM due to knock detection. But no real knock actually occurs (strange !) in this area. Above 6200 RPM (where knock detection is no more used), the ECU increases timming by about 5 to 10° and again no knock occurs, then you can feel the car take off like an aircraft up to the redline.
What concerns me more is that when I increase boost level to 1.45 bar, the engine starts knocking physically between 5600 and 6200 RPM, again while the ECU retards timming by 5 to 10° (20° absolute timming), and above 6200 RPM, the engine stops to knock while timming become much more agressive (30° absolute timming).
The question is why the engine knocks between 5600 and 6200 with low timmings (20°) while it doesn't above 6200 with stronger timmings (30°) ?!!!
Can the STi's active Valve timming system explain such a behaviour ? In such a case, it would means that Valves are inadequately controlled by the STi7 ECU. Does someone knows at which RPM active valve system activates ?
What about AG821-2593 and STi8 ECU and behaviour ?
Thanks
Following tests I've done on many STi7 with AG820-2591 and 2592 ECU, at reasonnable boost level (1.25 bar) I can see the ECU retarding timming between 5600 and 6200 RPM due to knock detection. But no real knock actually occurs (strange !) in this area. Above 6200 RPM (where knock detection is no more used), the ECU increases timming by about 5 to 10° and again no knock occurs, then you can feel the car take off like an aircraft up to the redline.
What concerns me more is that when I increase boost level to 1.45 bar, the engine starts knocking physically between 5600 and 6200 RPM, again while the ECU retards timming by 5 to 10° (20° absolute timming), and above 6200 RPM, the engine stops to knock while timming become much more agressive (30° absolute timming).
The question is why the engine knocks between 5600 and 6200 with low timmings (20°) while it doesn't above 6200 with stronger timmings (30°) ?!!!
Can the STi's active Valve timming system explain such a behaviour ? In such a case, it would means that Valves are inadequately controlled by the STi7 ECU. Does someone knows at which RPM active valve system activates ?
What about AG821-2593 and STi8 ECU and behaviour ?
Thanks
11 March 2004, 06:35 PM
#2
Scooby Senior
Join Date: Oct 2000
Location: Zurich, Switzerland
Posts: 3,105
Likes: 0
Received 0 Likes
on
0 Posts
Hi Xavier - VVT timing isn't changed between 2592 and 2593. According to EcuTek the code is exactly the same, although some of the knock correction parameters are different.
Applying the 2593 code to 2591 or 2592 ECU hardware does not fix the issue. The difference is in the hardware, but buggered if I can find it - side by side they look the same. It also only affects LHD cars, so I doubt you'll get much interest here - do you speak German? Try www.subaru-community.net if so.
Did you go as far as measuring EGT on a car pulling excessive timing with high boost? Could be the pulled timing is actually causing the knock
I put a manual boost controller on all these suspects now before mapping, if they exhibit the issue I stop there and then until ECU is replaced (still under warranty in CH, but getting harder by all accounts).
Richard
PS: why don't you become an EcuTek dealer and do it "properly"?
Applying the 2593 code to 2591 or 2592 ECU hardware does not fix the issue. The difference is in the hardware, but buggered if I can find it - side by side they look the same. It also only affects LHD cars, so I doubt you'll get much interest here - do you speak German? Try www.subaru-community.net if so.
Did you go as far as measuring EGT on a car pulling excessive timing with high boost? Could be the pulled timing is actually causing the knock
I put a manual boost controller on all these suspects now before mapping, if they exhibit the issue I stop there and then until ECU is replaced (still under warranty in CH, but getting harder by all accounts).Richard
PS: why don't you become an EcuTek dealer and do it "properly"?
11 March 2004, 08:03 PM
#4
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Hi richard,
I already have another interesting job. I'm just mad keen on cars
AFAIK, all STi7 have the problem, but it becomes really important above 1.3 bar @ 6000RPM. A few do it even with stock boost !
But you're right, only LHD cars are affected and my German is too poor for German website visit !
Didn't check EGT, the detcan is much easier to install and to use. How can you explain that with the same fuel mixture, same boost pressure and less ignition timming (20°), the engine dets between 5600 and 6200 while it didn't above 6200 and up to the redline with pulled timming (30°) ? That's seems to be incoherent to me !
If only I could have one AG820-2592 and one AG821-2593 (or STi8) ECUs to open, no doubt I would find the differences in the knock sensor signal path (I'm working in the analogue and digital signal processing buisness). But nobody want to loose warranty on its car !
The fact that LHD cars only are affected makes me beleive that there is a noise problem which might be sorted-out by very simple filtering.
I already have another interesting job. I'm just mad keen on cars
AFAIK, all STi7 have the problem, but it becomes really important above 1.3 bar @ 6000RPM. A few do it even with stock boost !
But you're right, only LHD cars are affected and my German is too poor for German website visit !
Didn't check EGT, the detcan is much easier to install and to use. How can you explain that with the same fuel mixture, same boost pressure and less ignition timming (20°), the engine dets between 5600 and 6200 while it didn't above 6200 and up to the redline with pulled timming (30°) ? That's seems to be incoherent to me !
If only I could have one AG820-2592 and one AG821-2593 (or STi8) ECUs to open, no doubt I would find the differences in the knock sensor signal path (I'm working in the analogue and digital signal processing buisness). But nobody want to loose warranty on its car !
The fact that LHD cars only are affected makes me beleive that there is a noise problem which might be sorted-out by very simple filtering.
11 March 2004, 08:29 PM
#5
Scooby Senior
Join Date: Oct 2000
Location: Zurich, Switzerland
Posts: 3,105
Likes: 0
Received 0 Likes
on
0 Posts
Subaru's are pretty sensitive through this range of peak torque - this, along with the retarded timing which will up EGT's, make any problems even worse. That's only a theory of course, I've done no analysis to back it up.
How long would you need access to original 2591/2 & 2593 boards for? Fancy a trip to Switzerland?
Richard
How long would you need access to original 2591/2 & 2593 boards for? Fancy a trip to Switzerland?
Richard
11 March 2004, 09:05 PM
#7
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Crichon, Ricou, the French guys are interested ?
Richard, thanks for the invitation. Should be for a couple of day, but I'll need electronic tools such as scope and other that I have at my work only.
Richard, thanks for the invitation. Should be for a couple of day, but I'll need electronic tools such as scope and other that I have at my work only.
Trending Topics
12 March 2004, 12:01 PM
#8
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Xavier,
There are some cars that respond a bit overly sensitive, but with the standard turbo on the STi you WILL see lower advance in the 5500 - 6100 RPM range. I have tested with a car to disable the knock correction into the ECU and the car DID actually knock when I tried to advance a bit in this RPM range.
It could be that you have an overly sensitive ECU, but it also could be that everything is absolutely normal and you are on the limit of knock.
1.3 bar with standard turbo is WAY too much, you can go as high as 1.2 around 6000 and afterwards you loose consistency and you risk massive ignition retard as you see.
In most cases the timing pull is realy because of the knock limit is reached, so you have to check what is the reason for the knock limit to be reached. Normally you should be able to run about 25° of advance in the 5500 - 6000 RPM region at 1.2 bar of boost. (with 98 RON fuel and outside temperatures below 20°C)
P.S. I´ve tested about everything with the STi, I am able to map these cars with very consistent results right now! (perhaps not the highest numbers, but consistent)
P.S.2 These cars are VERY sensitive to fueling as well, specially in the "knock friendly" RPM´s. To lean gives problems, but too rich also........
Mark.
There are some cars that respond a bit overly sensitive, but with the standard turbo on the STi you WILL see lower advance in the 5500 - 6100 RPM range. I have tested with a car to disable the knock correction into the ECU and the car DID actually knock when I tried to advance a bit in this RPM range.
It could be that you have an overly sensitive ECU, but it also could be that everything is absolutely normal and you are on the limit of knock.
1.3 bar with standard turbo is WAY too much, you can go as high as 1.2 around 6000 and afterwards you loose consistency and you risk massive ignition retard as you see.
In most cases the timing pull is realy because of the knock limit is reached, so you have to check what is the reason for the knock limit to be reached. Normally you should be able to run about 25° of advance in the 5500 - 6000 RPM region at 1.2 bar of boost. (with 98 RON fuel and outside temperatures below 20°C)
P.S. I´ve tested about everything with the STi, I am able to map these cars with very consistent results right now! (perhaps not the highest numbers, but consistent)
P.S.2 These cars are VERY sensitive to fueling as well, specially in the "knock friendly" RPM´s. To lean gives problems, but too rich also........
Mark.
Last edited by EMS; 12 March 2004 at 12:03 PM.
12 March 2004, 12:45 PM
#9
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Thanks EMS,
My car is fitted with Andy's wonderful TD05/06-20G, that's why I try to raise boost above 1.3 bar.
I agree with you on all points. My only concerns is that it dets (hearing with det cans) in the 5600-6200 rev area with ignition timming of 20° while it doesn't above 6200 and up to the redline with pulled timming of 30° ! Same AFR (10.8:1) and boost level (1.45 bar). That seems to me very strange !!
I'm not reluctant to remap, but as I developped my own AFR and boost electronic control devices, I prefer to use it ;-)
My car is fitted with Andy's wonderful TD05/06-20G, that's why I try to raise boost above 1.3 bar.
I agree with you on all points. My only concerns is that it dets (hearing with det cans) in the 5600-6200 rev area with ignition timming of 20° while it doesn't above 6200 and up to the redline with pulled timming of 30° ! Same AFR (10.8:1) and boost level (1.45 bar). That seems to me very strange !!
I'm not reluctant to remap, but as I developped my own AFR and boost electronic control devices, I prefer to use it ;-)
12 March 2004, 01:09 PM
#10
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Xavier,
1.45 at 6000 with the turbo you have is A LOT!!! I have seen no increase in HP with anyting over 1.3 bar at 6000 with this turbo on a MY99 with FMIC! I am not supprised with the 20° of timing you can "only" run at your boost. (assuming you still use the TMIC)
Within a few weeks I will remap an STi with this turbo also with the TMIC, I can tell you more after that!
The 10.8 AFR should be quite good.
The increase in timing after about 6200 RPM is because the knock correction is switched off there, look out to not have too steep increase in timing as it could cause knock also! (Timing maps have to be changed though)
Mark.
1.45 at 6000 with the turbo you have is A LOT!!! I have seen no increase in HP with anyting over 1.3 bar at 6000 with this turbo on a MY99 with FMIC! I am not supprised with the 20° of timing you can "only" run at your boost. (assuming you still use the TMIC)
Within a few weeks I will remap an STi with this turbo also with the TMIC, I can tell you more after that!
The 10.8 AFR should be quite good.
The increase in timing after about 6200 RPM is because the knock correction is switched off there, look out to not have too steep increase in timing as it could cause knock also! (Timing maps have to be changed though)
Mark.
12 March 2004, 02:18 PM
#11
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by EMS
Xavier,
1.45 at 6000 with the turbo you have is A LOT!!! I have seen no increase in HP with anyting over 1.3 bar at 6000 with this turbo on a MY99 with FMIC! I am not supprised with the 20° of timing you can "only" run at your boost. (assuming you still use the TMIC)
Within a few weeks I will remap an STi with this turbo also with the TMIC, I can tell you more after that!
The 10.8 AFR should be quite good.
The increase in timing after about 6200 RPM is because the knock correction is switched off there, look out to not have too steep increase in timing as it could cause knock also! (Timing maps have to be changed though)
Mark.
1.45 at 6000 with the turbo you have is A LOT!!! I have seen no increase in HP with anyting over 1.3 bar at 6000 with this turbo on a MY99 with FMIC! I am not supprised with the 20° of timing you can "only" run at your boost. (assuming you still use the TMIC)
Within a few weeks I will remap an STi with this turbo also with the TMIC, I can tell you more after that!
The 10.8 AFR should be quite good.
The increase in timing after about 6200 RPM is because the knock correction is switched off there, look out to not have too steep increase in timing as it could cause knock also! (Timing maps have to be changed though)
Mark.
With 1.45, it dets between 5600 and 6200, but it didn't above !
12 March 2004, 07:15 PM
#12
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Xavier,
I don´t see why you want to run more boost when it doesn´t give you any gain in power? You see yourself that more boost gives you det, which you can only cure by retarding the ignition. If you retard the ignition in combination with this higher boost, I think you will have less power than with lower boost.
You have measured 390 HP at 6500 RPM, does your power graph has a drop in power between 5600 and 6200 RPM?
Was there a fan directly on top of the intercooler to avoid very high charge temperatures? In my opinion it´s useless to measure an Impreza with TMIC on the dyno without a fan directly on top of the intercooler. A blower in front of the car causes the hot air from the engine compartment to leave through the intercooler and bonnet scoop...........
Don´t bother too much about the drop in advance between 5600 and 6200 RPM, it´s completely normal and possibly caused by the high VE of this engine in this region.
P.S. Try to run some octane improver!
P.S.2 I think this power achievement (390 HP) is VERY good, as the dyno´s on the main land tend to read a lot lower than those in the U.K.
P.S.3 Did you try with a little bit richer mixture (10.6:1) in this critical RPM range?
P.S.4 Do you run std. injectors and fuel pressure?????
Mark.
I don´t see why you want to run more boost when it doesn´t give you any gain in power? You see yourself that more boost gives you det, which you can only cure by retarding the ignition. If you retard the ignition in combination with this higher boost, I think you will have less power than with lower boost.
You have measured 390 HP at 6500 RPM, does your power graph has a drop in power between 5600 and 6200 RPM?
Was there a fan directly on top of the intercooler to avoid very high charge temperatures? In my opinion it´s useless to measure an Impreza with TMIC on the dyno without a fan directly on top of the intercooler. A blower in front of the car causes the hot air from the engine compartment to leave through the intercooler and bonnet scoop...........
Don´t bother too much about the drop in advance between 5600 and 6200 RPM, it´s completely normal and possibly caused by the high VE of this engine in this region.
P.S. Try to run some octane improver!
P.S.2 I think this power achievement (390 HP) is VERY good, as the dyno´s on the main land tend to read a lot lower than those in the U.K.
P.S.3 Did you try with a little bit richer mixture (10.6:1) in this critical RPM range?
P.S.4 Do you run std. injectors and fuel pressure?????
Mark.
Last edited by EMS; 12 March 2004 at 07:17 PM.
12 March 2004, 07:41 PM
#13
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Hi Mark,
I've tried octane booster, but it didn't seem to improve behaviour in that nasty area.
I know the dyno figures are very good, but I can't understand why it should det in the nasty rev area while it doesn't above to the red line. And the dyno clearly shows a problem in this area : power curve stops to increase (375 HP at 5600RPM as well as at 6100 RPM) and increases again after 6200RPM ! That make me mad ;-)
I've also tried 10.2:1 fuel mixture without any improvement (95% IDC in that case !)
A agree I'm still using stock fuel pump and injectors....but I'm convince they should be OK up to 400 HP. Still have to check that with wideband sensor, but the fact it doesn't det above 6200 makes me beleive there is no pb with them.
I've tried octane booster, but it didn't seem to improve behaviour in that nasty area.
I know the dyno figures are very good, but I can't understand why it should det in the nasty rev area while it doesn't above to the red line. And the dyno clearly shows a problem in this area : power curve stops to increase (375 HP at 5600RPM as well as at 6100 RPM) and increases again after 6200RPM ! That make me mad ;-)
I've also tried 10.2:1 fuel mixture without any improvement (95% IDC in that case !)
A agree I'm still using stock fuel pump and injectors....but I'm convince they should be OK up to 400 HP. Still have to check that with wideband sensor, but the fact it doesn't det above 6200 makes me beleive there is no pb with them.
12 March 2004, 08:19 PM
#14
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Xavier,
Octane Booster didn't make it any better? The NF octane booster?
I already gave up to look after the det problem of the STi in that RPM region....... I tend to live with it at the moment. Did you try colder spark plugs? Did you try removing cranck case ventilation from the inlet?
It's realy remarkable that the "problem" disapears at higher revs.....
Perhaps we should start the investigation again!
P.S. You can be happy not to have a massive power "hole" in that RPM range as it is the case with a power kit for the STi made by a well known tuning company...
Mark.
Octane Booster didn't make it any better? The NF octane booster?
I already gave up to look after the det problem of the STi in that RPM region....... I tend to live with it at the moment. Did you try colder spark plugs? Did you try removing cranck case ventilation from the inlet?
It's realy remarkable that the "problem" disapears at higher revs.....
Perhaps we should start the investigation again!
P.S. You can be happy not to have a massive power "hole" in that RPM range as it is the case with a power kit for the STi made by a well known tuning company...
Mark.
12 March 2004, 08:35 PM
#15
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
If you are at 95% DC that means 19 ms injector opening time @ 6000 RPM, you can NEVER run at 10.2:1 AFR! At these outside temperatures on a car with the standard turbo at about 1.2 bar @ 6000, you need about 19 ms for 10.5 AFR. (with uprated pump, which gave an noticable increase in fuel pressure at high revs!) Your turbo flows A LOT more air so I don't think you are richer than 11.5:1 at the boost you are running.
Did you verify this 10.2:1 AFR with a wideband?
There is a big possibility that your fuel flow won't increase from a certain injector opening time, simply because your fuel pump is at the limit! I have tested one time with a fuel pressure gauge in an STi, I saw a fuel pressure drop at high revs running at about 300 HP. Fitting a Walbro gave me a stable fuel pressure, even at 360 BHP. So I NEVER use the standard pump in any setup above 300 HP to ensure maximum safety!
If you realy are that lean, richening the mixture could probably help you!
P.S. I am going to use 650 cc injectors with the TD05/06 hybrid in the STi as I max. out (about 95% DC) the standard ones now already with the standard turbo and Walbro fuel pump.......... (@340 - 350 BHP)
Mark.
Did you verify this 10.2:1 AFR with a wideband?
There is a big possibility that your fuel flow won't increase from a certain injector opening time, simply because your fuel pump is at the limit! I have tested one time with a fuel pressure gauge in an STi, I saw a fuel pressure drop at high revs running at about 300 HP. Fitting a Walbro gave me a stable fuel pressure, even at 360 BHP. So I NEVER use the standard pump in any setup above 300 HP to ensure maximum safety!
If you realy are that lean, richening the mixture could probably help you!
P.S. I am going to use 650 cc injectors with the TD05/06 hybrid in the STi as I max. out (about 95% DC) the standard ones now already with the standard turbo and Walbro fuel pump.......... (@340 - 350 BHP)
Mark.
12 March 2004, 09:10 PM
#16
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
Agree Mark, but again, it doesn't det after 6200 (no anoid sound in det cans while there is between 5600/6200).
Still no wideband fitted neither EGT sensor so I can't check fuel mixture and temperature. But Det Cans if of very good help.
Octane booster I tested were Elf and Liqui Moly (French ones I suppose ;-)). None of them helped.
I will swapp my fuel pump for a Walbro. It seems the 255 high Power was not to be considered for STi7 as it drawns too many current from the controller. Can you confirm ? Will the 255 LHP work then ?
As far as I know, 550 cc injectors are good for up to 410 HP considering a fuel mixture of about 10.8:1. So I won't change my injectors.
I suppose you use a wideband. What fuel mixture are you running at 350 HP with stock injectors, VF35 and 95% IDC ?
Edit : forgot, my spark plugs NGK PFR7B
Thanks for your help ;-)
Still no wideband fitted neither EGT sensor so I can't check fuel mixture and temperature. But Det Cans if of very good help.
Octane booster I tested were Elf and Liqui Moly (French ones I suppose ;-)). None of them helped.
I will swapp my fuel pump for a Walbro. It seems the 255 high Power was not to be considered for STi7 as it drawns too many current from the controller. Can you confirm ? Will the 255 LHP work then ?
As far as I know, 550 cc injectors are good for up to 410 HP considering a fuel mixture of about 10.8:1. So I won't change my injectors.
I suppose you use a wideband. What fuel mixture are you running at 350 HP with stock injectors, VF35 and 95% IDC ?
Edit : forgot, my spark plugs NGK PFR7B
Thanks for your help ;-)
Last edited by XNWRX; 12 March 2004 at 09:11 PM.
12 March 2004, 09:40 PM
#17
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Xavier,
It is true that the car easily knocks in the 5600 - 6200 RPM region. That is something that you have to live with! Some tricks can make it easier to live with, like little bit retarding of the main ignition map so you avoid over correction of the ECU in pulling timing. An other thing is to make the mixture a tad richer in that area.
I am (quite) sure you run way to lean to get a consistent result.
The GSS342 works perfectly fine in a WRX!
You will never be able to run 10.8:1 at 410 HP with the standard 500 cc (!) injectors. (assuming you run standard fuel pressure)
I run somewhere between 10.5 and 11.0 with my setup @ about 95% IDC. (don't want to tell the exact number....)
Mark.
It is true that the car easily knocks in the 5600 - 6200 RPM region. That is something that you have to live with! Some tricks can make it easier to live with, like little bit retarding of the main ignition map so you avoid over correction of the ECU in pulling timing. An other thing is to make the mixture a tad richer in that area.
I am (quite) sure you run way to lean to get a consistent result.
The GSS342 works perfectly fine in a WRX!
You will never be able to run 10.8:1 at 410 HP with the standard 500 cc (!) injectors. (assuming you run standard fuel pressure)
I run somewhere between 10.5 and 11.0 with my setup @ about 95% IDC. (don't want to tell the exact number....)
Mark.
13 March 2004, 04:09 AM
#18
Scooby Regular
Thread Starter
Join Date: Oct 2001
Posts: 141
Likes: 0
Received 0 Likes
on
0 Posts
OK Mark,
I'll try different setup and let you know.
The GSS342 is OK for WRX, but not sure it works on STi because of the fuel controller (33%, 66%, 100%) which can not deliver enough current for it !
I'll try different setup and let you know.
The GSS342 is OK for WRX, but not sure it works on STi because of the fuel controller (33%, 66%, 100%) which can not deliver enough current for it !
13 March 2004, 09:00 AM
#19
Scooby Regular
Join Date: Jan 2002
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Xavier,
"normal" WRX is the same as STi. Absolutely no problems there. (I already fitted more than 20 GSS342 in STi´s)
The Walbro as a direct swap is completely up to his task, maxing out the standard injectors without any drop in fuel pressure.
Mark.
"normal" WRX is the same as STi. Absolutely no problems there. (I already fitted more than 20 GSS342 in STi´s)
The Walbro as a direct swap is completely up to his task, maxing out the standard injectors without any drop in fuel pressure.
Mark.
Last edited by EMS; 13 March 2004 at 09:03 AM.
Thread
Thread Starter
Forum
Replies
Last Post
Brzoza
Engine Management and ECU Remapping
1
02 October 2015 05:26 PM
