Can't get rid of knock Need more idea's

[DJ23GSX]

I am having some knock issues at 19psi and 93 octane that i can't tune out. I am pushing my idc's close to 95% on the 560's @ 42psi bfp and still can't get rid of it. I can't seem to get knock under 12-15 counts no matter how much fuel i add or take away. My mods are in my profile. I would love to turn the boost down but thanks to the great porting from sbr it creeps to 19psi by 5k.

So far i have fixed or tried these things

1. New Knock sensor retourqued a few times with no change in knock
2. New fuel filter
3. Fixed all boost leaks (including the AFPR)
4. Raised my base fp up to 45psi
5. Different 2g mas
6. New plugs
7. Rewired the 255
8. New caps in ecu
9. Air filter box with hose plumbed with ambient air
10. Took foglight out and trimmed the i/c duct to fit the mkiv i/c better
11. Checked the fuel pump seal and cleaned the pump bag filter


I get pk when the car is cold and driving normal but it goes away after the car is warmed up. I don't think its pk i am seeing under boost though as it does go up and down if i add or take fuel away. I just can't tune it for 0 knock. It doesn't seem like i should have to add this much fuel. The next thing i would like to try is bringing my base timing down a bit. I also have a set 650's but i don't think i should need that much fuel for this set up from what i've been reading in other threads with similar mods.

 

[DJ23GSX]

I thought my boost creep was my boost controller at first but even if i run a line directly from the turbo outlet nipple to the wastegate it still creeps to 19-21psi.


When i say a tiny boost leak i mean tiny. It takes a good couple of minutes to bleed from 25psi down to 10psi and then even longer from 10psi to 0. Not to mention the fact that fixing my horrible boost leaks at the injector insulators barely did anything to my idc's for some reason.

I am in the process of getting a timing light and dropping it back down to 5*. I think it can only help the situation.

 

[blazebolt]

How well is the wastegate passage ported in your 16G? That's where the boost creep problem lies with these turbos. I've ported the inlet and wastegate passage a good bit and was able to hold 7psi when directly connected my j-pipe nipple to the wastegate actuator. There's a vfaq with good pictures in the tech articles section. Doesn't sound like you have a boost leak if it takes a long time for it to reach 10psi. It shouldn't affect it that much. Post a log of your next run so we can see the progress

 

[oldman]

I thought my boost creep was my boost controller at first but even if i run a line directly from the turbo outlet nipple to the wastegate it still creeps to 19-21psi.

http://www.dsmtuners.com/forums/showthread.php?t=186525

I am in the process of getting a timing light and dropping it back down to 5*. I think it can only help the situation.

When you checked your timing via logger @ 7* BTDC, did you ground the terminal on the firewall?

 

[DJ23GSX]

http://www.dsmtuners.com/forums/showthread.php?t=186525


When you checked your timing via logger @ 7* BTDC, did you ground the terminal on the firewall?

I know what needs to be done to help the boost creep and will most likely be doing it this weekend.

I don't remember if i grounded the terminal on the firewall. I used a timing light not a logger to check the timing. I didn't know you could check timing with just a logger. I am going to swing by my mechanic's shop tomorrow and borrow his light and set it to 5*.

 

[oldman]

I don't remember if i grounded the terminal on the firewall. I used a timing light not a logger to check the timing. I didn't know you could check timing with just a logger. I am going to swing by my mechanic's shop tomorrow and borrow his light and set it to 5*.

Then I mis-understood your previous statements and thought that your 7* BTDC was simply based on your logger reading at idle without grounding the terminal.

I believe last time i checked my base timing was 7* but i don't have a timing light so i'm not going to mess with that.

As far as the timing goes i set the base timing in the datalogger to 7 so it should be reading the correct timing while under boost.

 

[Stapl3]

Yet another great SBR thread.

 

[DJ23GSX]

Then I mis-understood your previous statements and thought that your 7* BTDC was simply based on your logger reading at idle without grounding the terminal.

No sorry all i meant by that was because my base timing was 7* with the timing light i changed the base timing setting in the preferences section of the datalogger to 7* to give me the correct timing while datalogging. Does that make better sense?

My timing at idle without the terminal grounded bounces from 7-9* but i'm not sure thats worth anything to you.

What would cause my idc's on the 650's to be almost as high as the 560's both at the same bfp? Better yet i made two pulls today both with 6 counts of knock (back to back) and didn't touch the safc but the idc's jumped again from 88% on the first pull to 98% on the second pull?

Yet another great SBR thread.

Are you refering to there wonderfull portwork or his seat of the pants tuning methods??

 

[oldman]

No sorry all i meant by that was because my base timing was 7* with the timing light i changed the base timing setting in the preferences section of the datalogger to 7* to give me the correct timing while datalogging. Does that make better sense?

I understand that after your last post.

My timing at idle without the terminal grounded bounces from 7-9* but i'm not sure thats worth anything to you.

Sort of, timing should bounce between 8-10 when base is set to 5*, I'm not so sure if your base is set to 7*, my guess is more like 4*.

What would cause my idc's on the 650's to be almost as high as the 560's both at the same bfp? Better yet i made two pulls today both with 6 counts of knock (back to back) and didn't touch the safc but the idc's jumped again from 88% on the first pull to 98% on the second pull?

1. At 45psi with a 1G stock ECU, 560s flow 617cc and 650s flow 716cc, what are your before and after AFC settings between the change over?

2. Are you saying after installing the 650s and properly adjusted your AFC, the IDC is the same at the same PSI? Furthermore, are you saying your IDC changes between two pulls at the same PSI? If so, can you compare the airflow reading of 560s(617) vs. 650s(716) as well as pull to pull at the same psi and RPM? Your IDCs are directly related to your MAF readings, is your 2G mas hacked?

Are you refering to there wonderfull portwork or his seat of the pants tuning methods??

Most that I have helped over the years in porting for boost creep starts off with a SBR ported EVO3, the auther of the tech article included, it's best to save yourself the $50 for porting tools.

 

[DJ23GSX]

I understand that after your last post.


Sort of, timing should bounce between 8-10 when base is set to 5*, I'm not so sure if your base is set to 7*, my guess is more like 4*.


1. At 45psi with a 1G stock ECU, 560s flow 617cc and 650s flow 716cc, what are your before and after AFC settings between the change over?

2. Are you saying after installing the 650s and properly adjusted your AFC, the IDC is the same at the same PSI? Furthermore, are you saying your IDC changes between two pulls at the same PSI? If so, can you compare the airflow reading of 560s(617) vs. 650s(716) as well as pull to pull at the same psi and RPM? Your IDCs are directly related to your MAF readings, is your 2G mas hacked?


Most that I have helped over the years in porting for boost creep starts off with a SBR ported EVO3, the auther of the tech article included, it's best to save yourself the $50 for porting tools.

1. I was running 42psi bfp on both the 560's and 650's. I was able to lean out the hi throttle settings on the safc a ton when i put in the 650's. I'd say a good 15-20%. Although my fuel trims are still asking for fuel even when i have the low throttle settings at 0 on the safc for some reason.

2. My idc's on the 560's at 20psi were around 82-85% with 0-3 counts of knock before the heatsoak or knock starts and i have to add all that extra fuel to try and stop it which is why my idc's were sometimes over 100% on the 560's. On the 650's at 20psi my idc's dropped to around 77-80% until somthing happens at 6k rpms and they jump to 99% most of the time even though my 5k 6k and 7k safc settings are all the same at -14%. Like i said before something is happening at 6k. Up til 5-5.5k rpms my idc's are under 80% with 0 knock and i could probly even lean it out more.

Yes at the same psi of boost without touching my safc settings on back to back runs i've seen my idc's raise up 10-12% for no apparent reason at 6k rpms. My 2g mas is unhacked. It still has the honeycombs and the screw is turned all the way in. I've tried the screw half way out as well with no change in knock.

 

[Slowboy]

I understand that after your last post.


Sort of, timing should bounce between 8-10 when base is set to 5*, I'm not so sure if your base is set to 7*, my guess is more like 4*.


1. At 45psi with a 1G stock ECU, 560s flow 617cc and 650s flow 716cc, what are your before and after AFC settings between the change over?

2. Are you saying after installing the 650s and properly adjusted your AFC, the IDC is the same at the same PSI? Furthermore, are you saying your IDC changes between two pulls at the same PSI? If so, can you compare the airflow reading of 560s(617) vs. 650s(716) as well as pull to pull at the same psi and RPM? Your IDCs are directly related to your MAF readings, is your 2G mas hacked?


Most that I have helped over the years in porting for boost creep starts off with a SBR ported EVO3, the auther of the tech article included, it's best to save yourself the $50 for porting tools.

Port work does not cure boost creep.

Path of least resistance is what you need to think about, where does the air go - around the turbine wheel or out the WG hole?

There is not much material in the real MHI EVO III turbo to remove under the WG flapper, you can barely run a 1" end mill, or drill bit down the hole, and leave enough material to seat the 34 mm flapper.

Only way to make creep go away is a larger WG hole for air to escape, simple as that!

Mike Huml
SBR Inc.

 

[blazebolt]

I have the stock 32mm flapper and just opened up the pathway to the wastegate and was able to hold 7psi all the way to redline.

 

[oldman]

Port work does not cure boost creep.

Path of least resistance is what you need to think about, where does the air go - around the turbine wheel or out the WG hole?

There is not much material in the real MHI EVO III turbo to remove under the WG flapper, you can barely run a 1" end mill, or drill bit down the hole, and leave enough material to seat the 34 mm flapper.

Only way to make creep go away is a larger WG hole for air to escape, simple as that!

If that is the case, perhaps you shouldn't be selling an EVO3, especially charging a $60 port job and $60 34mm upgrade knowing that they do nothing, at the very least you should tell your customers that an external wastegate or having enough supporting mods to run 22psi+ will be required for the turbo.

In fact porting does work as myself and many others have proven over the years, the biggest contributors to boost creep is not the size of flapper or wastegate hole, rather it's the wastegate angle of entrance to the flapper in the inlet as well as the lack of actuator extension (opening the flapper more). It's more about wastegate path flow characteristics than capacity, if you know this maybe your port jobs would've been more effective.

1. I was running 42psi bfp on both the 560's and 650's. I was able to lean out the hi throttle settings on the safc a ton when i put in the 650's. I'd say a good 15-20%. Although my fuel trims are still asking for fuel even when i have the low throttle settings at 0 on the safc for some reason.

Because a 2G mas flows about 31% more and 617s flow 37% more, theoretically your low settings should be at about -6% and -21% between 4-7k on high settings.

2. My idc's on the 560's at 20psi were around 82-85% with 0-3 counts of knock before the heatsoak or knock starts and i have to add all that extra fuel to try and stop it which is why my idc's were sometimes over 100% on the 560's. On the 650's at 20psi my idc's dropped to around 77-80% until somthing happens at 6k rpms and they jump to 99% most of the time even though my 5k 6k and 7k safc settings are all the same at -14%. Like i said before something is happening at 6k. Up til 5-5.5k rpms my idc's are under 80% with 0 knock and i could probly even lean it out more.

Yes at the same psi of boost without touching my safc settings on back to back runs i've seen my idc's raise up 10-12% for no apparent reason at 6k rpms. My 2g mas is unhacked. It still has the honeycombs and the screw is turned all the way in. I've tried the screw half way out as well with no change in knock.

If the screw has been turned then the mas was hacked, once the screw has been messed with, it's difficult to get it back to stock and will not read correctly/consistently. I think there are several issues here.

1. Your ignition base timing is off.

2. It's pretty clear that heatsoaking is probably your major source of knock.

3. Your inconsistent/high IDC is probably caused by the hacked 2G mas.

4. Your tuning is most likely off due to adding fuel in the attempt to rid of knock, probably also the reason why you have not run into timing advanced issues associated with running 716s with an AFC as well as high IDCs, wideband will most likely show rich conditions.

This is what I would do, first set your base timing to 5*BTDC and BFP to 37psi, port your turbine housing and take control of boost, then properly retune (650s flow 31% more therefore low should be about 0 and high should be about -14%) to about 18psi until you're able to add the FMIC. If the inconsistant IDCs persists between pull to pull, borrow a virgin 2G mas and test. Good luck.

 

[DJ23GSX]

If that is the case, perhaps you shouldn't be selling an EVO3, especially charging a $60 port job and $60 34mm upgrade knowing that they do nothing, unless you tell your customers that an external wastegate or having enough supporting mods to run 22psi+ will be required for the turbo, might as well take your internal O2 dump housings off your product line.

In fact porting does work as myself and many others have proven over the years, the biggest contributors to boost creep is not the size of flapper or wastegate hole, rather it's the wastegate angle of entrance to the flapper in the inlet as well as the lack of actuator extension (opening the flapper more), if you know this, maybe your port jobs would've been more effective.


Because a 2G mas flows about 31% more and 617s flow 37% more, theoretically your low settings should be at about -6% and -21% between 4-7k on high settings.


If the screw has been turned then the mas was hacked, once the screw has been messed with, it's difficult to get it back to stock and will not read correctly/consistently. I think there are several issues here.

1. Your ignition base timing is off.

2. It's pretty clear that heatsoaking is probably your major source of knock.

3. Your inconsistent/high IDC is probably caused by the hacked 2G mas.

4. Your tuning is most likely off due to adding fuel in the attempt to rid of knock, probably also the reason why you have not run into timing advanced issues associated with running 716s with an AFC as well as high IDCs, wideband will most likely show rich conditions.

This is what I would do, first set your base timing to 5*BTDC and BFP to 37psi, port your turbine housing and take control of boost, then properly retune (650s flow 31% more therefore low should be about 0 and high should be about -14%) to about 18psi until you're able to add the FMIC. If the inconsistant IDCs persists between pull to pull, borrow a virgin 2G mas and test. Good luck.

This afternoon i happened to find the silencer for my apexi catback. I tossed it in in attempt to lower exhaust flow enough to control the boost. Well it did just that and i made a total of about 3 pulls at 15psi all the way to redline. The first few pulls i had the 650's at 42psi bfp and i was seeing 24* timing and little spurts of knock but nothing over 6 counts and idc's around 68%. After afew pulls and not being able to fully get rid of the knock i figured it was due to the high timing so i backed the bfp back down to 37psi and upped the boost to 16psi and made 2-3 pulls with 0 knock and idc's at 70%. The car definatly feels like its being held back but i am happy to see the knock gone.

One more thing what kind of issues will a bad fuel pump relay cause? I had the caps replaced in the ecu a few weeks ago and everythings been fine since but while i was tuning this afternoon the car shut off on me 3 or 4 times but i didn't hear the constant clicking like a bad ecu does. It was just one quick click and the car came right back on on its own. The check engine light came one everytime it did it but didn't stay on for more than 10 seconds so i could check it with the logger. It was kinda weird i think the stereo stayed playing but the dash lights and safc shut down and restarted.

 

[Slowboy]

If that is the case, perhaps you shouldn't be selling an EVO3, especially charging a $60 port job and $60 34mm upgrade knowing that they do nothing, at the very least you should tell your customers that an external wastegate or having enough supporting mods to run 22psi+ will be required for the turbo.

In fact porting does work as myself and many others have proven over the years, the biggest contributors to boost creep is not the size of flapper or wastegate hole, rather it's the wastegate angle of entrance to the flapper in the inlet as well as the lack of actuator extension (opening the flapper more). It's more about wastegate path flow characteristics than capacity, if you know this maybe your port jobs would've been more effective.

You are not correct.

You can only flow X amount of air through a given size hole - end of story - I do not care what angle the air enters the hole, this is not what is most important.

Also - we can open the stock hole of the EVO III - measure it stock, and tell me if 1" is a larger hole.

In case you forget, there is no one on this list with more dyno time under there belts on the EVO III than SBR - I am pretty sure over the years we have had our time with this turbo.

 

[blazebolt]

I think someone fell asleep in physics class. The angle of entry is VERY important or else air will not flow properly towards the wastegate. Sure you can make the exit hole bigger (wastegate) but it will do very little since you have not addressed the real problem which is the path that the air is traveling. I don't want to waste my time getting into detail. You can figure it out on your own.

 

[Slowboy]

I think someone fell asleep in physics class. The angle of entry is VERY important or else air will not flow properly towards the wastegate. I don't want to waste my time getting into detail. You can figure it out on your own.

Sure, then you add pressure into the equation and you are still wrong.

I am not saying the angle has NOTHING to do with it, but about 5% of the equation.

Fact is, if the hole is to small, not enough air can get out depending on pressure, etc.

Mike Huml
SBR Inc.

 

[oldman]

You can only flow X amount of air through a given size hole - end of story - I do not care what angle the air enters the hole, this is not what is most important.

Also - we can open the stock hole of the EVO III - measure it stock, and tell me if 1" is a larger hole.

Again, if this is true, why are you knowingly selling this turbo with internal wastegate? Worse yet, why do you charge customers $60 for a 34mm flapper upgrade? I'm not sure what is worse, your lack of knowledge on the subject of EVO3 boost creep or your lack of honesty for your customers.

You are not correct.

If you say so Mike, I guess all of us who can hold 12psi, through proper porting, to redline are just lying all these years. The internal wastegate path of the EVO3 can handle more than enough flow under 20psi, best example being the flapper pivot arm mod which changes the length of the flapper pivot arm to allow the flapper to open more than 45*, this method has also been proven to work without any porting, the reason it's not a popular fix is due to the reduction of the actuator leverage resulting in flapper being blown open earlier.

In case you forget, there is no one on this list with more dyno time under there belts on the EVO III than SBR - I am pretty sure over the years we have had our time with this turbo.

I failed to see the relevance of this with the subject at hand.

 

[blazebolt]

The reason why your assumption is flawed is because you are not looking at the entire picture. In stock form, the wastegate passage is practically perpendicular to the turbine passage. Air will not travel towards the wastegate until there is adequate backpressure given from the turbine. By making it easier for air to travel to the wastegate, less backpressure from the turbine is needed to make more airflow towards it. Although allowing a bigger exit passage allows more volume of air to escape, the air does not want to head in that direction. There are other factors that I am leaving out, but it gets more complicated and we'll just stop here for now.

 

[daren_p]

I would have to agree with everything Bruce has said. To the poster who said you can only flow so much though an X sized hole, sure this is true but only to a point. Obviously you don't know much about flow because for a given size sure that hole can only flow so much, but guess what when you add obstructions to the inlet & outlet of that hole, it no longer is able to flow the same amount as the same sized hole without any restrictions in the flow path. The evoIII WG inlet & outlet are flawed in design & it has been proven over & over again that with proper porting of these restrictions, the WG hole is then able to bypass enough gases to prevent creep. If you take the time to read some of the posted threads about proper porting to eliminate boost creep you would understand what people are referring to as they are not refering to the std porting that is done to a turbo. I know of a good number of people that had boost creep issues & after only performing proper porting to eliminate boost creep, the creep they had was no longer an issue. I too have personally done just proper porting & nothing else to creeping turbos & it too eliminated the creep but why whould you think this would work it only been proven time & time again

 

[GVR4592]

Port work does not cure boost creep.

Path of least resistance is what you need to think about, where does the air go - around the turbine wheel or out the WG hole?

There is not much material in the real MHI EVO III turbo to remove under the WG flapper, you can barely run a 1" end mill, or drill bit down the hole, and leave enough material to seat the 34 mm flapper.

Only way to make creep go away is a larger WG hole for air to escape, simple as that!

Mike Huml
SBR Inc.

You've obviously never PROPERLY ported a turbocharger then. You guys are really making yourselves look stupid in this thread.

 

[nightspeed87]

You've obviously never PROPERLY ported a turbocharger then. You guys are really making yourselves look stupid in this thread.

Sad but true....

 

[DJ23GSX]

Well back on topic now


I just got back from my mechanics shop and he say's there's something wrong timing wise cause according to the timing light its reading 50-60* advanced at idle with the terminal on the firewall grounded. He said if it were that far off i would be bending valves. He thinks the idler pulley might have slipped or is slipping. Does this sound right?

One thing i noticed as well when i was changing my power steering belt is that the two notches in the cam gears a lined up but are facing straight up and not facing each other in the middle like everything i've seen and read about those marks.

 

[daren_p]

Well back on topic now


One thing i noticed as well when i was changing my power steering belt is that the two notches in the cam gears a lined up but are facing straight up and not facing each other in the middle like everything i've seen and read about those marks.

Just wanted to clarify that you are first aligning the marks on the crank pulley & the mark on the timing cover before your looking at the marks on the cam gears? If you haven't done this first, the position of the cam gear marks don't mean anything. If you have aligned the crank pully & this is happening, I would be pulling the crank pulley off & making sure the real timing marks are ligned up as the crank pulley is known to crack & seperate which can make your timing marks off.

 

[DJ23GSX]

Just wanted to clarify that you are first aligning the marks on the crank pulley & the mark on the timing cover before your looking at the marks on the cam gears? If you haven't done this first, the position of the cam gear marks don't mean anything. If you have aligned the crank pully & this is happening, I would be pulling the crank pulley off & making sure the real timing marks are ligned up as the crank pulley is known to crack & seperate which can make your timing marks off.

No i didn't look or line up the crank pulley and timing cover mark when i noticed this. I just noticed it while the timing belt cover was off during the ps belt intall. You'll have to bare with me on this timing stuff i've never touched it before so i will try and explain things to the best of my ability. Thanks

 

[daren_p]

No i didn't look or line up the crank pulley and timing cover mark when i noticed this. I just noticed it while the timing belt cover was off during the ps belt intall. You'll have to bare with me on this timing stuff i've never touched it before so i will try and explain things to the best of my ability. Thanks

Hey no worries, everyones got to learn sometime . As I spoke about quickly above, the cam gear marks mean nothing as far as their position, unless you first line up the crank pulley (or sprocket underneath, there are actually 2 sets of marks on the crank). Basically when you shut the motor off it isn't going to stop in the correct position to check your timing marks. To line up the crank pulley you can insert a socket wrench into the end of the crank pulley bolt & manually turn the motor over till it reaches the correct spot for the marks to line up. You only want to turn it clockwise though, if you turn it in the opposite direction, you risk jumping timing. Instead of going over it all here, go to www.vfaq.com & check out the faq on changing a 1g timing belt. This should give you a better idea of what needs to be done to check timing marks.