EJ207 Subaru WRX 2011 Engine RPM problem

[andpdas]

Hello

Thank you for your attention.

I need help with my Subaru WRX STI 2011 JDM EJ207.

I have implemented my Link G4+ fury with my car.

But my engine does not increase rpm more than 5500rpm.

Under 5000 rpm works fine.

I've tried changing my ignition timing from 45 BTDC to 0 BTDC but it does not have any effects.

I've also double checked my dwell table, Ignition delay, Ignition charge time as same as shown in help in PCLink that it's same as the EJ20 quad AVCS configuration.

Do you have any guesses? Thank you very much.

 

 

[Adamw]

Your screenshot shows more than 8000RPM?

If you can you attach the actual log file rather than just the picture it would be much easier to help.

[andpdas]

Hi Adamw

Thank you for your immidiate responce.

I have attached multiple logs and my pclr file when I collected my logs. (although I changed the ignition timing for test)

My engine is 8000 rpm limit at factory.
And my link ecu is set to 8000 rpm max too.

You can see that, AP(Accel pedal position) is at a constant value (or could be increased slowly) but the rpm increces and decreses crazy.

Some times it shows like at 30000 rpm. but the noise from the engine doesn't sounds like 30000 rpm. (not 3000rpm. 30000rpm lol)

Best regards.
 

Log 2018-02-20 2;03;05 am.llg

Log 2018-02-20 2;01;51 am.llg

Log 2018-02-20 1;59;54 am回転が上がらない.llg

Log 2018-02-20 2;29;35 am.llg

Log 2018-02-20 2;25;35 am.llg

G4+Fury-initial19.pclr

[andpdas]

I forgot to say that, if I swap to stock ECU.
Than the engine works fine up to 8000 rpm.

So I don't think this is a mechanical problem.

[Adamw]

Ok thanks, now I understand.  You have massive trigger errors in all logs.

The next step is to do a triggerscope so we can see what is wrong with the signal (just at idle will be fine).  Here is a short video how to do the trigger scope:  https://1drv.ms/v/s!AiYbYlZQuRHPmieMTkwQDCXEb2LY

Please "save as log" and attach the file here.

[andpdas]

1 hour ago, Adamw said:

Ok thanks, now I understand.  You have massive trigger errors in all logs.

The next step is to do a triggerscope so we can see what is wrong with the signal (just at idle will be fine).  Here is a short video how to do the trigger scope:  https://1drv.ms/v/s!AiYbYlZQuRHPmieMTkwQDCXEb2LY

Please "save as log" and attach the file here.

 

Thank you so much Adamw.

I have attached my trigger scope log when the engine is idling at around 900 rpm.

Best Regards.
 

Trigger Scope Log 2018-02-20 2;07;25 am.llg

[Adamw]

Your camshaft "teeth" are upside down compared to normal subaru.  Can you try changing trigger 2 edge to "rising".

[Brad Burnett]

I did spot this as well, but the falling edge is much cleaner than the rising based on the trigger scope trace. 

If the timing is syncd correctly with the falling edge set, would it make much of a difference @Adamw

[Adamw]

2 hours ago, Brad Burnett said:

If the timing is syncd correctly with the falling edge set, would it make much of a difference @Adamw

We are expecting the falling edge on cam teeth to be just to the left of the "gaps" on the crank.  Then as the cams advance that cam edge moves further to the left of the gap.

Since his signal is inverted his falling edge actually starts off on the right hand side of the gap, and as the cam advances it crosses over to the other side of that gap.  I would say this is when it gets all upset.  Changing to the rising edge will put the trigger points back to the correct side of the gap where we designed them to work. 

 

2 hours ago, Brad Burnett said:

I did spot this as well, but the falling edge is much cleaner than the rising based on the trigger scope trace. 

A bit of rounding at the top or bottom a square wave is usually not a drama, the actual trigger point that the ecu uses is at a level of about 1.0 or 1.5V depending or if its rising or falling (cant remember which is which).  So provided the edge looks relatively vertical around the 1.0V mark you are usually good.

 

2 hours ago, Adamw said:

Your camshaft "teeth" are upside down compared to normal subaru.  Can you try changing trigger 2 edge to "rising".

Andpdas,  I should have also mentioned if you change the edge to rising and the engine will no longer start you may have to try adding or subtracting 360deg from your current trigger offset.

[andpdas]

HI, Adamw

Thank you very much for your advice.
I'll try that "Rising Edge" settings tonight.

Best Regards.

[andpdas]

Hello

I had a gear box trouble so I couldn't test until today.

@Adamw
I have changed the trigger 2 Edge as Rising.
But I still keep receiving trigger errors.
I feel it's getting even worse.
Also occures at around 1500rpm too.

I added the trigger scope at 1500rpm and the log I just collected today after changing triger 2 edge to rising.

Do you have any guess?

Thank you.


 

Log 2018-03-5 12;45;59 am-trigger-error.llg

Trigger Scope Log 2018-03-5 12;43;44 am.llg

 

Trigger Scope Log 2018-03-4 11;52;33 pm.llg

[Adamw]

Hmmm, you have something weird going on here.  In one of your triggerscopes the camshaft signal is inverted then in the other it is up the right way.  We know of some engines where the sensors are designed to do this but as far as we know this is not normal behaviour for the late EJ207 sensors.  So Im suspicious the LH Inlet cam sensor may be faulty.

Can you try swapping LH inlet cam sensor with the RH cam sensor?

One of your scopes correct way up:

 

The other is upsidedown:

[andpdas]

Hi

I noticed that both trigger scopes has same timing of when trigger scope 2 is falling.
Also these two trigger scopes are collected in different rpm.
My english is bad so I attached the picture what I'm trying to say.

I also attached an wire info.
Just in case if it will be an clue.
I forgot to say that my Fury is piggyback of my stock ECU.

I'll try exchange the Intake camshaft RH to trigger 2 tonight.

@Adamw
Could disconnecting Intake Camshaft position sensor from stock ECU could help?
Or the sensor's 5V(I guess) input is coming from the stock ECU.
Changing this from the Fury's 5V sensor voltage could help? 

Sorry for my newbie.
I am realy happy that you help me.
Thank you so much!

 

 

[andpdas]

I was reading the help browser in PCLink.

I found that there is a trigger errors in SubarV7-10.
Title is "Trig2: RPM Lockout".

```

This functionality, specific to the Subaru V7-10 Trigger Mode, is to eliminate trigger errors in some applications.

 

This value specifies the engine speed (RPM) at which the ECU ignores the trigger 2 signal for synchronization information. Because the ECU is already syncronized this will have no effect on engine operation.

 

If you are not experiencing trigger errors, set this value outside the engine operating range. e.g. 12,000 RPM.

 

If you are experiencing trigger errors, a typical value for this is 3,000 RPM.
```

I think my car is Subar Version 10.

Does this can be possible for this problem's reason?

[Simon]

That was a work around and really only masks another issue.  As if the triggers are correct and behaving as expected trigger errors should not occur. And if they do there is an underlying issue.

In this case the polarity of the sensors seems to change with RPM this will cause the timing of the set edge to move relative to the crank when the sensor flips polarity.

To our knowledge this is not normal for Subaru sensors.

I would suggest trying the sensor from the opposite bank (swap LH sensor in to RH sensor position on the engine) 

[andpdas]

@Adamw @Simon

Thank you for your advice.

I have exchanged the trigger.
now trigger 2 is at Intake AVCS RH and DI1 ad Intake AVCS LH.

But still not working well.

I couldn't start my engine.
I took a multiple trigger logs when I am cranking.
RPM was increasing and decreasing between 700 rpm to 900 rpm.

Could you help me solve?

one more question.
what trigger mode should i select?
I was trying with EJ 20 quad avcs.
is that ok ?

Thank you.

 

Trigger Scope Log 2018-03-7 1;06;41 am.llg

Trigger Scope Log 2018-03-7 1;06;07 am.llg

Trigger Scope Log 2018-03-7 1;03;44 am.llg

Trigger Scope Log 2018-03-7 1;03;28 am.llg

Trigger Scope Log 2018-03-7 12;58;53 am-IN-R-cranking.llg

[Adamw]

8 hours ago, andpdas said:

now trigger 2 is at Intake AVCS RH and DI1 ad Intake AVCS LH.

Sorry, I think you have misunderstood what I wanted you to do.  We still want trigger 2 connected to the LH intake cam but I wanted you to remove the sensor from the right hand cam and swap it with the LH sensor (this will prove if it is a faulty sensor).  I dont know how difficult this is in this car - can you actually access the sensors without too much work?

Also, we just noticed you have the pull-up resistor turned on on trigger 2.  Since you are piggybacking with the OEM ecu then that will already have a pullup connected, you should turn the pullup off on trigger 2.

[andpdas]

@Adamw

Thank you for your reply.

Sorry for my misunderstanding.

I asked the garages how much does it costs to swap camshaft sensor from left to right.
They said it costs around 20000yen (about 180 USD).

If it's possible, I want to check the other possibility of this problem first before I swap my sensors.
Maybe like, collecting the trigger-scopes at around 5000 rpm. Or anything else.

Also My stock ECU has a diagnostic system that can notice the camshaft sensor errors.
I rolled back my ECU to stock to check if there is an error.
But my ECU doesn't notice any types of errors.
I think the waveform from the camshaft sensor doesn't change even if I change it to new-new sensor.

I also took a log from Romraidar with my stock ECU installed.
But the engine RPM, camshaft position and any other logs does looks fine.

Any other opportunities?

How about changing the trigger mode to "Subaru V7-10(JDM)"?

↑Sorry if I was saying something misdirected.

I forgot to say but I've already changed the trigger 2 pull up to disabled.
The trigger-scopes that I mentioned here are all collected with trigger 2 pull up disabled.

Also, I have my camshaft position sensor's ground connected to stock ECU.
Is that ok?

Thank you.

 

[neil brown]

as an outsider to the thread looking in

 

why piggy back onto the original ecu when the link fury is more than capable running the engine and many other functions on its own

[Stevieturbo]

5 hours ago, neil brown said:

as an outsider to the thread looking in

 

why piggy back onto the original ecu when the link fury is more than capable running the engine and many other functions on its own

There may be legal/inspection matters that require the factory ecu to be in place ?

[Adamw]

6 hours ago, neil brown said:

as an outsider to the thread looking in

 

why piggy back onto the original ecu when the link fury is more than capable running the engine and many other functions on its own

 

1 hour ago, Stevieturbo said:

There may be legal/inspection matters that require the factory ecu to be in place ?

 The reason is that it is a 2011 STI, so it has significant CAN bus integration, certainly at least much of the dashboard is CAN driven and I think probably the "SI drive"/DCCD system will need CAN messages from ECU too.  We dont do a plugin for years beyond 2007 so we havent reverse engineered the CAN and therefore our firmware doesnt have a "2011 STI" CAN bus mode.

 

 

12 hours ago, andpdas said:

@Adamw

Thank you for your reply.

Sorry for my misunderstanding.

I asked the garages how much does it costs to swap camshaft sensor from left to right.
They said it costs around 20000yen (about 180 USD).

If it's possible, I want to check the other possibility of this problem first before I swap my sensors.
Maybe like, collecting the trigger-scopes at around 5000 rpm. Or anything else.

Also My stock ECU has a diagnostic system that can notice the camshaft sensor errors.
I rolled back my ECU to stock to check if there is an error.
But my ECU doesn't notice any types of errors.
I think the waveform from the camshaft sensor doesn't change even if I change it to new-new sensor.

I also took a log from Romraidar with my stock ECU installed.
But the engine RPM, camshaft position and any other logs does looks fine.

Any other opportunities?

How about changing the trigger mode to "Subaru V7-10(JDM)"?

↑Sorry if I was saying something misdirected.

I forgot to say but I've already changed the trigger 2 pull up to disabled.
The trigger-scopes that I mentioned here are all collected with trigger 2 pull up disabled.

Also, I have my camshaft position sensor's ground connected to stock ECU.
Is that ok?

Thank you.

 

Changing to V7-10 mode will not help.  The problem is the cam signal inverting.

My thoughts are the inverting cam signal is likely to be due to one of the following:

1. The piggy-backed OEM ECU is influenecing the signal.  My next suggestion would be to try running the engine with pin A21 at the OEM ECU temporarily disconnected.  It might even be a good idea to power the sensor independent of the OEM ECU.
2. The sensor is faulty - that is why we suggested swapping L & R sensors, it is unlikely to have 2 faulty sensors.
3. Maybe this signal inverting behaviour is normal/by design and is just something we havent seen before.  However we know of quite a few dealers/users running this late engine without drama so I'm not yet convinced it is normal.

So my suggestion is first to try disconnecting OEM ecu (pin A21).  If that doesnt help then the next step is to swap the sensors.

[andpdas]

 

11 hours ago, neil brown said:

as an outsider to the thread looking in

 

why piggy back onto the original ecu when the link fury is more than capable running the engine and many other functions on its own

 

6 hours ago, Stevieturbo said:

There may be legal/inspection matters that require the factory ecu to be in place ?

Thank you for looking the thread.

The reason that I choosed the piggyback instead of indipendent Fury.
Is because One is the CAN and DCCD.
Also, my engine needs a lot of output and inputs.
Like tumble generator, secondary air conditioning valve, charcoal canister, air conditioner system, e-throttle, power steering, inter cooler water spray system, input and out put camshaft VVT(subaru call this AVCS) etc...
So I need Thunder to support all of these input and outputs.
But I only wanted to control stuffs that makes more TORQUE and HORSEPOWER!!!

Thats why. thank you!
 

5 hours ago, Adamw said:

 

 The reason is that it is a 2011 STI, so it has significant CAN bus integration, certainly at least much of the dashboard is CAN driven and I think probably the "SI drive"/DCCD system will need CAN messages from ECU too.  We dont do a plugin for years beyond 2007 so we havent reverse engineered the CAN and therefore our firmware doesnt have a "2011 STI" CAN bus mode.

 

 

Changing to V7-10 mode will not help.  The problem is the cam signal inverting.

My thoughts are the inverting cam signal is likely to be due to one of the following:

1. The piggy-backed OEM ECU is influenecing the signal.  My next suggestion would be to try running the engine with pin A21 at the OEM ECU temporarily disconnected.  It might even be a good idea to power the sensor independent of the OEM ECU.
2. The sensor is faulty - that is why we suggested swapping L & R sensors, it is unlikely to have 2 faulty sensors.
3. Maybe this signal inverting behaviour is normal/by design and is just something we havent seen before.  However we know of quite a few dealers/users running this late engine without drama so I'm not yet convinced it is normal.

So my suggestion is first to try disconnecting OEM ecu (pin A21).  If that doesnt help then the next step is to swap the sensors.

Thank you so much.

I will try the no1.
although I have already disconnected the pin A21 from the stock ECU.
But I haven't try the power and the ground from the sensor independent of the stock ECU.

Even that didn't fix it, I will change LH input camshaft sensor to new-new one.
Even the wage of swaping these sensors are expensive, The price of the sensor is not so expensive.
I'll change it to new-new one just in case.

Best Regards.

 

[andpdas]

On 2018/3/8 at 10:10 AM, Adamw said:

 

 The reason is that it is a 2011 STI, so it has significant CAN bus integration, certainly at least much of the dashboard is CAN driven and I think probably the "SI drive"/DCCD system will need CAN messages from ECU too.  We dont do a plugin for years beyond 2007 so we havent reverse engineered the CAN and therefore our firmware doesnt have a "2011 STI" CAN bus mode.

 

 

Changing to V7-10 mode will not help.  The problem is the cam signal inverting.

My thoughts are the inverting cam signal is likely to be due to one of the following:

1. The piggy-backed OEM ECU is influenecing the signal.  My next suggestion would be to try running the engine with pin A21 at the OEM ECU temporarily disconnected.  It might even be a good idea to power the sensor independent of the OEM ECU.
2. The sensor is faulty - that is why we suggested swapping L & R sensors, it is unlikely to have 2 faulty sensors.
3. Maybe this signal inverting behaviour is normal/by design and is just something we havent seen before.  However we know of quite a few dealers/users running this late engine without drama so I'm not yet convinced it is normal.

So my suggestion is first to try disconnecting OEM ecu (pin A21).  If that doesnt help then the next step is to swap the sensors.

@Adamw

Hi 

I have swapped my Left Intake Camshaft Sensor to a brand new one.
But I'm still getting trigger errors when RPM goes over 5000 rpm.

I have tried both, piggy back left intake camshaft sensor and also independent ( disconnected pin A21 from stock ecu and enabled pull-up for trigger 2).
I have also tried changing the trigger edge to rising too. but it didn't work neither.
So I think, the possibility you raised, which no 1 and no 2 could not be the reason.

What should I try next?

I have attached.
- the my settings file
- log file that I drove and tried to reach up to 5000+ rpm
- trigger scope around 900 rpm
- trigger scope around 2000 rpm. (only this looks weird )
- trigger scope around 3500 rpm. 
- trigger scope around 5000 rpm. 

* all of these trigger scopes are collected in neutral gear. 
Best Regards

PS.

Althow the idle is unstable, I just haven't setup the closed loop idle control correctly.

So please don't care about it.

 

this.pclr

Log 2018-03-21 3;26;47 pm.llg

Trigger Scope Log 2000RPM 2018-03-21 3;06;57 pm.llg

Trigger Scope Log 900RPM 2018-03-21 3;06;40 pm.llg

Trigger Scope Log 5000RPM2018-03-21 9;31;10 pm.llg

Trigger Scope Log 3500RPM2018-03-21 9;31;25 pm.llg

[andpdas]

I noticed that I am receiving the camshaft position error.

• Error: 2nd Cam pulse
• Error: Extra pulse

But I have no idea whats going on.

And I don't think the trigger scopes are inverted.
Although the timing of camshaft signal rising is different.
But the camshaft signal falling timing are both(900 rpm and 2000 rpm) same.
( If you meant this as inverted signal, it's my misunderstanding. sorry )

[Adamw]

Hi Andpdas, sorry for my slow reply, I wanted to have a good look over all your data before replying.

I must admit I cant see much of a reason for your problem, that cam sensor is still inverting but I agree that the falling edge is remaining in the right place so should work regardless.  

Im running out of ideas quickly but I still have a couple of thoughts that I would like to try.  

1. Can you try changing to "EJ25 AVCS" trigger mode, see if any difference.
2. Can you temporarily disconnect trigger 1 from the OEM ecu to test?