5th Gen Subaru Impreza Forum banner
1 - 20 of 52 Posts

·
Registered
Joined
·
20 Posts
Discussion Starter · #1 ·
I have a 2019 Impreza with manual trans.

I've been chasing down a slight stutter during both cruise and acceleration for quite some time. Previously thought it was misfire so I took it to the dealership and they couldn't find anything but gave me a tsb that they said pertained to my issue(tsb had nothing to do with it). Issue was still there though. Irritated and determined, I bought a launch scan tool(I had worked at dealerships/shops for a while so I am fairly mechanically inclined).

I had been checking different things until I came across knock correction. For the last 3 months I've been monitoring knock correction with other things to find a correlation/cause. The stutter I'm feeling is the knock correction pulling timing. I don't feel it until it hits -3 to -7 degrees retard, which happens atleast a few times a drive, but I can't seem to find any other readings that correspond with that other than it happens around 2300 RPM when I'm cruising at 55MPH.

In the time I was working on cars, I didn't deal a whole lot with timing so this is a somewhat new thing for me. Looking for any direction I can be pointed to look or anyone else that's had a similar issue.

It is under warranty and I am going to be talking to them about it but I drive 1 hour to work and another hour back home 5 days a week. I don't always feel it and am already guessing they're not going to find anything on their 10 min drive around the block so I want to try and narrow this down as much as possible for them.
 

·
Registered
Joined
·
554 Posts
Others in the forum have noted that running higher-octane fuel can improve things. Especially in hot ambient temperatures which will make lesser fuels knock the engine.

However, at this point, you are only assuming that the knock-sensor is causing the "stuttering". You have not mentioned anything conclusive.

As an experiment, you can try a tankfull of higher-octane fuel. This should reduce or eliminate knocking so the sensor is no longer impacting things. You can still monitor to make sure the sensor is not activating.

The goal of this experiment would be to determine if the "stuttering" goes away during this time.
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #3 ·
Others in the forum have noted that running higher-octane fuel can improve things. Especially in hot ambient temperatures which will make lesser fuels knock the engine.

However, at this point, you are only assuming that the knock-sensor is causing the "stuttering". You have not mentioned anything conclusive.

As an experiment, you can try a tankfull of higher-octane fuel. This should reduce or eliminate knocking so the sensor is no longer impacting things. You can still monitor to make sure the sensor is not activating.

The goal of this experiment would be to determine if the "stuttering" goes away during this time.
]
I've ran 91 or 93 for the last 3 months exclusively, the knock sensor isnt detecting anything, only knock correction. Nothing has been an assumption as I have 3 months of recorded live data showing that my knock correction is pulling the timing.

I was just on another thread that went on for about 3 years of people with 2017-2019 having this issue and one of the last posts of the thread talked about doing an idle relearn procedure with positive results. Going to try that this weekend.
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #4 ·
]
I've ran 91 or 93 for the last 3 months exclusively, the knock sensor isnt detecting anything, only knock correction. Nothing has been an assumption as I have 3 months of recorded live data showing that my knock correction is pulling the timing.

I was just on another thread that went on for about 3 years of people with 2017-2019 having this issue and one of the last posts of the thread talked about doing an idle relearn procedure with positive results. Going to try that this weekend.
This is a screenshot from one of my recordings.
25907
 

·
Registered
2018 Impreza base. 5MT
Joined
·
47 Posts
25914



I've gone done the same road with my 18 5mt car. Here is an image of one of many data recordings. The upper right box is actual ignition timing. Lower right is knock signal, upper left is engine rpm and lower left is throttle plate opening angle. You can see the ECM pulling/retarding ignition timing without a knock signal, sometimes as far as -20 degrees. I spent many hours reviewing data, and the best I can tell is the mapping for emissions is what does it. The problem did seem to get slightly better after the first ignition coil recall. I haven't gotten the latest one done yet. I've mostly given up on the issue, except for keeping an eye on this forum for updates, and I upshift around 4k rpm now instead of 3k. I know this isn't the best example of what the timing is doing, but it was the first recording I looked at. I drove myself (and my wife) crazy trying to figure this out.
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #8 ·
View attachment 25914


I've gone done the same road with my 18 5mt car. Here is an image of one of many data recordings. The upper right box is actual ignition timing. Lower right is knock signal, upper left is engine rpm and lower left is throttle plate opening angle. You can see the ECM pulling/retarding ignition timing without a knock signal, sometimes as far as -20 degrees. I spent many hours reviewing data, and the best I can tell is the mapping for emissions is what does it. The problem did seem to get slightly better after the first ignition coil recall. I haven't gotten the latest one done yet. I've mostly given up on the issue, except for keeping an eye on this forum for updates, and I upshift around 4k rpm now instead of 3k. I know this isn't the best example of what the timing is doing, but it was the first recording I looked at. I drove myself (and my wife) crazy trying to figure this out.
That's about where I'm at. As I said in another post, I saw a few people had positive results with doing an idle relearn procedure. I'm going to try that today or tomorrow and is basically my last attempt until I just completely ignore it. I'll update on results this coming week.
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #9 ·
That's about where I'm at. As I said in another post, I saw a few people had positive results with doing an idle relearn procedure. I'm going to try that today or tomorrow and is basically my last attempt until I just completely ignore it. I'll update on results this coming week.
 

·
Registered
Joined
·
500 Posts
Automotive electronics in the last 25 years or so have become so invasive that actually chasing-down an issue like yours takes a lot of time and expensive equipment. Now that you've uncovered the 'problem', what can you do about it? Completely reprogram the ECU? Slap in an illegal ECU, and risk a failed inspection and federal fines? Find some 'work-around' which may or may not either work or even be legal? Fine-tuning engines today is just about impossible; it's better to just adjust your driving technique to work with your engine's shortcomings.

I really don't like how my CVT goes too quickly into a too-high gear, and how reluctant it is to shift down in anything short of nearly flooring the gas. As a result, I tend to corner with greater throttle than I'm used to, just to keep the transmission in a more drivable range. I also often use the shift-paddles as an over-ride even in normal traffic. Work with what you've got...
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #11 ·
Automotive electronics in the last 25 years or so have become so invasive that actually chasing-down an issue like yours takes a lot of time and expensive equipment. Now that you've uncovered the 'problem', what can you do about it? Completely reprogram the ECU? Slap in an illegal ECU, and risk a failed inspection and federal fines? Find some 'work-around' which may or may not either work or even be legal? Fine-tuning engines today is just about impossible; it's better to just adjust your driving technique to work with your engine's shortcomings.

I really don't like how my CVT goes too quickly into a too-high gear, and how reluctant it is to shift down in anything short of nearly flooring the gas. As a result, I tend to corner with greater throttle than I'm used to, just to keep the transmission in a more drivable range. I also often use the shift-paddles as an over-ride even in normal traffic. Work with what you've got...
Either today or tomorrow I'm going to essentially erase the learned memory from the ecu. From what Ive read and understand, at some point it learned the conditions of what caused a previous knock and is working to prevent that by pulling timing when it sees those certain conditions again. I won't be doing any actual "work" on it as I refuse to tear into it until the warranty is up. Simply doing a hard reset on the computer components if you will. It's easy, takes little time and it's worth a shot. If this doesn't work then I will be done trying to mess with it. Already have it set up so it's paid off before the warranty is up so if I want/have to I can sell it when the time comes.

Just hoping they'll come out with a wrx hatchback before it gets to that point lol.
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #12 ·
Automotive electronics in the last 25 years or so have become so invasive that actually chasing-down an issue like yours takes a lot of time and expensive equipment. Now that you've uncovered the 'problem', what can you do about it? Completely reprogram the ECU? Slap in an illegal ECU, and risk a failed inspection and federal fines? Find some 'work-around' which may or may not either work or even be legal? Fine-tuning engines today is just about impossible; it's better to just adjust your driving technique to work with your engine's shortcomings.

I really don't like how my CVT goes too quickly into a too-high gear, and how reluctant it is to shift down in anything short of nearly flooring the gas. As a result, I tend to corner with greater throttle than I'm used to, just to keep the transmission in a more drivable range. I also often use the shift-paddles as an over-ride even in normal traffic. Work with what you've got...
Day 2 after the idle relearn procedure. It's still showing knock correction but not nearly to the frequency or degree it was before. The most it's shown is -2.3 degrees (I could physically feel it around -4 degrees and up). Its usually staying around -.6 to -1 degree and that was onlyhand full of times in the 120 miles I've driven since yesterday morning. Still curious as to why it's trying to correct a knock that isn't there. Going to see if I can talk to the tech instead of the service writer this weekend during my oil change to pick their brain a bit.
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #13 ·
Update: First 2 days were good, after that it's back to how it was before. Hitting -7 degrees and just as often. Talked to the service advisor and hes herd of the issue. Checked my oil level which was good and I told him Ive used premium fuel exclusively for the last 3 months. Going to schedule an appointment with their master tech and see what happens from there. If they can't fix it I'm just going to drive it until something actually gives. I am already on track to have this paid off before the warranty is up so I can either trade it in or sell it.
 

·
Registered
Joined
·
21 Posts
Update: First 2 days were good, after that it's back to how it was before. Hitting -7 degrees and just as often. Talked to the service advisor and hes herd of the issue. Checked my oil level which was good and I told him Ive used premium fuel exclusively for the last 3 months. Going to schedule an appointment with their master tech and see what happens from there. If they can't fix it I'm just going to drive it until something actually gives. I am already on track to have this paid off before the warranty is up so I can either trade it in or sell it.
Hey guys, Marty1 here from the Crosstrek forum. SOA has finally come out with a TSB (09-77-21) to address this issue. Below is the first page. In a few weeks the link to the entire TSB will show up under "Manufacturer Communications" for the Impreza and Trek on the NHTSA site.

Font Parallel Screenshot Rectangle Number
 

·
Registered
Joined
·
20 Posts
Discussion Starter · #15 ·
Hey guys, Marty1 here from the Crosstrek forum. SOA has finally come out with a TSB (09-77-21) to address this issue. Below is the first page. In a few weeks the link to the entire TSB will show up under "Manufacturer Communications" for the Impreza and Trek on the NHTSA site.

View attachment 26079
I'm glad to hear that they've found the issue, however, I'm in the process of trading it in lol. This definitely throws a wrench in my decision making.
 

·
Registered
Joined
·
21 Posts
I'm glad to hear that they've found the issue, however, I'm in the process of trading it in lol. This definitely throws a wrench in my decision making.
Sadly some Crosstreks have also been traded in over this issue.
The good news is that Trek owners are already reporting that the fix has solved the problem.
 

·
Registered
Joined
·
193 Posts
I wish they would fix the jerkiness in the CVT that my 18 Legacy doesn't have. It is much worse with the AC on but is noticeable without also. Things like moving at slower speeds or sometimes even when going faster around 50 and hitting the gas it will jerk but like I said much worse with the AC on.
 

·
Registered
Joined
·
21 Posts
I wish they would fix the jerkiness in the CVT that my 18 Legacy doesn't have. It is much worse with the AC on but is noticeable without also. Things like moving at slower speeds or sometimes even when going faster around 50 and hitting the gas it will jerk but like I said much worse with the AC on.
That could be due to the CVT chain slipping. Have you seen this TSB? MC-10188160-0001.pdf (nhtsa.gov) The dealer can perform tests to confirm.

Here's a list of possible symptoms, as listed in the TSB:
Font Circle Rectangle Number Magenta
 
1 - 20 of 52 Posts
Top