So as many of you know and have helped me through some really strange issues with my car this spring. The one that has been plaguing me has been instances of KR. The car just seemed to be pulling timing whenever it felt like it. It did have a pattern though, lugging in a high gear and and low throttle input was most noticeable. It did pull some timing at WOT sometimes, but not all the time. Last weekend I was doing some maintenance on my SHO, oil change, checking the CAC and catch can. When I was done I took the car out for a test drive and I hit a spike the likes I have never seen 7.5KR on minimal throttle input on the first turn out of my drive way. Total freaked out I pulled the car back into the garage and re checked everything. I couldn't see anything wrong. So again a spike of 5KR. Brought the car back in and started looking for something loose that could be rattling. I tied everything loose that could be possibly rattling. I even went as far as to put some rubber washers on the belly pan because it was loose as hell. No impact. Still the same issue. I then remembered about the chaffing knock sensor wires from last year (Mike bdp had a good how to regarding the chafing). I did however fix that by securing them away from the block. Now this is where it gets interesting. I pulled the Airaid intake off and the intake piping away from the area and the zip tie was still doing its job. So I decided to add a few more just for good measure. Took the car out and the issue was even worse?? Ok so now at least I know where the issue is. After much trial and error I finally came up with a method to secure all the wires and hoses to avoid any movement in the area. I am convinced that in my case at least the connector that the knock wires plug into has some sort of issue. You really shouldn't need to go to the length that I did to make this issue go away. The wires themselves showed to sign of chaffing, I even wrapped each wire with e-tape and that had no effect. 

So on to the how to.
Just in front of the thermostat housing you will find the plug in question. you can follow the wires down under the intake from there. My harness where the knock sensors plug into was particularly loose. I attached a few tip ties to the knock sensor wires going down under the intake to what seems to be a vac line. This is a good one since it moves with the motor and is will not change depending on load of the car. You want to stay away from securing it to say the rad hose as it flexes when the engine is under different loads. I Then secured all the rubber vac lines away from the harness and wires. The next step was to secure the harness itself. I used a piece of 1/2 inch fuel line about 4 inches in length split the hose down the center and slid this over the wires directly under the connection. I then tip tied the hose to the main wire bundle directly under the connector. I also tip tied the top of the connector just for good measure. The biggest issue was isolating the bundle from everything else that flexes during the normal operation of the car. But with some perseverance it is possible. If and when I pull the intake itself I will replace the knock sensors and the wires. But I know that some of you have had similar issues and this could potentially solve your issue to. Since yesterday morning I have had 0 spikes and KR is almost non existent. I will get like 0.5 once in a blue moon. Even on the expressway lugging in a high gear at 9 psi of boost its adding timing same as last year.   



   

Sent from my iPhone using Tapatalk

Great idea. I'll be doing this later tonight. Thx for the write up. I'll post my results.

Very interesting! I will be attempting also once my car cools down. I have the same exact symptoms you did...totally random.

One question. Is your split fuel line acting as a wire loom touching the knock sensor itself? Can't tell from the pic...

Quote from: ecoboostsho on May 28, 2015, 04:42:55 PM
One question. Is your split fuel line acting as a wire loom touching the knock sensor itself? Can't tell from the pic...

Correct I am using it as both as a loom and insulator. This also stabilizes the harness too. It does not go down very far. Only enough to cover the first few inches of the wires. This also prevents the zip tie from pulling on the wires coming out off the harness.


Sent from my iPhone using Tapatalk

So, just to be certain, you basically anchored the knock sensor wire harness and connector, so that it would no longer rub or bump up against other objects. Is that correct?

Seems very feasible. False knock is easily detected by the sensor, essentially from anything that may be tapping or rubbing or clanking in the engine compartment. The sensor basically "listens" for knock. It isn't foolproof.

Anything under the hood, especially around the sensor, should be stowed securely. Check out some of the false KR findings Mazdaspeed owners have dealt with over the years on MSF.

Got it. Just didn't want to mute the vibrations to the sensor itself so to speak. Doesn't sound like you are doing that at all. I really hope this works for me!

Quote from: IHeartGroceries on May 28, 2015, 04:48:07 PM
So, just to be certain, you basically anchored the knock sensor wire harness and connector, so that it would no longer rub or bump up against other objects. Is that correct?

Seems very feasible. False knock is easily detected by the sensor, essentially from anything that may be tapping or rubbing or clanking in the engine compartment. The sensor basically "listens" for knock. It isn't foolproof.

Anything under the hood, especially around the sensor, should be stowed securely. Check out some of the false KR findings Mazdaspeed owners have dealt with over the years on MSF.

Correct and exactly.


Sent from my iPhone using Tapatalk

Good finding, dude. Power delivery will feel much, much more linear now. XD

Yes it does, and thanks!


Sent from my iPhone using Tapatalk

Great info, just did it myself even though I was not experiencing any of the symptoms. Took all of 5 minutes to accomplish. Thanks for the heads up.

Well, this certainly dispels the theory that only the 2010-12's are vulnerable to knock sensor related issues!

Awesome job!

Glad to see you've figured it out.

Rich

That is some sweet news to say the least,makes me wonder if your DTC code for loss of communication can be tracked down to the knock sensor,and usually one code will be set,but could have hundeds stored in the PCM,for the same code if its that eratic. Z