knock sensor bypass

[RX8SE3P]

Edit: updated on my latest post further down.

Was brainstorming some ideas with a mate who also has a ZE.

I was wondering, with my AWR engine mounts and vibrations increased, would things like that affect the knock sensor and cause it to pull timing? He also mentioned that older ZE engines have a metal plate covering the area and that could lead to vibrations and throw the sensor off. Also if the sealant around the knock sensor degrades, it may pick up vibrations and think it's engine knock...

So what we were thinking was to somehow fool the ecu into thinking it's got the right readings for optimal timing. I know a bunch of people will argue that the knock sensor is important. The idea would mainly be for testing purposes on a car that is known to already be running quite well. If performance increases, then perhaps the sensor is picking up the wrong vibrations etc and then you could look into fixing that so the sensor reads properly. It could also save you from having to take half the car apart just to make it run good again, as like a temporary fix.

I could be wrong about it's function though. Does it retard timing permanently and store a CEL code permanently if knock is detected, or does it go away and run fine once knock is no longer detected?

[Nd4SpdSe]

I've never heard of anyone having issues using hardened or solid motor mounts, how about MazdaSpeeds own mounts? By that definition, your KS would go crazy over bumps and pot holes, and gravel roads too. Knock sensor senses a different type of vibration, that from misfiring within the motor, not vibrations around the motor. I've never seen anyone have problems or having to bypass the KS on an Mx-3, there shouldn't be a need to.

It's important but it's not. If you religiously run the required octane for your motor (ZE), than you should comfortable be able to remove it.

On my Xterra, however, that's a known problem. It's fairly common for people to bypass the KS, even myself I've done it after a new KS, with relocation, than unbolted, wrapped in foam and tied to a wire harness, was still detecting knocks, so I got pissed off and bypassed it. A few of my buddies have followed suit after me. I've ran without a knock sensor for about 3 years on my Xterra (did it before I jointed the military, and that's been 2yr 5mo), and I'm Supercharged. I however religiously put in 91 octane, as I'm suppose to. I can also feel when she retards the timing and limits power, and just replaced a faulty seconday 02 this past weekend that was causing it, but I think if she was automatic, I wouldn't be able to feel the motor like I can with stick.

What we do, which would work in the same principle on the Mx-3, is to find out the acceptable range of the KS in resistance; there's a high range and a low-range, add both together and devide by 2, that'll split the difference. Use that resistance and get a resistor of that (or close) value. Use that resistor to bypass/trick the ECU that all is good and well.

Again, this isn't a problem on MX-3's, and I would suggest against it, as that's probably not the cause, directly or indirectly, of any issues you're having.

[davmac]

Knock sensor is a microphone and output is oscillating dc voltage. I've heard of people connecting a 500K to 1M ohm resistor in place of the knock sensor. The resistor fools the ECM to thinking that a knock sensor is there even if it is not. That avoids having the ECM retard timing thinking the circuit is open. BTW the only code that a knock sensor will throw is open or bad circuit. It doesn't throw a code for knocking it just uses the signal to dynamically control timing.

[RX8SE3P]

Update - Seems I was throwing the knock sensor code and my knock sensor has died.

I just put the manifold back on and cannot be bothered taking everything apart for a piece of crap part that can just fail again.

My mate and I hooked up a 560K resistor in it's place and the code went away. Tested with it disconnected and code came back, then reconnected after resetting ecu and it went again. So putting a resistor there will work it seems. Will test further, car did sound stronger with the sensor bypassed.

Also, this confirms that the code will be stored unless you clear the ECU.

[davmac]

^ makes sense. The ECU normally only uses the knock sensor to retard timing if it thinks the engine is knocking. The ECU probably treats an open (effectively unplugged) knock sensor the same as a knocking engine. The knock sensor provides protection, but does not enhance engine performance.

[Nd4SpdSe]

If the ECU sees no knock sensor, it's probably freaking out, and going into safe mode.

[Caleb T]

I will be tryin this bypass method soon! Thanks for the post up.
I'm going to run this car as a DD till it blows up, so knock sensor is not needed, lol.

[RX8SE3P]

I have ran mine bypassed with the resistor for about 700kms now and the code has not come back. It's a supercheap and fast solution. My car has not ran this good ever before.

Go for it and let us know how it went afterwards. Our engines don't really knock ever, so I think the sensor is more of a wank factor item.

[Silkwyrm]

So since my KLZE swap I've gotten a KS CEL code. Installed a new sensor and still got the code, didnt care much since I could tell the engine wasnt knockiing. Also I've been getting a strange sluggishness in the in the mid-upper RPM range. Never put it together that these were related. Last night I tried this and wired in a resistor. The weird sluggishness is gone! So happy I saw this thread. Thanks guys.