Can somebody please explain how they work and what the frequency(s) are that it looks for and picks up?Also how they are wired to the ECM and what the ECM does when it picks up knock.Im doing a project for Junior Design and this is all much needed info.The more info the better.Thanks guys,Nick

 

I know I'm way out of my territory in this part of the forum, but from what I've seen, they pick up frequencies in the 6 kHz range and report that as knock. The sticky situation comes in when you install something like a poly tranny mount that causes this type of resonance, at which point you have to dig into the PCM calibration and desensitize them. I couldn't get anything over 19-20* of WOT timing until I desensitized them about 10-20%.

I know I've seen a pretty good description around here on them. Might dig around in the tuning section too.

 

i searched here and in tuning and got nothing...i saw the thread about the trouble u were having,glad u fixed that.So sorry if its been posted but i havent seen it,please post up any and all info!

 

TTT i need some info or a link or something fellas!

 

Did some googleing and came up with this:
http://www.megasquirt.info/ms2/knock.htm
Not much else applicable out there... I'd like to find out more about this too.

 

that was pretty informative,i just wish some of the guru's on here would post with some more info.ie how they connect to the PCM and are wired in and freq etc.

 

The Ls1/Ls6 are located in the valley, the new Ls2's are located on the side of the block.

ls1/ls6 sensors are in the 6K range. The new Ls2 sensors are a broad range sensor.

A lot of this has been covered when the Ls2 came out and folks became aware of the fact ls2 sensors won't work.

You might read some of those threads as the talked about freq. response.

 

The two factory knock sensors, located beneath the intake manifold, are microphones that are used by the PCM to "listen" for certain frequency sounds that are determined to be knock. It does so by monitoring the voltage output of the sensors and associating certain voltage changes with knock. Both "false" and "real" knock cause sudden changes in the voltage output of the knock sensors, and the PCM will then retard our timing to compensate and prevent potential damage. This will always result in decreased power output, sometimes 10-30 rwhp easily depending on your individual engine setups. "Real" knock is caused by detonation, which could be caused by running excessively lean (poor tuning), a bad batch of gasoline, or other problems related to higher than normal compression ratio etc. Now, lets understand the difference between "Real" and "False" knock. False knock can be caused by several things, including:

1. Over-sensitive Knock Sensors
2. Exhaust components contacting the chassis and transmitting vibrations
3. Loud exhaust systems transmitting vibrations back to knock sensors
4. Noisy valvetrain components (lifters, rockers, valves, valvesprings, pushrods, etc..)
5. Noisy internal motor components or drivetrain components

Tony

 

Excuse me if I misinterpreted your question at all. Our knock sensor location and frequencies were posted above but I read your question like you wanted to know how it actually detected knock? Knock sensors are Piezoelectric sensors. The Piezoelectric effect is an example of how energy is converted from one form to another, "mechanical to electrical." A knock sensor is made up of a polarized crystal encased in a sturdy conductive principle. When a polarized crystal is compressed it forms a voltage through mechanical deformation. The conductive principle directs the newly aquired AC current to the computer. First it has to go through a DC converter so the computer can read it. The harder the knock, the more the crystal is compressed, which determines the amount of voltage sent to the computer. The higher the voltage the more drastic measures the computer takes to correct the situation.

 

Also, Piezoelectric sensors are 1 wire sensors. No source voltage is required to operate, only feed voltage is reported.

 

Dang he's good......and the funny thing is he is realy from Alabama....go Alabama public schools....hell yea

 

Could a chattering clutch cause false knock? I seem to be getting some knock retard events on hard take-offs when the clutch is slipping. (I still have a stocker, it is total crap!!)

 

i have a knock sensor bank 2 code that has been coming and going for about 2 weeks. would bad spark plugs be a good place to start looking for the cause?

 

I've felt a chattering clutch and seen it on the log take away timing, picked up as knock.

 

Yes very possible

 

very interesting

 

this thread is old but a lot of good info here. very interesting

 

Any sensors that generates its own voltage signal without a source voltage is known as active sensors.Those that rely on a source voltage (example:Map sensor) is known as a passive sensor.As mentioned already a knock sensor is a piezoelectric sensor that generates its own voltage signal to the ECM when it senses the vibration caused by detonation.The ECM in return retards the ignition timing to prevent engine damage from the occuring detonation.

 

Two things, we hooked up the factory knock sensors to the radio in a vehicle that was running megasquirt and we could hear the knock through the speakers. Same vehicle when it was naturally aspirated had a siezed u-joint in the drive shaft that caused it to pull timing out over 3500rpm on the dyno so they are very sensitive. We have since gone back to a factory PCM and HPtuners so we can get up on the edge of boost by watching the knock closer.

 

Didn't want tp start a new thread so i'm bringing this one back with another question.

Should the PCM show knock if the Knock sensors are disconnected.

When the TPS hits 60% I'm getting 8* of knock even if I unplug the knock sensors.