WHAT IS A KNOCK SENSOR FOR AND IS IT HARD TO REPLACE WHEN MY CAR GOES INTO SECOND GEAR AT FULL THROTTLE IT MAKES A WEIRD RATTLING NOISE THEN IT CLEARS UP AND TH MOTOR RUNS GOOD I DONT KNOW WHAT IT IS

 

Knock Sensor is there for detected Engine Knock (Pinging). Basically when the Knock sensor detects knocking in the engine, the voltage from the ESC Module to the ECM drops, and this signals the ECM to retard timing.

No, there not hard to change at all. I'm pretty sure your Knock sensor is located by your starter, on the block.

 

isnt the hole that it screws into directly in the water jacket areas?

 

Your Knock Sensor is located on the passengers side of the block between the two middle cylinders, between 4 & 6 and yes water will flow out when you replace this sensor.
Do not over tighten I believe its only 15 or 20 lbs of torque recomended But dont hold me to that exactly and dont use any teflon tape or anything that might prevent this sensor from feeling the detonation.

If you run a strong motor with out this sensor working properly you can easily blow a head gasket or worse due to excessive detonation under a load or high RPM.

 

It sounds like you are encountering detonation already. A knock sensor will not fix that. You are either running too much compression, your cam timing is off, too much timing or too low of octane.

If you think BECAUSE you hear detonation that the Knock Sensor is bad (because it went deaf), you are partially correct. A knock sensor is specifically tuned to a particular engine (and a particular ECM for that matter too). A 305 KS is differnent from a 350 KS and a MAF type KS is different than a SD type. This is because the internal harmonics and frequency they operate are different. (Though there is a way to make a MAF 350 KS work on a SD 350 engine).

A knock sensor will hear noises the human ear does not hear and begin to pull timing out. The audible knock the human ear hears is NOT the same frequency the KS works at. This is why a number of people will say "My knock sensor is picking up false knocks". They think that the knock sensor is not working properly by pulling timing out when they cannot hear the detonation. But the fact is, the knock sensor IS detecting the start of detonation, just not at the same frequency as the human ear.

But if the noise persists, the knock sensor can become saturated with noise and stop pulling timing out. This is about the time when you start to hear the engine detonating.

Lastly, make sure the "rattle" is in fact detonation and not some other noise. The most common place detonation is first heard is on the "upshfit" - somewhere just before the shift to just after the shift.

 

This is not correct. The KS has a pipe thread and if you just screw it into the block, it will eventually leak water out of the treads. You need to apply some sort of paste or pipe tape to the treads of it. IIRC, the correct torque spec for the sensor is 16ft lbs. If it gets tightened more then that then the sensor gets more sensitive. vs versa for too loose. I am 99% sure on that.

Also, some guys have problems with the sensor picking up false sounds coming from headers. not sure is an exhaust leak would be sensed by the sensor but I guess anything is possible. I read somewhere on here that you can check to see if your sensor is working by having a buddy hit the side of the block or the header with a hammer. If you have a timing light hooked up, you should see the timming retard just after the hammer hits the block. Not sure how true or accurate this is but I have read it on this board somewhere AND I have never actually tried this techinique. just my .02 cents there.

if your gonna change it then I suggest prepping the new one with pipe tape and keepingithandy since water will come rushing out of the hole once the old one is removed. If your quick about it you might not make a big mess. otherwise, you will need drain the coolant from the car and I cant gaurentee that you will get it all out to eliminate water spewing from that hole. Good luck.

PS DONT BUY A KNOCK SENSOR FROM AUTOZONE. SPEND THE MONEY AND GET IT AT THE DEALER.

 

Thats funny the check engine lite was comming on in my truck after the dealer installed the new motor.Took it back and thay said the tech that put the motor in put teflon tape on the KS and it would not ground.Thay took the teflon tape off screwed it back in and in three years it has not leaked yet.

 

All the ones I have seen new come with some sort of red coating on the threads to keep it from leaking coolant. I haven't had any problems with leaks in the past, and I have yet to use teflon on a knock sensor.

 

possibly Loctite? Also I have heard aswell to use that white paste stuff on it . It hardens up after applying

 

Do you really believe that? That has got to be the dumbiest thing I have heard this week. No offense on you or anything but think about it for a second. They claim that the sensor was not grounding out on the block because of the tape being on it? Im sorry but even if you apply tape or paste to the treads, there is going to be at least some spot on the threads where it is making direct contact to the block. it is impossible to put tape on threads and NOT get the threads of the sensor to touch the block. The dealership will tell you anything just to get you out the door thinking they saved the day for you.

As for the red stuff on the treads , yes itis a form of locktite. If you remove your sensor after it has been installed then you will need to use locktite again. So, you have a choice or loctitie, thread paste, or thread tape.

 

Bottled - Dont use teflon tape thats not good advice it can defeat the purpose of having a properly functioning sensor.
The threads are designed with a tapered pitch that prevents leaking similar to a NPT fitting. I have replaced several on different motors and never heard of a leak on one.

Yes the sensor can be tested with tapping on the block while idling with any type of metal object and a timing light the sensor works (if it works)and pulls timing back and it can be seen at idle with a timing light

Detonation is not some type of frequency not noticeable by the ear its very noticeable to any one who knows motors or motor knock while your motor is under a load or WOT its not that complicated.

Knock if not controlled will rattle a motor apart period.

 

This is new news to me. I had a sensor leak on a buddys car a while back after a engine overhaul and we ended up using one layer of pipe tape and all was good. I dont understand what you mean by " defeating the purpose of a properly functioning sensor."

does the same advice apply to all sensors of just the KS.

 

The knock sensor is designed to pick up the slightest amount of rattle or ping before it gets out of control by pulling timing back.

Yes theres water there and it does seem like teflon tape is the average or normal thing to do in this case but its not and your not plumbing. I personally have made the same mistake many years ago myself.
This sensors job is designed to pick up a motors worse enemy (detonation) wich gets worse and more prevelant when you build a strong motor with improper timing curves or poor quaility fuel or to high of compression with out race fuel or a lack of fuel under load a lean condition.

If you soften or cushion the sensors ability to hear, feel or sense knock by placing thick teflon tape there than that can defeat the purpose of having a good properly functioning sensor..

 

You CAN have detonation and not hear it. Its not always audible. I myself have experienced this.

Also, it is possible to have detonation, have a functioning knock sensor, and the sensor not pick it up every time.

As far as being able to test the knock sensor with a timing light at idle, I don't think you can do it on all thirdgens. I haven't tested it myself to see if it will keep from pulling timing, but the $8D code has a parameter for minimum vehicle speed before it will retard timing. In AUJP for example, its set to 2 or 3 mph. That means that the ecm will not pull timing even if knock is found unless you are over that required speed. Hitting the header with a wrench and watching the timing light should not show a change in timing. (Again, atleast in this instance, not saying its like this for all thirdgens)

 

[QUOTE]Originally posted by BOTTLEDZr28
[B]Do you really believe that? That has got to be the dumbiest thing I have heard this week.

The check engine has not come on for three years now and that is all thay changed.besides that the service manager was my brother and i don't think he would lie to get me out of there.

 

92blue do you realize everything you say is a contradiction within in its self ! Choose a side and stand behind it period!

[QUOT E]Originally posted by 92blue
1.[B]You CAN have detonation and not hear it. Its not always audible. I myself have experienced this.
1. Well if you couldnt hear it then do pleases tell us how you knew you had a detonation problem "with out hearing it".
[QUOTE]Originally posted by 92blue
2.[Also, it is possible to have detonation, have a functioning knock sensor, and the sensor not pick it up every time.
2. If your sensor doesnt pick it up every time then its not functioning properly or you have a short some where.
[QUOT E]Originally posted by 92blue
3.As far as being able to test the knock sensor with a timing light at idle, I don't think you can do it on all thirdgens. I haven't tested it myself to see if it will keep from pulling timing, but the $8D code has a parameter for minimum vehicle speed before it will retard timing. In AUJP for example, its set to 2 or 3 mph.
3. I have personally tested it on more gm vehicles than I can count including V'6's and all v 8's TPI's,TBI"s MPIand yes with a wrench or pipe against any bracket that is connected to the blockwhile idling! If you dont know just say so.

 

I don't see a contradiction in anything I said. I stand behind everything I said in my last post.

Ever though of hooking up a laptop or a scanner? They come in pretty handy if you want to see if you have knock counts, and how many. I HAVE had detonation before, and not heard it. My knock counts went up, and the ecm has retarded timing. I heard nothing, and neither did anyone else in the car with me.


Have you ever wondered why its frowned upon to tune a spark table by "advancing the timing until it knocks, then back off 2 degrees"? Its because the knock sensor doesn't pick up detonation 100% of the time it occurs. You can have knock and not know it. Don't believe me, do a search. Its been covered many times over by alot of the "engine gurus" on this board.


Good for you. Would you like a cookie?

I said that I can't speak for all thirdgens, but the $8D code has a parameter for minimum vehicle speed before it will enable knock retard. That I do know. Since you seem to think I pulled that out of my ***, here you go, straight out of the ANHT hac...

*********************************************
* KNOCK PARAMS
*
*********************************************
L820A: FCB 52 ; 650 RPM, IF >= ENAB KNOCK
; ARG = RPM/12.5
;
L820B: FCB 2 ; MPH, IF >= THEN ENAB KNOCK
;
L820C: FCB 48 ; -3.9 DEG C, IF < DISABLE KNOCK
; ARG = (DEG C +40) * 256/192
;
L820D: FCB 53 ; 0.24 DEG C, IF DIFF COOL SINCE START >
; ENAB KNOCK, ARG = DEG C * (256/192)

In case you are wondering why I referred to AUJP and now I am quoting ANHT, I had ANHT handy, and they are both set to the same thing: 2 mph.

Where do I contradict myself?

 

My apologies 92blue for thinking that you were talking out of your bum..
Every time I run my scan I have found that my ecm pics up knock and how much timing is being pulled back, sometimes a few degrees at idle sometimes a whole lot at WOT.
I have seen several things send a knock sensor a false signal thinking that it was my motor and it actually wasnt such as exhaust hitting or rubbing underneath the car body, noisy valvetrain - roller rockers, and cracked flywheels as well.

 

Thats cool, don't worry about it.

I agree with you on the false knock comment. I think the LT1s have a module or something that they change when they install roller rockers to keep the noise from registering as false knock, don't they? I've heard something like that before, but don't recall the specifics.

 

I helped with the install of two prochargers on two seperate SS camaro's about 5 years ago, bothe having strong sound low mile motors and no sign of trouble(knock). But we noticeably picked up knock at idle and at low RPM when we also upgraded the valve train with roller rockers and stronger rigid push rods.
The funny thing to me at the time was a buddy suggested to put several layers of teflon tape on the sensor to keep the sensor from being so sensitive and it did just that.
I do remember we made some peremiter calibration changes but Im not sure right off what they were to correct the problem.

 

The knock sensor detects the knock sound produced by pre-ignition.. Pre-ignition happens when the fuel/air mixure ignites too early during the compression stroke. (The gas explodes before the spark plug sparks).. Knock can be caused by things like fuel with too low an octane, the ignition timing too far advanced or glowing hot spots on the heads caused by excessive carbon buildup.

FWIW, before I replaced my knock sensor because I heard knocking, I'd check a few other common problems first.. Off the top of my head I have three questions:

Is your timing set to specs ?

Are there any vacuum leaks ?

Is your PVC valve OK.. (If it gets stuck open, you'll run lean).

--
Wrapping a knock sensor with telfon won't stop knock.. It just makes the sensor less sensitive to knock. As a result it won't report the knock to the ECM which won't retard the timing as much and you'll get a bit more horsepower.

.