Some of these may sound dumb but bear with me, I am in training.

1. What is the Knock Sensor and what is it's job?

2. I have a '86 carbed LG4. Where is my KS? Post pic if possible

3. When I accelerate I often hear a knock or pinging. Is this because of something related to the KS?

4. When do you know when to replace the KS?

5. Is this something I can tackle or is lot's of diagnostic timing equipment and the such needed to repair.

:hail: :hail: :hail:

You guys are the BEST. Thanks in advance.

 

1. It's a little microphone, literally. It's tuned to the particular audio frequency produced by knocks in a given size engine. Its job is to listen for spark knock and transmit instances of it to the ECM, which can then back off the timing (brutally) to prevent it. This way, the ECM can in theory dial in as much timing as possible while knowing when it has to back it off on the fly, for the best gas possible mileage.

2. It screws into the coolant drain plug hole on the pass side of the block, just above the oil pan rail, near the starter.

3. Sort of... your static timing (distributor location) may be too far advanced, and the ECM may already have pulled as much as possible out, and it's still too much advance. That engine should not knock at its normal timing setting even when fed typical gas pump watered-down cat pee instead of fuel, so something's not right.

4. They don't ordinarily go bad, except when over-tightened or dropped or hit or something like that. They're kind of fragile that way. They're not terribly expensive, so it might not be a bad idea to change yours out. Use about 4 wraps of Teflon tape (not thread sealer or silicone, use the tape) on the threads, and only tighten it to about 8 or 10 foot-pounds; in other words, don't gorilla it into there. It's the coolant drain plug, i.e. all your coolant will come out when you remove it, so it would be a great idea to drain & flush your cooling system and put in new coolant at the same time.

5. Yes. No. Maybe.

 

Thanks

 

I've done a LOT of experimentation with the Knock Sensor AND the eprom. Inside the eprom you can control the retard "attack rate" (how fast it retards the spark), the "decay rate" (how quickly it restores the timing), and the "max limits" by either RPM (when in WOT/PE) or via the load (MAP reading for SD).

I can tell that the Knock Sensor is NOT a perfect device. First, as RB said, it hears a "specific spectrum" NOT ALL. This is to avoid false knocks, but it also excludes the same spectrum WE HEAR which is also a lot of the mechanical noises of the engine. So your "EARS" are another important Knock Sensor that needs to be relied upon. You can think of it this way, "the knock sensor hears frequencies we cannot hear and we can hear frequencies the knock sensor cannot"....you need both. So YES you can get knock which the knock sensor cannot detect.

This is also why a lot of times people THINK they have false knocks (they don't). They are watching the KS pull timing yet they hear nothing. They don't realize the knock sensor hears a different frequency than the human ear. As I said, the frequency that the human ear hears IS NOT what the knock sensor hears. You need BOTH sensors (the KS and your ears).

Also, the knock sensor can get saturated with noise and go deaf. At that point, the "attack rate" goes to zero and the "decay rate" starts to pull the timing and you will get actual detonation. You can actually see the knock retard decay on a scan tool while you ACTUALLY hear knock.

Lastly, there is code within the ECM to make the Knock Sensor "a little tone deaf". During a "shift" with a transmission is a great example, ESPECIALLY automatics during an upshift. If the ECM didn't "go tone deaf", it would trigger the knock sensor EVERYTIME it upshfited and pull a ton of timing out. However, this is also when you are often going to hear the first instance "audible knock" because the knock sensor is purposely "tone deaf" and not as sensative.

 

Can a motor/tranny make a noise that the knock sensor may mistake for a knock? I have 157,000 miles on the motor, tranny, ans clutch. When I am in gear and the motor is hot, in gear and not, when I punch it from a low rpm I hear my motor make a noise which may be a knock and then my computer throws out code 43. I am running 89 octane. When it is cold my car has no problems. Can anybody help?? Mods are in my sig.

 

Yes the knock sensor can be fooled by other "noises". That is why the "knock sensor routine" has "filtering logic" for "upshifts".

As for your question, engines are far more prone to "ping" during hot summer weather. Depending how hot it is, some people have to step up a grade or two in the gas during the summer to stop "ping" especially when they have "load" like climbing up a hill.

 

How does your engine know that you are shifting a manual transmission so that it doesn't pull timing. There is no sensor, so does it just infer this from the inputs from the other sensors

 

I am not sure if my computer knows if I am up or down shifting.
I was just wondering if I should upgrade my ing system to solve this problem. Right now it is all stock. I am going to look into the fuel regulater so I can turn it up about 2 more psi. I have good plugs but fatory replacement everything else. I even think the timing is set a 0*!! Thanks for the help.

 

How does your engine know that you are shifting a manual transmission so that it doesn't pull timing. There is no sensor, so does it just infer this from the inputs from the other sensors.

 

I thought I mentioned it earlier but most of this "filtering logic" SO FAR appears to be more for automatic cars. I've been working with the source code and I have been more concerned with my setup (automatic) than standard.

For SD cars, "standards" appear to be basically the same as the "automatic" code except it doesn't lock the TCC (values all set to zero) and there is a wire lead for the Shift Light.

The code has a number of "flags and constants" where it compares "where it's been, and where it is now".

But, also, there is NO UPSHIFT RETARD ROUTINE. (That is something I wish to add to my code one day). What the filtering routine does is make the ECM less sensitive to the Knock Signal as the rpms are increase in anticipation of shift. Thus the Knock Sensor goes "a little deaf". Unfortunately, this is also when detonation is it's most critical and why guys notice "audible knock" first right as they are about to/or have just started "the upshift".

I've heard that some of the Ford ECMs incorporate a "upshift retard routine", and some of the later GM ECMs may also. But, there is only so many hours in the day, and I have too much to do as it is.

I personally am not a fan of the "Filtering logic" and, as I said, hope to replace it with a "Upshift retard" routine.