Explanation. My 383's block is filled up to the first freeze plug. My knock sensor was moved to the driver side because of my 2210 LT headers. My ecm is a 749 and I'm running the $59 code. Tunerpro RT and Ostrich 2.0 are my tools. I took the car out for a 15 minute jog a few weeks back and the car ran pretty well BUT, looking at ANY datalog my knock sensor readings look like a saw tooth blade in the histogram. I mean they start at 5000 or so and go all the way up to like 400000 before dropping back to 5000 and doing the same thing over and over and over and over. I'm a little stumped as to what I can due to correct what I feel is a false knock reading. Everything is brand new and I mean brand new...oh yeah....93 Octane is the only thing this engine sees.

 

I get readings like that when its not connected.
Goes from 0-255 constantly.
Try putting a resistor in place of the sensor to see if the wiring may be a problem. 3.9K IIRC is for 350 ci.

 

jp,excuse my ignorance but how do you wire up a resistor to the connector?

 

bump.

 

I'm going to follow the factory service steps and just check the sensor with an ohmmeter and look for anywhere between 3500-4500 ohms.

 

Sorry I didn't see this earlier.
I just connect one end of a resistor to the connector (ball up the end and jamb it in) and then ground the other side to the motor/chassis.
3900 ohms is for a 350 KS. and should allow you to run without the sensor to see if the sawtooth goes away.
If it does you may want to desensitize the sensor by installing a 1/4" npt 90* fitting. I've heard that can help with bigger cubes. Not sure how it will react to the filled block though.

 

No prob on not seeing.
I've got the feeling my filled block is going to be a gremlin with knock sensing....SO, I'm searching when I can for an answer but for now I'll try the resistor tip and see what my knock counts end up at.

 

The factory KS is useless on anything other than a stock engine(for the most part). It's just crystals that vibrate at a certain frequency, creating a signal. The harmonic frequency of a 383ci motor is not exactly the same as a stock L98. The knock sensor would go crazy on my externally balanced 400 sbc.

 

Hmmm...good points. What was your solve for the erractic readings on the 400?? I assume you're still using EFI?

 

Just didn't use a knock sensor. None of my cars run one now. I learned the "real" ways to sense detonation...

 

"Real" being plug reading???? I'm only running a 8.7:1 motor without boost and ALWAYS run 91 or better gas. Once I put my blower belt on (after extensive tuning naturally aspirated) I'll run only 93 or better and obviously retune.

 

bump

 

Plug reading certainly helps...

Once you see salt and pepper insulators, you are in dangerous territory.

 

My thought is plug reading is an "after effect" of what using a knock sensor is supposed to be used for. IF, you don't use the knock sensor then the ECM can't pull timing, resulting in the detonation not being combatted.
Plug reading isn't doing anything to show you when detonation occurs (datalogging), only evidence if its occuring.
I'm just vary curious to see what other ways there are.....I mean how'd the carb guys do it before EFI and knock sensors.

 

I've heard you can solder a resistor to the knock sensor wire then solder the resitor to a ground anywhere and it will fool the ecm into thinking there is a knock sensor.

 

Ears !
Take a hunk of garden hose. Stick one end in your ear, and the other shoved tightly ove something on the block.
Might be surprised what you hear.

A knock sensor is basicly a microphone, doing essentially the same thing.
The one on my big block came from an 87 350. Works fine !

 

Guys, these are all fine methods while idling or blurping the throttle by hand but what about on the fly and driving?? I drive / datalog with my laptop. Once back in the garage I playback the run and would look for knock counts then reference what RPM / VE areas were producing the knock and adjust accordingly. Someone had mentioned the "real" way and I was curious to what his definition of that was.

 

In the good ol days you set your advance curve on the dist , and just kept winding the dist around until you heard "audilbe" knock when under load ;
say WOT in top gear at low revs.Take it back a fraction.
That was the definitive max advance the engine could handle.

Maybe the electronics have taken us to a higher level of measurement where we are getting knock counts /false knocks registering that are not detrimental to engine operation.
As stated the harmonics of a modded engine are going to different from a stock one.

The resistor takes the sensor out of the equation altogether.You will get no knock counts as it is seeing a fixed resistance.No counts = no spark retard Just fools the ECM into thinking all is right so it doesn't throw a code .

Maybe adjust your knock settings so you can log indicated knock counts without ECM pulling any spark until you determine whether or not you actually have true detonation.