91slowrunner

I've been dealing with the classic code 52 knock sensor problem, but after a replaced sensor and pigtail i still had the problem. Well i started testing continuity and found that a had corrosion within my wiring harness on the engine ground wire. The knock sensor has two wires a sensor and an engine ground. I found that the engine ground is wrapped around the sensor wire but does not ground to anything, in fact it stops at the end of the pigtail before it plugs into the sensor. Shouldn't this wire be ground out to something? Otherwise it seems like this engine ground that runs all the way to the ecu is pointless?

Cyberman

The "ground" that is wrapped around the sensor wire is a sheild. It is grounded at the sensor housing, which is screwed into the block, which is a ground point. The shield helps to prevent interference on the knock sensor signal.

91slowrunner

But if you look at the pigtail you see that the ground/shield wire ends before the sensors plug. The plug only has one pin for the sensor wire. At least this is the way all the pigtails I have seen are. Am I all wrong?

91slowrunner

Stole this off another forum, sound right?
Anonymous: "The sensor works on a process where it vibrates at high frequency. This frequency creates a voltage which is then sent to the computer for inspection. So its putting out a voltage even when disconnected. You can put on a good voltage meter and see what its output is. Then rev the engine and continue checking. If the frequency and voltage are OK then the computer sets the program accordingly for fuel injection. If the reading is not OK then it turns on the light and sets the fuel injection volume down.
The signal is critical to performance and needs to have a good ground. I ran an extra ground wire from the battery to a location near the sensor. Then I attached it to the grounded shielding on the sensor wire. This helped for awhile but the real problem was the wire itself. I think the heat and high frequency caused internal leakage in the old wire. So I replaced the whole wire and ran it along the rear of the fire wall instead of in the wire harness. Then reattached the wire to my good ground. This fixed the problem.

Because of the high frequency you have to have a good meter that can read the voltage produced. Some shops will inspect this for free. I will try and find the information and give an update"

muddpigg

iTrader: (1)

Strange everything I read about knock sensors had to do timing advancement that is controlled by the ECU. Advancing till it "knocks" for better excelleration. We may be talking about to different engines and sensors. For us 22re guys, forged pistons play hell with it when cold so you need to relocate the sensor to lower in the block or it retards the timing till issues araise.

91slowrunner

Ive got a 3.0 but I have heard horror stories about those forged pistons. Anyone else? To ground or not to ground?

abecedarian

cyberman is correct about the shield being to keep interference off of the knock sensor signal wire... interference from the ignition, injector pulses, etc. there is no terminal for the sheild to connect to- it should be crimped in the connector. it is known though for the shield to separate and loose connectivity to ground. yota should've grounded it at either the firewall or at the ecu though.

muddpigg is correct about the forged pistons- they tend to make more noise than cast pistons and the ecu often confuses this with detonation so moving the knock sensor lower in the block isolates that noise.

the knock sensor is a piezo-electric device which functions as a microphone which the ecu uses to hear all the mechanical noise within the engine, listening for sounds similar to detonation and when it hears such noises, retards the timing. once the noise is back to 'normal' the ecu begins re-advancing the timing. if detonation is heard again the process repeats. the ecu's range of adjustment is limited though and with too much initial advance cannot retard the timing enough to eliminate detonation in those cases, or with other serious detonation.

there is no real way for a home mechanic to test one. typically, you'd use an oscilloscope and computer software to analyse the signals to determine it is operating properly.

apalmer1

as stated above the KS is a piezoelectric crystal tuned to produce voltage when exposed to the vibrations typically caused by knocking. If you think that your aftermarket pistons are causing this knock (which i doubt they are but its a possability) you can try either moving the sensor as recomended above, or putting a few of the absolute smallest resistors possible in parallel in the sig wire from the sensor to dampen the signal to the PCM.

If you have any friends with a DSO (digital storage oscilloscope) or a very very good multimiter (fluke 88, et.) you can have them probe the sensor wire and then hit the block with a hammer or whatnot and see the signal, then just watch it when its running and see if the sensor actually is picking anything up.

khoopes

All good sounding suggestions, I'd like to add that it is important to keep ground loops from occurring and flowing through your shield. Often one end of a shield is floated in electronics to prevent this. If both ends are grounded make sure all of your block and body grounds are secure, corrosion free, and that none of them are missing. The idea is to keep all of the electronics at the same ground potential. You can use a good digital multimeter / voltmeter and check the voltage from the negative terminal of the battery and probe the various ground points on the vehicle's body and engine and check for any readings above a few millivolts. In an ideal world the readings would all be zero but I don't know what the acceptable range is. Although it may not solve the problem specifically this is good practice for other problems too like alternator noise in the radio, ignition noise, etc..

My2cents...

Oh, yes, you need to do the above procedure with the engine running so watch out for the fan and pulleys...

91slowrunner

Ok, so to update. I grounded out the shield wire and eliminated the knock sensor problem. Unfortunately in doing so I have created a horrible engine noise through my amplified stereo system. It is grounded to my seat mount bracket, any suggestions?

abecedarian

how did you ground the knock sensor shield? is it grounded at both ends or just one?
if so, break the ground at the engine end, or ground the stereo system directly to the battery.

91slowrunner

One end is grounded to the center of the block and the other end to the engine ground on the ecu.

a4runnerfreak

Should always ground sound systems to the batt. Same CB radio's, and linear's.

Bill54880

So what are the symptoms of a bad or over reactive knock sensor?