Bimmerfest - BMW Forums

Bimmerfest - BMW Forums (http://www.bimmerfest.com/forums/index.php)
-   E39 (1997 - 2003) (http://www.bimmerfest.com/forums/forumdisplay.php?f=103)
-   -   Knock Sensor code? Causes Replace? (http://www.bimmerfest.com/forums/showthread.php?t=542640)

redbull713 05-19-2011 11:35 AM

Knock Sensor code? Causes Replace?
 
Hello forum, I have some questions regarding knock sensors and their corresponding codes. I had scanned my 97 540i with a peake reader and had recieved a bunch of codes related to multiple cylinder misfires and fuel trim. Now i have torn the intake mani out to do the valley pan job as well as replace a bunch of gaskets etc. The code I got was this: "D5 knock sensor Cyl. 7-8" My question is: Does this mean that my sensor is bad?? OR does it mean that the computer registered a knock/ping from this sensor on these cylinders?? I dont know if I should just replace the bugger while im in there. Thanks in advance gang.

redbull713 05-19-2011 02:10 PM

Also it looks as if the wire insulation that goes from the sensor to the connector is smashed/warped a little like it was caught between something. Has anybody else had this problem before?

96 GGM 528I 05-19-2011 06:38 PM

The wire normally used for knock sensors is a shielded wire, so in short if it is bare at all you need to replace it. Shielding prevents the wire from hearing electrical interference given off from the alternator and a few other electrical components that contain motors. When the codes go off and it says the knock sensor is bad then it probably is. The factory knock sensors attach to the side of the block and listen to a certain frequency to interpret as pinging in the engine. When it hears this pinging it will pull timing to the motor as an effort to lower cylinder pressures which relieves stress on the engine. When you reduce timing it fires the plugs sooner meaning the compression stroke is not quite complete but this relieves stress on the engine. I would make sure everything else is up to snuff. Also, I would pull your plugs and 'read' them. You need to look at the tips and look for little specs of metal looking similar to pepper on the plug. This indicates pre-ignition of the cylinder which=knock.

bluebee 05-19-2011 10:11 PM

Quote:

Originally Posted by 96 GGM 528I (Post 6077569)
it will pull timing to the motor as an effort to lower cylinder pressures

Are you sure about that?

I thought knock sensors feel the vibration cased by the flame front igniting the gas "too fast" because the bulk of the explosion occurred while the piston was still moving up on the compression stroke. I had thought that retarding the timing merely moved the point of ignition later in the compression stroke so that maximum pressure occurs when the piston is at top dead center.

My point is that I always thought the compression was the same (it's a mechanical thing ... how can it 'not' always be the same?) whether or not the timing was retarded.

Anyway, I looked in the bestlinks for an explanation of how the knock sensors worked but nobody put a reference link there.

Searching for "knock sensors", I find a few that might have more information for the OP:

rdl 05-20-2011 04:42 AM

Quote:

Originally Posted by bluebee (Post 6077943)
Are you sure about that?

I thought knock sensors feel the vibration cased by the flame front igniting the gas "too fast" because the bulk of the explosion occurred while the piston was still moving up on the compression stroke. I alway thought that retarding the timing merely moved the point of ignition later in the compression stroke so that maximum pressure occurs when the piston is at top dead center.

My point is that I always thought the compression was the same (it's a mechanical thing ... how can it 'not' always be the same?) whether or not the timing was retarded.

Anyway, I looked in the bestlinks for an explanation of how the knock sensors worked but nobody put a reference link there.

Searching for "knock sensors", I find a few that might have more information for the OP:

Wikipedia article describing ping/knock
http://en.wikipedia.org/wiki/Engine_knocking

From BMW's WDS - description of knock control
###############
Knock Control


Tasks of the knock control system
Operation of an engine with knocking combustion over a prolonged period can lead to serious damage. Knocking tendency is increased by:

•Increased compression ratio

•High cylinder charge

•Poor fuel quality (RON/MON)

•High intake air and engine temperatures

The compression ratio can also reach excessively high values due to deposits or production-related scatter.

On engines without knock control, these unfavourable influences must be taken into consideration in the ignition design by providing a safety distance to the knock limit. However, this results in unavoidable losses in efficiency in the upper load range.

The knock control can prevent knocking engine operation. For this purpose, if there is an imminent risk of knocking, the control retards the ignition timing of the corresponding cylinder(s) (cylinder-selective) as far as necessary. This makes it possible to adapt the ignition characteristic map to the consumption-optimum values without having to take the knock limit into consideration. A safety distance is no longer necessary.

The knock control system carries out all knock-related corrections to the ignition timing and enables perfect operation also with regular grade fuel (minimum RON 91).

The knock control provides:

•Protection against knocking damage also under unfavourable conditions

•High efficiency due to optimum utilization of the fuel quality and consideration of the relevant engine status

•Logistics advantages with regard to fuel availability

•Lower consumption and higher torque over the entire upper load range (corresponding to the fuel quality used).

Design of knock control system
The M54, the M52 and the M43 are equipped with a cylinder bank-selective, adaptive knock control system. A double knock sensor detects combustion knock. The sensor signals are evaluated in the DME control unit.

The knock sensor is a piezo-electric structure-borne noise microphone. It picks up the structure-borne noise and converts it into voltage signals.

Function of the knock control system
If knocking occurs, the ignition is retarded for a certain number of working cycles and then gradually approaches the original value. The retard setting can be controlled individually for each cylinder bank (cylinder bank selective).

In the event of the knock sensor failing, a fault code is entered in the fault code memory of the DME control unit. In the case of fault, the engine is protected by constant retard setting of the ignition timing.

Installation Locations/Conditions
The double knock sensor is secured by means of an 8 mm screw on the water jacket of the engine block between both cylinder banks. It is arranged such that each sensor monitors one cylinder bank.

Only screw locking compound may be used to lock the screws. Washers, spring washers or serrated lock washers must under no circumstances be used.

Self-diagnosis and emergency operation of the knock control system
Self-diagnosis of the knock control system includes following checks:

•Check for sensor signal interference/line break, plug connector defective etc.

•Self-test of entire evaluation circuit

•Check of basic engine noise level detected by the knock sensor

The knock control system is switched off if a fault is found during the course of one of these checks. The emergency program adopts the task of controlling the ignition timing. At the same time, a defect code is stored in the defect code memory. The emergency program ensures damage-free operation as from minimum RON 91. It depends on the engine load, speed and temperature.
#################

Regards
RDL

bluebee 05-20-2011 07:05 AM

Quote:

Originally Posted by rdl (Post 6078166)
description of knock control

Very interesting.

I'll add this to the bestlinks.

BTW, notice it didn't say anything about retarding the timing reducing compression. The only statements about compression are the typical ones for what causes knocking in the first place (which is a different thing).

96 GGM 528I 05-20-2011 07:28 AM

"If the spark occurs too retarded relative to the piston position, maximum cylinder pressure will occur after the piston is already traveling too far down the cylinder. This results in lost power, high emissions, and unburned fuel."~Wikipeda

Like I said the pulling of timing reduces cylinder pressures lowering stress on the motor so that it will stop knocking. As the poster above me and I previously said the sensor interprets a certain frequency as knock and retards timing. If your car is throwing the code it has been reading knock so often that it no longer assumes it to be cause by bad fuel, vacuum leaks, or any other problem but rather lies in the sensor itself. Like I said just to double check that it is the fact that some of the shielding is damaged causing the bad reading I would read the plugs that will tell you if the cylinder is pre igniting. I have been using stand alone computers to tune modified engines for a number of years and am quite familiar with knock sensors and such. Also on extremely modified engines the sound that represents knock can change as your piston surfaces change to machined forged aluminum and cams change to aggressive cuts. So that is one reason why we use stand alone systems on extremely modified engines because factory knock sensors will kill timing on a built motor in some cases.

Below are some links about the shielded wire on knock sensors. Like I said in my first post in short Yes wire damage even insignificant will cause interference.

http://www.supraforums.com/forum/sho...-sensor-rewire

http://www.dsmtuners.com/forums/tuni...esistance.html

http://www.dsmtuners.com/forums/newb...ck-sensor.html

http://www.yotatech.com/f116/knock-s...ed-huh-105393/

How to read plugs http://www.nightrider.com/biketech/spkplghnbook.htm

bluebee 05-20-2011 08:51 AM

Quote:

Originally Posted by 96 GGM 528I (Post 6078432)
maximum cylinder pressure will occur after the piston is already traveling too far down the cylinder

Ah, my mistake. I read "compression" when you were actually talking about "cylinder pressure". My fault.

I don't equate 'compression' with 'maximum cylinder pressure' because the former is mechanical (measured without any spark plugs or spark involved - but with valve timing a key component) while the latter is due to the result of flame front dynamics inside the cylinder (where both spark & valve timing are involved).

96 GGM 528I 05-20-2011 09:32 AM

"When it hears this pinging it will pull timing to the motor as an effort to lower cylinder pressures which relieves stress on the engine." ~Quoted from my first post

Sorry for the confusion and I will try to be more clear in my post. Sometimes I get all wrapped up thinking about it and it comes out poorley. I am a Mechanical Engineer and our college curriculum had Technical Report Writing instead of a normal writing class and they defiantly didn't focus to much on sentence structure but rather report material. Again I apologize for any confusion.

redbull713 05-20-2011 10:28 AM

Thanks guys for all of the input/help. This forum is an great asset. However, nobody seems to have answered my question. Does the knock sensor code appear due to a faulty knock sensor OR does it appear when/if the knock sensor has registered a ping and then store this as a code?

96 GGM 528I 05-20-2011 10:48 AM

"If your car is throwing the code it has been reading knock so often that it no longer assumes it to be cause by bad fuel, vacuum leaks, or any other problem but rather lies in the sensor itself. Like I said just to double check that it is the fact that some of the shielding is damaged causing the bad reading I would read the plugs that will tell you if the cylinder is pre igniting." ~Quoted from above

Thought that would answer it for you. If your car reads knock it does not throw a CEL. It throws the CEL because the resistance in the wire is high (Due to damage or corrosion inside the casing) causing it to always hear a minute amount of knock. It also could get a bad signal because if you followed my links the wire has a steel mesh wrapped around the copper core. The steel mesh is there to capture noise from various electrical components. That mesh should be grounded on only one side of the sensor so that the noise it is hearing does not become harmonic and grow when transfered through the casing. But like I said above READ UR PLUGS TO BE SURE. follow the link and look for salt and pepper on the plugs.

bluebee 05-20-2011 10:49 AM

Quote:

Originally Posted by redbull713 (Post 6078857)
OR does it appear when/if the knock sensor has registered a ping and then store this as a code?

I don't know the answer ... but I can tell you this ...

I've been using 87AKI Costco gas since I bought my 2002 525i and I'm sure the knock sensors are going crazy as the DME retards my timing ... yet there is no fault code.

96 GGM 528I 05-20-2011 11:53 AM

I have a car with a stand alone computer and it allows me to set it to illuminate the light when it sees over 5 deg of knock. On most cars 3deg of knock and below the computer quites advancing timing and leaves it where it is. Above 5 Deg it starts to retard timing. On my race car when the CEL does come on due to knock it goes off as soon as the knock goes away.

rdl 05-20-2011 06:53 PM

Quote:

Originally Posted by bluebee (Post 6078381)
Very interesting.

I'll add this to the bestlinks.

BTW, notice it didn't say anything about retarding the timing reducing compression. The only statements about compression are the typical ones for what causes knocking in the first place (which is a different thing).

You have a good point.

The key point is that at a critical level of pressure and temperature, the mixture of air & gasoline will ignite spontaneously - without any spark or flame being present. The entire volume of mixture burns intantaneously like an explosion, with a bang => knock.

When the piston comes up on the compression stroke the air/fuel mixture rises in temp and pressure due to this compression. Once the spark ignites the mixture, usually before TDC, a flame front progresses throught the mixture, releases heat energy which increases temp and pressure futher. We then have a race between the rising temp and pressure due to further compression and the additional heat from combustion against either:
- the flame front burning all the mixture before TDC
or
- the piston passing TDC, allowing the gases to expand which reduces temp and pressure.
If conditions lead to some of the unburned mixture reaching the critical temp and pressure that remaining volume goes bang. If it is a small volume we get a ping, if a large amount knock.

The solution to this problem is to retard the timing, i.e. make the spark a little later on the next cycle. Not necessarily to after TDC, but enough that the flame front wins the race on the next cycle and there is no unburned mixture by the time the flame burns itself out with all the fuel consumed.

Interestingly, many engine designs walk to the edge of ping/knock and normally run with a small amount of ping, usually inaudible. The reason for this is that thermodynamic laws tell us that the higher the max temp and pressure, the higher the thermal efficiency. Higher thermal efficiency give us better fuel ecomomy and higher power from a smaller engine. So the designer makes the engine parts strong enough to tolerate the little bit of shock from a ping, and puts in controls to ensure that heavier, damaging knock is avoided.

Compression ratio is not the be all and end all of ping/knock. The gases in the combustion chamber don't know what the compression ratio is; they only know temp and pressure. If one fills the cylinder with less mass (fewer molecules) for instance at 10,000 ft versus sea level, compression temp and pressure will be lower and no ping. This explains why an engine can run happily at light throttle but knock badly at WOT. Light throttle mimics high altitude (thin air), WOT lets lots of air/fuel mass into the cylinder. Conversely, turbocharging packs the cylinder "fuller" than possible with normal asperation with pre-compressed air/fuel. Therefore they need to run lower compression ratios so that the critical temp and pressure is not reached.

Similarly, the race between the flame front racing against the piston passing TDC explains why knock is more common at low RPM with WOT. At WOT lots of air/fuel gets into the cylinder and compression results in high temp and pressure before the spark. The flame front progesses at (nearly) constant speed, but if the piston is slow, the critical temp and pressure is achieved and we get the bang => knock. At high RPM, the piston wins the race and no knock.

Regards
RDL

bobdmac 05-20-2011 10:13 PM

Quote:

Originally Posted by bluebee (Post 6078902)
I don't know the answer ... but I can tell you this ...

I've been using 87AKI Costco gas since I bought my 2002 525i and I'm sure the knock sensors are going crazy as the DME retards my timing ... yet there is no fault code.

Your engine management may be compensating by varying your valve timing. Double VANOS allows modification of both intake and exhaust valve timing. If that's sufficient to stop the detonation, it wouldn't produce a code.

Burning2nd 05-21-2011 03:40 AM

Quote:

Originally Posted by bluebee (Post 6078902)

I've been using 87AKI Costco gas since I bought my 2002 525i and I'm sure the knock sensors are going crazy as the DME retards my timing ... yet there is no fault code.

This is why your car really doesnt require premium,
It adjusts Itself

Ping, knock, Detonation.... All stem from the same set of variable

rdl 05-21-2011 05:49 AM

Quote:

Originally Posted by redbull713 (Post 6078857)
Thanks guys for all of the input/help. This forum is an great asset. However, nobody seems to have answered my question. Does the knock sensor code appear due to a faulty knock sensor OR does it appear when/if the knock sensor has registered a ping and then store this as a code?

As I re-read this thread I see that you have a good point.

Generally, a fault code listing a specific component indicates that that component is faulty, i.e. values being reported are inconsistent with a properly functioning part. Symptoms from apparently good sensors are reported as such. E.g.
Component => HO2S Heater Control Circuit (Bank 1 Sensor 1)
Symptom (from sensor) => System Too Lean (Bank 1)

Combined with your description of a damaged wiring harness, the sensible conclusion is that the fault is reporting a bad component. But, this is a conclusion based on an inference, not definite knowledge.

Regards
RDL

Flug540 05-21-2011 06:56 AM

Quote:

Originally Posted by redbull713 (Post 6078857)
Does the knock sensor code appear due to a faulty knock sensor OR does it appear when/if the knock sensor has registered a ping and then store this as a code?

You can try measuring resistance of your sensors to verify if there is a short, I believe it should be about 4.96 MOhm. I'm addressing my sensors today by reinsulating the entire length of the cable. I did not have any codes yet, but as I was going through other maintenance I noticed that the insulation is failing badly. At first I thought I'll just reinsulate the part that is closest to the plugs, but the closer you get to the sensor, the worse it becomes. I got another pair of sensors from a member on this board and it was in exact same shape. There was no short yet, but I won't be surprised if some of them do, you just need to bend the cable a couple of times, and you may get it.

Flug540 05-21-2011 08:00 AM

2 Attachment(s)
To give you an idea of what it looks like if you pull the outer insulation. As you can see there is still some yellow insulation, which is very brittle, left towards the plug, but as you get closer to the sensor, it literally turns into dust as you pull the outer shell off.

redbull713 05-29-2011 03:57 PM

Wow, thank you very much to all who posted here. This thread is valuable in the discussion of knock sensors, maybe place in best links?

bluebee 05-29-2011 04:13 PM

Quote:

Originally Posted by redbull713 (Post 6098236)
maybe place in best links?

I had added RDL's specific post to the bestlinks for knock sensors ... but ... I see it's all lost now (due to a weird problem that is truncating the best links).

So I lost everything I had added about knock sensors. Sigh.

Flug540 05-31-2011 07:06 AM

9 Attachment(s)
I finished rebuilding my sensors over the weekend. Wasn't too bad, the first one took the longest time while I was trying to figure out the best technique. Below are some photos of the process and what they look like now. On the first two photos you can see what happens to the original internal insulation when you pull the outer shell. Resistance on all sensors actually gone up slightly, now it's about 4.98 MOhm, which I think shows that the original insulation already was beginning to conduct.

redbull713 05-31-2011 12:34 PM

Wow, good work Mr. 540Alex. I really love this forum, I never cease to be surprised by the stuff you guys do! I am dumb when it comes to wiring/electronics so I shy away from this type of stuff, but maybe I could learn...

Burning2nd 05-31-2011 02:11 PM

see... people who do work like this... deserve a standing applause.... big up's (that means respect for you older crew) sir

keep in mind that the threshold of the sensor reading is critical to the computers function

redbull713 05-31-2011 05:20 PM

Quote:

Originally Posted by burning2nd (Post 6101783)
see... People who do work like this... Deserve a standing applause.... Big up's (that means respect for you older crew) sir

keep in mind that the threshold of the sensor reading is critical to the computers function

+1


All times are GMT -7. The time now is 12:26 AM.

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2014, vBulletin Solutions, Inc.
© 2001-2011 performanceIX, Inc. All Rights Reserved .: guidelines .:. privacy .:. terms