2116 replies to this topic

[CHILL Gone DUTCH]

make sure the knock sensor is in the right orientation to 

 

=some body told me it was crucial but cant remember who it was 

[CHILL Gone DUTCH]

 

This is a simple one but a lot of others have suffered with a simple mistake 

After you built the ignition up are you sure you made it permanent on to the ecu 

if not it can reset back to the original settings when you turn ignition off

 

 

 

 

Nope, I've read the manual   

 

 

[TFD]

I have not touched the knocksensor during my build.

[TFD]

@Vocky:

 

http://www.j-body.or...=40&t=85381&a=1

 

 

[fezzasus]

Z22SE uses a single knock sensor which means the ECU takes any input from and pulls back ignition. More modern engines use dual knock sensors which allows identification of which cylinder is causing knock, and also eliminates any false knock events.

 

That said, comparing knock from different days doesn't really help, as any ambient conditions could cause changes in absolute fuel delivery and therefore change how close to the knock limit you're running

[Jason]

Does the orientation really matter as long as its got the correct surface touching its just a microphone..

[CHILL Gone DUTCH]

Does the orientation really matter as long as its got the correct surface touching its just a microphone..

I was sceptacle also but put it as it came of

[vocky]

@Vocky:

 

http://www.j-body.or...=40&t=85381&a=1

 

 

 

I knew it was a few years ago, but 2005

 

There were other threads about it as well, but it's probably not worth doing 

 

ps: but most aftermarket ECUs don't even have a knock function, mine certainly does not

Edited by vocky, 05 July 2016 - 07:55 AM.

[NickB787]

Ok off with the sensor head then, that should stop retard then :-) Actually it will be interesting to see if I get a reduction in knock with the new exhaust which has a flexi joint in my old one didn't have one and the whole exhaust moved slightly under load

[smiley]

For simplicity sake, saying do a long logging run, and then seeing if a 5 degree spike is not repeated in, or near that exact location, can be treated as a false flag, thus not requiring manual ignition correction?

 

[CHILL Gone DUTCH]

Straight up down spikes are considered false knock anyway General knock tends to look more right angles and over a longer period of time

[smiley]

Straight up down spikes are considered false knock anyway General knock tends to look more right angles and over a longer period of time

 

Does the ecu also think that?  

[CHILL Gone DUTCH]

The Ecu ignores anything up to "1" As for it ignoring spikes I think the answer is no As the Ecu won't no what's coming next

[TFD]

The ECU just handles every knock it senses... which makes sense. I've got 7 degree retard knock spikes during boost were I have non during a powerrun... That doesn't make sense.

 

@Vocky: I found more threads about it and yes it works to put a rubber washer or some simular gasket between the KS and the engine. Wrapping the bolt in a teflon coat helps as well. This method seemed to work without shutting the KS completely. Offcourse the "Najj.."-sayers disagreed to do this since you're shutting down a safety device to a sertain degree.

 

Now suhting it down to a sertain degree should not be all wrong if the KS itself is bsicly TOO sensitive.

 

For people without a socal life    : Search for "Phantom knock" and spend a good night reading.

 

I ordered NGK LTR7IX-11 spark plugs and gap them to .0030" (0,8mm). These are 1 step colder then the [color=rgb(40,40,40);font-family:helvetica, arial, sans-serif;]OEM NGK LTR6B-10T. Just to rule out spark blow out or pre-ignition by too hot plugs (hot spot). Then I'll have a look at the knocksensor and start with a little tiny washer and see what happens.[/color]

[CHILL Gone DUTCH]

You could introduce some high ignition just to make sure it still picks knock up in a very small area of the ignition just for one run and to double check you haven't over dampened the knock sensor

[TFD]

Thats what I'm gonna do.

[oakmere]

I also suffer from what I think is false Knock. I have been unable to solve it. I run high lift short duration cams and I think this may give false knock due to lifter noise. Some more advanced systems take a base reading at idle to take into account engine noise.

[Exmantaa]

 

Does the orientation really matter as long as its got the correct surface touching its just a microphone..

I was sceptacle also but put it as it came of

 

 

Somewhere on the LNF (= turbo Ecotec) engine forum the intstallation manual says to mount the Knock senor with the pigtail exacty horizontal with the correct torque. (19ft/lbs?)

Some guys there who had bad false knock experimented with poisition and found that orientation 180* the other way horizontal made a huge difference on his car.

On another car type the exact torque made a lot of difference, so try it out?

 

(Can you not distinguish false knock from real cylinder knock by checking every cylinder? real knock at a certain load should come from the same knocking cylinder...)

Edited by Exmantaa, 05 July 2016 - 10:35 PM.

[TFD]

http://www.speedster...and Install.pdf

 

The manual for the Z22SE says nothing about location of the knocksensor. Just bolt it back at 25Nm.

I wrapped the bolt in 2-layers of teflon and put a small paper gasket between the engine and the sensor, turned it firm hand tight back to the block.

We'll see what happens. Do a few runs on my current map as reference.

 

Will also fidle with AFR's. Leaner AFR for more retarded timing, richer AFR for more advanced timing (still confused why Chris is running 12:1 AFR with 5-7 degree more advanced timing than me, the difference is severe. Log from Smiley does this as well.).

Edited by TFD, 06 July 2016 - 09:09 AM.

[Exmantaa]

Yup, 25Nm/18ft.lb.

(I would try the horizontal orientation too.)

 

Only difference with a supercharger is that your engine will see quite some exhaust pressure before the turbo. Not sure what this does to your knock behavior if you still have the NA cam overlap...