Technical discussion...

[Tex Arcana]

Alright, I was on another forum, where the talk turned to knock sensors, and why Lightnings dind't have them (tho someone said they're there, just turned off). Another person posted posted this link (http://www.jandssafeguard.com/index.html) to a product that claims to use one knock sensor to detect individual cylinder knock, and retard spark timing for that cylinder. Which in and of itself is a good idea, especially for boosted applications.

But that brings up a question: why *not* set up individual sensors for each cylinder?? Yes, I know it could be expensive: ignore that aspect. Discuss why this is a good idea; or, if not, why it's a bad idea.


A related discussion: why *not* set up individual O2 sensors, to detect individual cylinder air/fuel conditions? Again, ignore cost, and discuss the relative merits, and implementation obstacles.

Thanks in advance.

EDIT: To clarify, I'm not talking about *just* our motors, I'm talking about motors in general; if you *have* to talk about ours, consider one with a centriblower or turbo(s). :)

[Silver_2000]

Alright, I was on another forum, where the talk turned to knock sensors, and why Lightnings dind't have them (tho someone said they're there, just turned off). Another person posted posted this link (http://www.jandssafeguard.com/index.html) to a product that claims to use one knock sensor to detect individual cylinder knock, and retard spark timing for that cylinder. Which in and of itself is a good idea, especially for boosted applications.

But that brings up a question: why *not* set up individual sensors for each cylinder?? Yes, I know it could be expensive: ignore that aspect. Discuss why this is a good idea; or, if not, why it's a bad idea.


A related discussion: why *not* set up individual O2 sensors, to detect individual cylinder air/fuel conditions? Again, ignore cost, and discuss the relative merits, and implementation obstacles.

Thanks in advance.

If you look back in time like 4 or 5 years you will see the issue was covered ad nauseum...
The summary is ....
The mod motor is too noisy for a knock sensor to work reliably

[wesman]

I don't think that the mod motor is as much the reason as the sueprcharger.

--wes

[wesman]

BTW- I had a J&S on my SC'd 350Z....very nice little unit. I can go on about it's capabilities if there are questions.

--wes

[Tex Arcana]

If you look back in time like 4 or 5 years you will see the issue was covered ad nauseum...
The summary is ....
The mod motor is too noisy for a knock sensor to work reliably

Sorry, I meant motors in general, not just ours. Edited post to reflect correction. :)

[Tex Arcana]

BTW- I had a J&S on my SC'd 350Z....very nice little unit. I can go on about it's capabilities if there are questions.

--wes

What sort of blower? Tell me more about it. :)

[Silver_2000]

I don't think that the mod motor is as much the reason as the sueprcharger.

--wes

Well I was close

either way the lightning is too noisy

[Beaudee]

The heck with the sensors,all you need is a good ear or a Terry Cole.:tu: The more whistles and bells the more stuff to be a nusance.Good post tho Tex!!

[wesman]

Well I was close

either way the lightning is too noisy


Indeed. Hell, I can't hear my radio half the time over the motor/exhaust/SC. I know a little sensor on top of the engine would have trouble.

[Tex Arcana]

interesting update from the other thread:


There are systems today that allow individual cylinder knock control, but those generally rely on one sensor per cylinder bank to detect vibrations from outside the cylinder. TI is making a new device that is actually inserted inside the cycliner to measure different situations. They say "External noise or signals from other sensors often can confuse the knock sensors", which the TI one is immune from.

I assume as it stands now, putting a knock sensor in the block outside each cylinder would cause confusion because they are so senstive..it probably wouldnt know if it's cylinder 1 or 3 that is pinging half the time. From all the things I hear, only 1 is needed per side at most because they can tune the response to try to detect which cylinder needs a timing change.


i would think the reason you dont need 1 per cylinder is that the engine knows which cylinder was firing when the knock is detected. since all cylinders are not firing at the same time, i can tell which cylinder needs to be retarded. i would assume one per bank would just make the signal the KS picks up a little more defined. thats just guess though, i could be way off


iirc detecting detonation is just a frequency sensor that detects the natural resonation of the metal that the engine is made of. Which is chimed like the string of a guitar when there is detonation. So detecting the actual source might be hard, since all the knock sensor really sees is the result of the actual detonation. So one sensor per sizeable chunk of metal is not only all that is needed, but already at the highest resolution of the knock sensor so to speak.



ard|On;20599280']They can sense it before you can hear it no less.

I saw an LED knock light in someone's car once, and wanted to put the same in my vette. But you could whack away at the manifold with a mallet, and watch that thing light up just before you heard a noise. Impressively weird.





Tex you should know that knock sensors are outdated and inaccurate. Ion sensing is the newest method.

http://www.fs.isy.liu.se/~larer/Projects/main.html
http://www.swri.edu/3pubs/IRD2002/03-9305.htm

GREAT DIY article.
http://vvnet.fi/ville/ion/DIY-Ion-Sensing-2.pdf

Damn, good info here. :D Thanks, folks