Jump to content

DIT knock detection questions


turbogrill
 Share

Recommended Posts

Hi

 

There are no good bulletproof  knock detection systems for my predetenetion prone engine (l28). The stock turbo model has some but its not viable to use.

 

Playing with the idea of having a knock sensor connected to a raspberry pi audio input and do realtime dsp operations to detect knock. This could then be fed to ms for timing adjustment.

 

A few questions:

 

1. Are some cylinders more knock probe than others or is it randomly distributed?

 

2. Is 1 knock sensor for all cylinder good enough? That is what stock has. Makes it easy.

 

3. Is there such a thing as a harmless light knock? 

 

4. Many algoriths use a window 10-70 degrees atdc to detect knock. Wouldn't a detenation happen before that?

 

5. Is the knock it self very different from the normal combustion? Is the flame front much faster?

 

6. My plan was to not use a knocking window, is there any other engine events that could be mistaken for a knock?

 

There is a TI app note on how to use DSP for it. They simple do a DFT analysis for a few different predefined frequencies

and compare with a threshold. The frequencies are a function of engine size and expected harmonics.

They also adaptivly update the threshold when no knock can happen. That requires crank position info.

 

Thanks

 

Link to comment
Share on other sites

Many engines will "knock" to some degree under light load, part throttle.  Ive also experienced false knock detected by the computer caused by rattling clutches, or loose heat shields.

 

What is it that makes the L28 inherently knock prone?  Could it be that the knock is a symptom of another issue?

 

The only reason I ask is, Subarus motorsports ECUs, for say Group N competition have knock control disabled, even though the boxer engine is known for being "weak" to repeated knock events.

Link to comment
Share on other sites

16 minutes ago, LAMR2 said:

In your L28, cylinder 5 is going to knock before any of the others. There's a localized hot spot due to how the coolant circulates, so that cylinder is always the hottest unless you drill some holes and reroute the coolant externally.

 

I guess that is good in a way, since then I only need one knock sensor :)

 

14 minutes ago, SonsOfIrony said:

Many engines will "knock" to some degree under light load, part throttle.  Ive also experienced false knock detected by the computer caused by rattling clutches, or loose heat shields.

 

What is it that makes the L28 inherently knock prone?  Could it be that the knock is a symptom of another issue?

 

The only reason I ask is, Subarus motorsports ECUs, for say Group N competition have knock control disabled, even though the boxer engine is known for being "weak" to repeated knock events.

 

The first generation cylinder heads was poorly designed but it got much better after 1981 I think. Also the stock EFI is terrible in spark and fuel control but a modern management system obviously solves that.

 

 

Link to comment
Share on other sites

I simple way to test is it a real knock is use a lap scope with a RPM point. Some thing we did on circle track before all this control stuff is watch our spark plug if we saw some metal balls we =where too lean/timing = knock

I did a lot of CNG design in kit building on a few engines to slow down the ECU to measure the knock I used copper washers on both sides of the sensor bolt . This soften up the signal .  

Link to comment
Share on other sites

2 hours ago, DRVOLKS said:

I simple way to test is it a real knock is use a lap scope with a RPM point. Some thing we did on circle track before all this control stuff is watch our spark plug if we saw some metal balls we =where too lean/timing = knock

I did a lot of CNG design in kit building on a few engines to slow down the ECU to measure the knock I used copper washers on both sides of the sensor bolt . This soften up the signal .  

 

What do you mean by lap scope and RPM point?

Link to comment
Share on other sites

10 hours ago, turbogrill said:

5. Is the knock it self very different from the normal combustion? Is the flame front much faster?

Yes, very different. Detonation occurs when the pressure and temperature from normal combustion inside the combustion chamber reach a point that causes the remaining mixture to spontaneously combust. This usually starts close to the exhaust valve, which adds enough energy to light off the mixture before the regular flame front gets there. The two pockets of combustion raise the pressure much quicker than normal and the pressure spike hits like a hammer. The knocking or pinging sound is the metallic ringing of the cylinder block from the spike.

 

(P.S. no such thing as "predetonation"; it's either pre-ignition or detonation)

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...