Notices
ECU Flash

Knock Control Questions (2.4L w/ Phantom Knock Issues)

Old Jun 9, 2015, 11:02 PM
  #1  
Evolving Member
Thread Starter
iTrader: (2)
 
Live4Redline's Avatar
 
Join Date: Apr 2008
Location: Richmond, VA
Posts: 381
Likes: 0
Received 2 Likes on 2 Posts
Knock Control Questions (2.4L w/ Phantom Knock Issues)

Looking for someone to explain the intricacies of the EcuFlash/Tephra "Knock Control" maps before I start randomly adjusting things to see what happens.

My Tephra Rom has..

Knock Load (knock ignored below these values)
Knock Sum Decay Timer
Knock Multiplier Low Rpm
Knock Multiplier Mid Rpm
Knock Multiplier High Rpm
Knock Adder Single Gain
Knock Adder Triple Gain


Qu:
  • Knock Load is self explanatory.
  • I don't imagine I'll need to play with the decay timer.
  • When I raise the multiplier like others have suggested, what is happening?
  • I'm not sure what the Adder's do? Or when its a single or triple event?

Also, if I'm trying to tune the knock sensor.. is there any clues if i break down the knock sum and log all the knock values where i can start to see the difference between the phantom knock and the real knock. Obviously I'm worried that if i de-sensitize the knock sensor too much I'll never see the real knock.

Hints, suggestions, advice, links, quotes, game plan...?



Why I'm asking...
So I have a 2.4L, 10:1 compression motor, with lots of goodies running a stock 9.8 turbo that was supposed to be temporary but.. I've got about 40 miles on the motor in the past four or so years and the car keeps getting put on the back burner because of this phantom knock issue. Its during normal driving, WOT pulls, some times its hardly there and the next run its in the 23 knock sum a crossed the entire curve and it doesn't mater what i set the timing too. The car drives awesome, doesn't buck, doesn't fall on its face, plugs look good, the compression numbers would make you jealous, and not an issue to be found. Ya know other then not being able to tune the car to make any power. I've read that a bunch of 2.4L's have had similar issues which makes me feel better but still a pain.

I replaced the knock sensor a while back, and tomorrow morning I'm gonna check one more time to ensure its still snug like it should be, just to make sure. I've looked at different things that could be causing noise but really I've just come to the conclusion its just a noisy motor and I'm gonna have to figure out a way to tune this knock sensor or simple bypass it all together and tune with some Det-Cans.

Last edited by Live4Redline; Jun 18, 2015 at 12:19 AM.
Old Jun 23, 2015, 09:10 PM
  #2  
Evolving Member
Thread Starter
iTrader: (2)
 
Live4Redline's Avatar
 
Join Date: Apr 2008
Location: Richmond, VA
Posts: 381
Likes: 0
Received 2 Likes on 2 Posts
Cause why not...


So I have been tuning the knock sensor on my car and just gonna post what i have so far to see anyone wants to make some comments. I've been tuning the knock sensor in combination to reading the spark plugs and even recording audio feed back off the knock sensor (Det-Can's).

Here's what I'm getting on the logs atm...




The Blue log is with Knock Multiplier settings of 19 / 19 / 31.
The Red log is with Knock Multiplier settings of 20 / 20 / 31.

I'm trying to keep the multiplier settings as low as possible but I'm gonna have to keep raising the high RPM setting till I can get some better results. I originally was getting 24-32 knock counts up top and with the setting at 31 I'm now seeing 17-19. By the time I tune it all out I doubt there will be much left to listen for any true knock but we will see.

I recorded some audio of a stock knock sensor in a non-stock location. This was a secondary sensor and not the ECU's sensor that was graphed in the logs. I plan to record the factories sensor tomorrow for comparisons.

This was a recording of me cruising through gears and what not.
http://www.filedropper.com/knockmap001cruise

This is a quick pull from 3K to redline.
http://www.filedropper.com/knockmap002pull


Conclusion so far... The spark plugs look perfect, the car even when its pulling timing for knock is running great and pulling strong, the audio is very noisy but I don't hear anything I would think would be knock but its hard to tell. I do think that up top the noise level picks up a great deal and its understandable that the knock sum is failing so hard, I'm just not sure why. When i test the factory positioned knock sensor maybe it'll sound a little clearer. I question if maybe the valve-train is responsible for the noise in the higher RPM but not exactly sure how to test it or why.


Well that's what I have so far... notes, comments, suggestions, or questions welcomed.
Old Jun 25, 2015, 04:19 PM
  #3  
Evolving Member
Thread Starter
iTrader: (2)
 
Live4Redline's Avatar
 
Join Date: Apr 2008
Location: Richmond, VA
Posts: 381
Likes: 0
Received 2 Likes on 2 Posts
Alright so went out and tested my new Knock Listening Harness and it works. I took some unlogged audio recordings and well I'm probably gonna have to take some more to be sure.

This is the stock sensor in the stock position while its connected to the ECU.

A random cruising audio recording. Nothing out of the ordinary but it does sound better then the remote location sensor I used before.
http://www.filedropper.com/knockmap003cruisestock

A low rpm till bounce off redline pull. I slowly increased the throttle to redline in an effort to extend the length of the recording. In the top end I do hear some spikes which may or may not be knock. Its hard to tell as of yet but the timing is rather conservative and plugs are saying no issues so there shouldn't be any knock.
http://www.filedropper.com/knockmap004pullstock

I'm gonna try to add some resistance to the harness in an effort to clean up the background noise and clear up what I'm hearing in the top end.



The Knock Listening Harness I made...

A trip to the junk yard to get some knock sensor plugs. The V6 Galants, Eclipses and Diamante locate the sensor plugs on the left side of the motor in an easy to access location right there in plain view so that's what I grabbed. There a two wire plug with a shiny silver sticker with a large A or B on them.



Then I made the below tieing in a female RCA cable plug (on a quick disconnect plug) for audio out and the yellow ground wire.
  • I will note that if you plug this up to the knock sensor and do not plug the ECU harness up too it, you need the ground wire.
  • If you plug this up to the knock sensor AND the factory ECU Harness, you do NOT use the yellow ground wire (feed back issues).

I'm also using an RCA to HeadPhone jack adapter to get it too my PC and or Headphones. RCA is just easier to work with and extend from under the hood too the passenger seat.

RCA Wiring: Center peg is signal. Outer ring is ground (black wire).



I plan to tie in some resistance in an effort to clean up the sound a bit. If it improves the quality I will update the post with the changes.
Old Jun 25, 2015, 05:14 PM
  #4  
EvoM Guru
iTrader: (9)
 
Abacus's Avatar
 
Join Date: Jun 2012
Location: FL
Posts: 1,339
Received 355 Likes on 240 Posts
Nice work on that harness. Here are some of my suggestions and ideas.

Does the engine have any noise on start up?

What is the piston to wall clearance?

Really aggressive camshafts?

Was the head decked? If so was the chamber measured and compression ratio calculated?

spark plug gap near .020-.024
check physical timing

check ignition timing with a light

How does your ignition timing look for a 2.4. You have a 10:1 engine so it won't take as much as a 9:1 engine.

Do you have aftermarket motor mounts or anything thing hitting anywhere ( piping) The 2.4 block is a little taller so is the exhaust hitting anywhere under power?

If physical timing is good then why not go to a chassis dyno and bring some high octane fuel with you. You can look at the engine moving around and see if anything is hitting anywhere. You can add/reduce timing and make some small wastegate pulls and see if the graph is wavy and showing knock. You will also be able to physically hear audible knock if its real. If the graph is wavy and pulls timing you can numb the knock sensor and see if it cleans up .

You can also add 5 gallons of high octane fuel and if the knock is real it will be reduced.

I'd recommend running the AFR 11.1-11.4 on pump gas. Too rich and it can knock. 11.5-11.8afr on high octane.
Old Jun 26, 2015, 10:53 AM
  #5  
Evolving Member
Thread Starter
iTrader: (2)
 
Live4Redline's Avatar
 
Join Date: Apr 2008
Location: Richmond, VA
Posts: 381
Likes: 0
Received 2 Likes on 2 Posts
Originally Posted by Abacus
Here are some of my suggestions and ideas.
Does the engine have any noise on start up?
  • No noise at start up.

What is the piston to wall clearance?
  • Don't have the number off the top of my head, but stock'ish clearance.

Really aggressive camshafts?
  • Second hand HKS 272's, they were laying on the self when I put the motor together.

Was the head decked? If so was the chamber measured and compression ratio calculated?
  • Head was not decked but I did have a professional double check the measurements for the compression ratio for me.

spark plug gap near .020-.024
  • Yes, checked it twice.

check physical timing
  • Checked it on the motor stand, drove it around 40miles and tore the timing side down to check it again.

check ignition timing with a light
  • No, not yet.

How does your ignition timing look for a 2.4. You have a 10:1 engine so it won't take as much as a 9:1 engine.
  • This is what I'm using right now to test on. The load peaks at 310 @ 4250 RPM. I was originally pulling timing out of any where it knocked which ended up being the entire curve and map. No amount of lowered timing seemed to fix the issue. I was worried the low timing might have its own adverse effects so I went with this which I felt should be safely conservative. I'm open to suggestions how ever.



Do you have aftermarket motor mounts or anything thing hitting anywhere ( piping) The 2.4 block is a little taller so is the exhaust hitting anywhere under power?
  • Stock mounts and exhaust it clear. The transmission mount is the only mount that's not new. I may replace it just cause, but it doesn't show any signs of issues.

If physical timing is good then why not go to a chassis dyno and bring some high octane fuel with you. You can look at the engine moving around and see if anything is hitting anywhere. You can add/reduce timing and make some small wastegate pulls and see if the graph is wavy and showing knock. You will also be able to physically hear audible knock if its real. If the graph is wavy and pulls timing you can numb the knock sensor and see if it cleans up.
  • I've done wastegate street pulls and the knock is still there. Boost levels and ignition timing seem to not effect it. Unfortunately the dyno is not an option at the moment. I may be able to get it on there in a August at best. So just doing the best I can, while I can. I agree though, some race fuel and some time on the dyno would help to rule a couple of things out.

You can also add 5 gallons of high octane fuel and if the knock is real it will be reduced.
  • Can I dump a bunch of E85 in it? No I'm kidding, but I'm probably gonna grab some 110 just so I can say for sure.

I'd recommend running the AFR 11.1-11.4 on pump gas. Too rich and it can knock. 11.5-11.8afr on high octane.
  • At the moment I was shooting for 11 to 11.5 AFR but its dropped back down on me with some of the previous ignition changes and the knock sensor pulling max timing out every couple of cells. lol

Last edited by Live4Redline; Jun 26, 2015 at 10:57 AM.
Related Topics
Thread
Thread Starter
Forum
Replies
Last Post
S.Brewer
General Engine Management / Tuning Forum
6
Dec 13, 2014 10:05 AM
stealthMR
Evo X Engine Management / Tuning Forums
7
Jun 6, 2011 06:56 PM
nismo n9ne
Evo Engine / Turbo / Drivetrain
22
Feb 9, 2011 01:21 PM
erggy
General Engine Management / Tuning Forum
2
Aug 6, 2006 09:18 AM


Thread Tools
Search this Thread
Quick Reply: Knock Control Questions (2.4L w/ Phantom Knock Issues)



All times are GMT -7. The time now is 04:26 AM.