2006+ Honda Civic Forum banner

1 - 20 of 21 Posts

·
Registered
Joined
·
120 Posts
Discussion Starter #1
I removed the bonnet insulation a long time ago because I thought it was pretty useless. [smilie=hammerit.gi: After getting flashpro and getting more and more familiar with maps and datalogging I noticed I always had some knock counts in the datalogs .
Nothing major , typically I would get up to 10 knocks in a 30 min drive.
Most of them were on the second cylinder, occasionally on the third , never on the first or fourth. I tried to get rid of this by pulling ignition and adding fuel but this had no effect.
I recently refitted the noise insulation pad under the bonnet to see if this would have an effect on (ghost) knocks. I've only done one datalog so far but the knock count has completely disappeared. Seems this insulation pad is best left in place.
 

·
Premium Member
Joined
·
6,643 Posts
Might be coincidence. What mods you got?

Did you try playing with knock tables?

Hondata do say if you start fiddling with knocks dont expect to be able to picj them up accurately anymore.

Hondas Active knock control system isnt great, ive had many battles with it
 

·
Registered
Joined
·
120 Posts
Discussion Starter #3
The only engine mods are a custom map and a Seeker air intake duct. I tried decreasing the knock sensitivity a bit but that had little to no effect. Everything is back to standard now. As far as I know the microphone that monitors possible knock is located near the second cylinder. That probably explains why you get a higher knock count on the second cylinder and none on the first and fourth. Individual cylinder trim for the knock sensitivity (like you have for fuel and ignition) would be a possible solution perhaps ?
I tried a modified (better A/F) GroupN map but that also recorded about the same amount of knocks.
Could indeed be coincidence that the knock count has gone to zero after refitting the insulation , need some more datalogs to confirm this.
 

·
Premium Member
Joined
·
6,643 Posts
Is it retarding the ignition, if so by how much?
 

·
Registered
Joined
·
120 Posts
Discussion Starter #5
Is it retarding the ignition, if so by how much?
Maximum 5°. Even a fairly low knock count is enough to raise knock control so the ignition retard will be greater as well. It seems to take a long time for knock control to drop so if you have occasional knocks knock control stays high.
Just compared two logs (same length) , one with 9 knocks on cyl.2 and the other with no knocks. Getting rid of all the knocks does make a difference.
 

·
Premium Member
Joined
·
6,643 Posts
5 is quite a bit.

The reason why it is doing it is it is listening for a certain frequency that indicates a knock is about to occur it then pulls the ignition.

I had the same issue, mine was caused by my map at the time.

what octane rating are you using? the higher the more likely combustion is to occur when top dead centre.

Id try to establish if knocking is actually occurring first. TDI south use a stethescope on the intake manifold and filter the frequencies to listen for knocking.

There are two parts to the knock system. you need to change the active knock control tables which are responsible for retarding the ignition. these tables are only visible with show advanced options turned on if i remeber correctly.
 

·
Registered
Joined
·
120 Posts
Discussion Starter #7
Always use 98 RON , that's the highest we can get over here.
If you would get a similar (and high ) knock count on all cylinders then your ignition map will have to be retarded in the areas where knock occurs.
But our car seems to have a system that's clearly more sensitive to one particular cylinder .
 

·
Premium Member
Joined
·
6,643 Posts
Always use 98 RON , that's the highest we can get over here.
If you would get a similar (and high ) knock count on all cylinders then your ignition map will have to be retarded in the areas where knock occurs.
But our car seems to have a system that's clearly more sensitive to one particular cylinder .
Not really sure what you mean by "your ignition map will have to be retarded in the areas where knock occurs", doesnt make sense.

Ive not noticed any particular cylinder detecting "knocks" more than others.

My ignition was being retarded by 7 degrees without any knocks being counted.

Have you had a valve clearance check done?
 

·
Registered
Joined
·
120 Posts
Discussion Starter #9
Not really sure what you mean by "your ignition map will have to be retarded in the areas where knock occurs", doesnt make sense.

Ive not noticed any particular cylinder detecting "knocks" more than others.

My ignition was being retarded by 7 degrees without any knocks being counted.

Have you had a valve clearance check done?
If you get knock counts at a particular point (RPM-MAP-CAM angle) the ignition value for that point is probably too advanced.

I never get knock counts on cylinder 1 or 4 , only on cylinder 2 (very rarely on 3). The total count is low and the uneven spread of the knocks across the cylinders makes me seriously doubt it is actual knock.

Ignition retard is triggered by knock control and ignition limit , not directly by knock counts. Knock control is affected by knock counts and knock level. Knock control seems to take a very long time to drop , even without knock counts. Knock control is a calculation for the fuel octane. At 70% knock control the ECU thinks it is getting 93 RON (0%knock control = 100RON / 100%knock control = 90RON) . I think knock level is the noise level , if the knock level (again at a certain RPM-MAP-CAM point) exceeds the knock sensitivity value for that point it is recorded as a knock count. If the knock level is high but not high enough to record any knocks I suppose it will still keep knock control high which has an effect on ignition retard.

Valve clearance was checked last summer.
 

·
Premium Member
Joined
·
577 Posts
Hondas Active knock control system isnt great, ive had many battles with it
I have noticed tons of false (ghost) knocks after new cams and my tuner said it's normal and should not worry about it... Almost impossible to set...
 

·
Premium Member
Joined
·
6,643 Posts
I have noticed tons of false (ghost) knocks after new cams and my tuner said it's normal and should not worry about it... Almost impossible to set...
same here
 

·
Registered
Joined
·
120 Posts
Discussion Starter #13
The standard knock sensitivity tables have lower values (more sensitive) near the stock VTec point (5400RPM). Using these stock sensitivity tables on an engine with a lowered VTec point probably makes the detection system less accurate. Decreasing the sensitivity around the stock VTec point and increasing the sensitivity near the new lowered VTec point is maybe the way to go ?
 

·
Premium Member
Joined
·
6,643 Posts
The standard knock sensitivity tables have lower values (more sensitive) near the stock VTec point (5400RPM). Using these stock sensitivity tables on an engine with a lowered VTec point probably makes the detection system less accurate. Decreasing the sensitivity around the stock VTec point and increasing the sensitivity near the new lowered VTec point is maybe the way to go ?
But dont expect to pick up knock correctly in the future.

My cams are noiser hence it picking up false knocks
 

·
Registered
Joined
·
120 Posts
Discussion Starter #15
But dont expect to pick up knock correctly in the future.

My cams are noiser hence it picking up false knocks
I would use the lower stock sensitivity values at the new VTec point which would make the system more accurate. I wouldn't just start adding or removing sensitivity values randomly.
 

·
Premium Member
Joined
·
1,368 Posts
same here
mate this is strange for me because i have the same cams as You (FD2) plus SC which makes a lot of noice and i have after 30 min of logs 4-8 knoks. 80% of these knoks are on cilinder #2 which is typical

when i had cams and my profi tuner in Poland made for me map i had about 3000 knoks and i also heard that these knoks are foulse knoks. Proplem was that my map was garbage! With good map You should not have more that 10 knoks after 30 min log

BTW ASH very helpfull mod to lower knoks is change sparks plugs to one step colder (from 7 to 8)
 

·
Premium Member
Joined
·
1,368 Posts
You should read this

One of the hardest things to do in your calibration is to eliminate knock counts. Sometimes its just false knock on cylinder #2, sometimes its a valid.
When trying to eliminate knock count, theres a couple things to ask yourself in this regard:
1) With reference to the engine bay, is everything installed correctly, and secured tightly?
2) Do I have anything hitting the block during driving, like an intake, or shifter related mechanism?
3) Is there an Ingalls Torque Dampener or similar modification like solid engine mounts increasing the noise the knock sensor is hearing?
4) Is my suspension, sway bars and linkages setup appropriately?
5) Is my valve lash setup correctly? (Neccessary after installing aftermarket cams).
After you have eliminated the obvious causes of false knock, what remains is most likely valid. The knock itself is typically caused by lean conditions combined with too much spark advance. The affect is amplified by low quality gasoline.
6) Try switching to a 93+ Octane gasoline (Ethanol increases octane as well) and see if this helps. Typically it should help even on the loudest setups.
Recently on my personal car, I was having issues with 1000's of knock counts as logged when tuning with the Hondata Flash Pro. They would only occur under certain situations, and would not go away, even when the timing was pulled in those affected areas.
Its worth noting that the IPS K2 cams I was running greatly increase cylinder pressure due to their efficiency and therefore require less timing then stock in numerous areas, so I while I was focus'd on this I overlooked the fact that I was running stock spark plugs. Typically NA setups do not require changing plugs, to a colder plug, as this is usually reserved for Forced Inductions setups.
7) However, being unable to correct the knock counts and having to reduce timing so drastically that I felt it was unacceptable as I was losing significant power, I finally decided to try the upgraded to BkR8EIX Iridium plugs, over the 7 series stock platinum plugs.
These plugs are here: NGK Iridium IX® Spark Plugs
The change was near instant, before:

after:



As you can see by the reduced knock count the plugs themselves make a large difference, and the increased dynmaic compression caused by the IPS K2's along with the increased redline, directly affected the ability to the stock 7 series plugs to shed heat.
The old plugs can be seen below, if you look close you can actually see Cyl #2 and #4 are running richer then the rest, the affect is more obvious in person, I think this is due to the knock sensor "hearing" those cylinders more due to sensor placement, and therefore compensating more often. All cylinders are running hotter due to the cams.

8 ) Another key in eliminating knock is to track the knock counts and see WHERE they are increasing. Typically you would monitor Knock Count in your datalogs in Flash Pro, and find the instants where the knock counts jump up significantly. In these areas you will have to pay attention to the ignition values, and make sure the surrounding values and changes are done smoothly, and that the ignition being run in those areas is typical.
9) If the results are repeatable, and it is always knocking in certain portions of the ignition map, try and remove 1-2 degrees of timing, and ensure the changes are smoothed to the surrounding squares.
10) Check the the AFR being run is typical for either closed loop or wide open throttle.
11) Lastly, if all else fails. And you have checked your spark plugs and verified the knock counts are indeed valid. See if they are only valid on one cylinder. If so... you have a couple options. You can use the cylinder trims to adjust fueling and decreasing timing on that cylinder and this will mask the symptoms temporarily. Or, can can go after the root cause. This could be compression related on that cylinder, or a bad coil above the spark plug, or even a clogged injector.
All it takes in one piece of debris sitting at the mouth of that injector to reduce the flow drastically. You may have to take them off and inspect.
I am sharing this information for those that are having issues tracking down knock counts in their calibrations.
 

·
Premium Member
Joined
·
6,643 Posts
BTW ASH very helpfull mod to lower knoks is change sparks plugs to one step colder (from 7 to 8)
Thanks for your very informative post .

Ill read it properly tonight
 
1 - 20 of 21 Posts
Top