A question for Hondata regarding knock

K-Series Programmable ECU installation questions / support issues
Locked
fonz600
Posts: 239
Joined: Sun Jul 30, 2006 11:57 am

A question for Hondata regarding knock

Post by fonz600 » Thu Jun 05, 2008 9:31 am

I was wondering if someone from Hondata can answer this question for me?

I have been plagued with ghost knock in my JRSC JDM EP3 (9PSI) since I bought it.
I've tried everything from removing up to 12 degrees of ignition to changing the manifold, exhaust, intake, charger belt and tensioner and spark plugs in an effort to eliminate any kind of noise which may be triggering it.
I know it's ghost as I can get over 100 counts from 6k-8k in 1 WOT run but with no hesitation whatsoever.

Now I was always pulling ignition in a belief it'd reduce knock but someone on another forum suggested adding ignition to columns 11-13 as it'll reduce ghost knock.

He was right and after checking previous datalogs the more ignition I pulled the more knock I got.
Thing is I don't wanna advance too far and get genuine knock but with all the ghost knock it's hard to tune for it.

A long story for a simple question but why does advancing the ignition in columns 11-13 reduce my ghost knock and pulling ignition increase it?

blundar
Posts: 16
Joined: Thu Feb 10, 2005 12:10 pm
Location: cincinnati
Contact:

Post by blundar » Tue Jun 17, 2008 8:29 am

I had a similar issue with a 15psi JRSC K20A2 I did last week. I'm betting that misfires are showing up as knock. I had several spots where I am convinced the knock sensor was lying. I would turn on watermeth and then pull 4 degrees of timing and see the knock counts increase. Pulled plugs and they looked perfect, even under magnification. Obviously, there wasn't knock happening at either timing level. Added timing back in, knock counts diminished.

User avatar
Spunkster
Site Admin
Posts: 18059
Joined: Fri Sep 20, 2002 5:06 pm
Location: Hondata
Contact:

Post by Spunkster » Tue Jun 17, 2008 1:48 pm

We have seen this happen with superchargers, and it is one of 2 things, either a resonance caused by a change in the cumbustion process with too little ignition timing that then casues some mechanical noise in the supercharger, or it just makes a noise which is in the frequency range of the knock sensor. This also happens in engines with very aggressive cams and can be cased by valve train noise. It is also possible that this is what is happening with the supercharger as well. When this issue arrises it is a good idea to have a good aftermarket knock monitor that you can actually listen for knock.

fonz600
Posts: 239
Joined: Sun Jul 30, 2006 11:57 am

Post by fonz600 » Tue Jun 17, 2008 3:20 pm

Yeah at the dyno last week I used Det cans to listen to my engine while we tuned. Although Kpro was showing over 100 counts of knock in a run myself and the dyno operator couldn't hear any sound of knock with the Det cans.
I have a new knock sensor on order but not holding my breathe it's gonna make any difference.

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Tue Jun 17, 2008 11:24 pm

With my jrsc rsx, I can hear knock down low, even with a vtec point of 3k (which really helps detonation down low). I've verified it by hooking my knock sensor up to a listening device and it is most certainly detonation. The best course of action I've found is to really richen up the boost columns on the low cam and try to stay away from them. Your gas pedal is your friend in this case.

Wayne
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

carbonctr
Posts: 128
Joined: Thu Feb 21, 2008 1:34 am
Contact:

Post by carbonctr » Wed Jun 18, 2008 1:23 am

I filled my car with the usual Shell V-Power and Dyno-Tab octane booster combination yesterday. I did a usual "quick" 40 mile drive last night and the knock count was horrendous! I achieved about 900 knock on the drive home as opposed to the usual 100 odd for the same drive. I wonder if the UK fuel problems are resulting in crappy V-Power...?

I'll post a datalog when I get my usb drive back from my girlfriend so I can download it from my carpc.

Also, I experienced the same problem previously when reducing the ignition in the boost columns...it increased the ECU's knock count...

Can anyone suggest a decent knock sensor at a sensible price?

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Wed Jun 18, 2008 3:56 pm

carbonctr wrote:I filled my car with the usual Shell V-Power and Dyno-Tab octane booster combination yesterday. I did a usual "quick" 40 mile drive last night and the knock count was horrendous! I achieved about 900 knock on the drive home as opposed to the usual 100 odd for the same drive. I wonder if the UK fuel problems are resulting in crappy V-Power...?

I'll post a datalog when I get my usb drive back from my girlfriend so I can download it from my carpc.

Also, I experienced the same problem previously when reducing the ignition in the boost columns...it increased the ECU's knock count...

Can anyone suggest a decent knock sensor at a sensible price?
Are you using a short ram intake?
The oem knock sensor is pretty decent.
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

carbonctr
Posts: 128
Joined: Thu Feb 21, 2008 1:34 am
Contact:

Post by carbonctr » Thu Jun 19, 2008 1:07 am

Razathorn wrote: Are you using a short ram intake?
The oem knock sensor is pretty decent.
I have the GruppeM Ram Air System with an SFS Silicon hose between it and the throttle body. The only modification I made before the knock increased was shortening the part on the hose where the IAT sensor sits so it is further in to the intake hose. I don't think temperature compensation is used so don't think this will make a difference...

Is the OEM knock sensor not too sensitive?

fonz600
Posts: 239
Joined: Sun Jul 30, 2006 11:57 am

Post by fonz600 » Thu Jun 19, 2008 3:20 pm

Exactly where is the knock sensor located Razathorn? Should have my new one tomorrow so gonna swap it over weekend.

Locked