Poll: Do you desensitize you knock sensors?

[markislive]

Ok, I’ve always left my knock sensors alone to be on the safe side…I can only manage 22* of timing up top in my 5,000lb sled and not trigger KR (.85 WOT Lambda). Just curious what everyone else is doing?

[Nick 2K2 SS]

If you are not tuning on a dyno so you can see if you are actually making more torque by increasing the timing, then no I would not desensitize the knock sensors. If you don't know where you are at torque wise with the timing then you are just shooting in the dark and I would play it safe and leave the knock sensors as they are. If it is true knock that is occuring then you might have actually already overshot the MBT timing in those ranges.

[markislive]

If you are not tuning on a dyno so you can see if you are actually making more torque by increasing the timing, then no I would not desensitize the knock sensors. If you don't know where you are at torque wise with the timing then you are just shooting in the dark and I would play it safe and leave the knock sensors as they are. If it is true knock that is occuring then you might have actually already overshot the MBT timing in those ranges.

Makes plenty sense...Thanks

[LSX378]

Yes but only at lower rpm/airflow values when running forged pistons.

[icicleboy]

this brings up an interesting point which I've pondered. I'm sure my truck has some false knock. I've read that numerous tuners start with 10% out of the sensitivity (raise numbers in the tables 10%) to make them less sensitive. If you think about it, a GM calibrator might pour in 100 octane, take the truck out and floor it. They look at the voltage throughout the pull and set that (or somewhere about there) as the threshold by rpm, etc.

Now you add FI. There is way more air at that rpm then there was from the factory. This makes the cyl pressure higher, hp higher, and possibly the noise the motor makes at that rpm louder. Now the threshold might need to be increased just to keep it from falsly reading something which is normal engine noise as knock.

Just a thought.

[BBA]

I don't desensitize the sensors unless working with something that has a noisey cam and headers.

I do, however increase the KR removal rate.

[CAMAROZ28AQUINO]

I have like 6* of knock on my Z and i want to decensitize

[Snyper]

a GM calibrator might pour in 100 octane

....why would they do that?

[S2H]

I don't desensitize the sensors unless working with something that has a noisey cam and headers.

I do, however increase the KR removal rate.

noisy cams dont cause knock... detonation causes knock...random noises arent picked upand dont make the sounds that the knock sensors are looking for...and you can just have too much timing and not know it because everybody seems to think that our engines like lots of timing...which is not always true.


find the real cause of your knock issues...could be a transient fueling issue causing your knock(a very likely cause that a lot of people overlook



I personally leave the knock sensor setting alone..
the only thing I change if I change it at all is the knock recovery time...make it quicker sometimes

[S2H]

I have like 6* of knock on my Z and i want to decensitize

find the real issue...that baby cam might not want more than 27* at WOT and only 21~23 at peak torque...

and as I just mentioned... get transient fueling correct and some of that will go away..
compression ratio? gas? 89/91/93?
pcv sytem putting a lot of oil in the engine?
have an injector spraying leaner than the others?

[LSX378]

noisy cams dont cause knock... detonation causes knock...random noises arent picked upand dont make the sounds that the knock sensors are looking for...and you can just have too much timing and not know it because everybody seems to think that our engines like lots of timing...which is not always true.


find the real cause of your knock issues...could be a transient fueling issue causing your knock(a very likely cause that a lot of people overlook



I personally leave the knock sensor setting alone..
the only thing I change if I change it at all is the knock recovery time...make it quicker sometimes

True on the cams largely dont cause knock other than effect on cyl pressure but to be most effective the knock detection function needs tuning for a specific engine and accessories in a specific chassis. This helps establish what GM calls the "noise channel" where knock is applicable and outside it "background" noise.

If a different valvetrain setup or forged pistons or a ring prone header setup or different powertrain vibrational characteristics add or detract or move the "background" noise both sides (low level noise and high level noise) of the "noise channel" then the knock area needs work to ensure it is working on real knoc.

Its not at all easy which is why GM typically uses 150+ data points per cylinder (1200 per V8 engine and only applicable around TDC on the firing stroke so a pretty narrow window) to get it in the right space then applies at least a couple different multipliers.



Interestingly GM almost Never varies learn adjustment minimum and initial voltage.

[Dragman]

I have like 6* of knock on my Z and i want to decensitize

I think u should find out if it real or false before doing that!

[TiredGXP]

I see up to 6* of KR when DOD/AFM deactivates, so knock sensors are picking up some valvetrain noise.

I dealt with this primarily through increasing the recovery rate, rather than desensitizing the knock sensors

[Alvin]

A plug check will tell you right away if you have any deto at all. I had a 98 Z in last week that would only take 22 degrees of timing. The 98's are bad about knock sensors going nuts, so installed fresh plug and ran it at 24 degrees and the plug showed light knock.

[icicleboy]

Update :

I just took a knockbox and bolted the sensor (mic) directly next to a knock sensor on my L92. I would say that 99.5% of the time that I would see knock on the screen (HPT scanner), I would hear it in my ear. This pretty much confirmed for me that the sensors are well calibrated and my particular vehicle has no false knock. Playing with the attack/recovery is fine, but I'm leaving the sensitivity stock.

[Blackdevil77]

an easy way to find out if knock is false or not is to put some race gas or octane booster in your gas. If you still get knock, it's probably false, if it goes away or noticeably decreases, then the knock was probably real. I had to desensitize mine a bit with my cam, valvetrain noise increased a lot which triggered some false knock

[bluegoat06]

I see up to 6* of KR when DOD/AFM deactivates, so knock sensors are picking up some valvetrain noise.

I dealt with this primarily through increasing the recovery rate, rather than desensitizing the knock sensors

I had the oppposite problem, i'll see 4-5* when AFM was engaging. I tried messing with the knock sensors in the low rpm mid/high g/cyl areas while in DOD, because i was getting that much with the stock G8GT timing tables. i started lowering the timing little by little until i saw a decrease in KR. loaded the stock knock sensor settings (only modified cyl#2 and 7 under the impression that they were nosiy), and simply continued lowering the DOD timing until the crazy KR went away. i get .4* here and there, but i can live with that. they were not crazy like i thought, the factory spark tables have too much advance in it, in both V8 and DOD modes. at least in my car, running with 91 octane. i didn't have this problem when i was living in Ohio and pumping 93. Cali gas Sucks

And the first thing i always do, i changed the recovery/attack knock settings, it helps in the GTO's, and it helped the G8 when i was in Ohio running 93 octane, and by then i even had 2* more across the board. but in Cali with 91 octane, things changed, i was getting knock out of the burst settings as well while in DOD. took me 4 days to say screw the factory table, i need to lower the timing, the car is even responding well at the track with less timing, so the knock i was having in the logs was real. the car doesn't knock with 93, but it does with 91, crazy, and i think the manual says we can run 87 or 89 i don't remember. we have a guy in our G8GT forum who runs 87 octane with the factory timing, and he's getting up to 14* knock, not even in DOD. somebody at GM was drinking a lot when they approved our timing tables.

[ssdungeon]

How much do you usualy change in the tabe?
mult by 1.04?

I personally leave the knock sensor setting alone..
the only thing I change if I change it at all is the knock recovery time...make it quicker sometimes[/QUOTE]