Pontiac Solstice Forum banner
1 - 15 of 15 Posts

·
Premium Member
Joined
·
389 Posts
Discussion Starter · #1 ·
So...the knock sensors that GM use are extremely sensitive, and have a tendency to interpret external noises as knock/ping and as a result dial back timing which, of course, negatively effects performance.

There is currently no way to adjust gain or sensitivity using HP Tuners. You do have the ability to turn them off completely, but that can be dangerous.

Laying in bed last night, I was thinking if there would be a way to 'tap in' an external electronic device that would allow you to adjust sensitivity? Some type of potentiometer or ???? (I really have no idea what a potentiometer is! I just like using that word.)

Seems simple enough to me, but I wouldn't know where to start. Maybe TS has some idea if something like this would be possible.

What you all think?
 

·
Registered
Joined
·
143 Posts
why do you want to adjust them ? what I did not understand . I use 93 plus octane no spark pings . leave them alone if you buy bad fuel just once and you relocate the knock sensors it will not adjust the spark timing correctly. my tune does allow for spark timing adjustment . you can kill your engine . my opinion only
 

·
Premium Member
Joined
·
389 Posts
Discussion Starter · #4 ·
why do you want to adjust them ? what I did not understand . I use 93 plus octane no spark pings . leave them alone if you buy bad fuel just once and you relocate the knock sensors it will not adjust the spark timing correctly. my tune does allow for spark timing adjustment . you can kill your engine . my opinion only
I guess I didn't make myself clear. I don't want to turn them off, I just want to be able to make them LESS SENSITIVE TO DRIVELINE, EXHAUST, ENGINE NOISE, RATTLES that are producing false or phantom knock retard.

If you have ever tuned or captured logs of these LNF engines, it will become immediately obvious that false knock is a real problem and unlike other ECM's there is no provision to adjust sensor sensitivity.
 

·
Registered
Joined
·
3,882 Posts
You might want to check this site out:

LNF False Knock- Discovery might lead to cure???

I think Shabby is the guy to talk to.
I saw no difference rotating them, i basically gave up. I saw "knock" after picking up the car and still see "knock" now, i autocross the car every summer and run 22lbs or so on 94 octane, engine hasn't blown up yet so i doubt its real knock.
 

·
Registered
Joined
·
7,505 Posts
Knock sensors are (usually) piezoelectric devices that generate a signal based on, effectively, noise. Kinda like a microphone, only not as linear. So a small noise will generate a small signal, but a larger noise over some threshold, generates a very much larger signal.

You might be able to attenuate (lower) the signal with a potentiometer (=variable resistor :)) but I wouldn't think that would work well, due to the non-linear response of the sensor.

Bottom line, there's not a lot I could recommend that you do.
 

·
Premium Member
Joined
·
389 Posts
Discussion Starter · #7 ·
Thanks TS. That was the technically relevant response I was looking for. I was, however, looking for a more productive answer:). I guess I'm stuck with this problem, and will continue to look for the cause of the false knock.

I just did a run using 114 Octane race fuel. The results are exactly the same as when running 94. There is no doubt in my mind this is false knock. Timing is super conservative and I am running super rich. Damn sensors!!!
 

·
Registered
Joined
·
750 Posts
Maybe something could be attached to the sensor to muffle the sounds it picks up. When I had my turbo setup we had to remove the knock sensor on the dyno because it was getting false knock. We did notice that we had more false knock on the dyno than on the street. Probably due to the way the car is strapped down.
 

·
Premium Member
Joined
·
389 Posts
Discussion Starter · #9 ·
After reading several dozen articles, studies, and watching various videos; I have come to the conclusion that knock detection (regardless of make & model) is a fine electronic art. It appears that various experts throughout the industry all agree - the only way to 100% detect knock is to listen for it via a listening device attached to the engine block, a set of high quality headphones, and some type of specifically engineered electronics that both filter and amplify engine noise and knock.

To get an idea on how complicated and involved knock detection is, take a look at the below link. This is WAY beyond me.....

http://oatao.univ-toulouse.fr/154/1/Vincent_154.pdf
 

·
Registered
Joined
·
398 Posts
On our cars you can hear real knock with your ears. have you checked with John?
 

·
Registered
Joined
·
1,158 Posts
On our cars you can hear real knock with your ears. have you checked with John?
I agree with this, BUT, long term PREdetonation is going to do damage to the engine too. I don't think you can hear predetonation, but this is where you need to know the limits of the tune/engine combo. A forged piston will take this predetonation better than the hyper pistons that the motor comes with stock.

That said, every GM tune I have done aside from the LNF has a way to change the sensitivity of the knock sensors. I really wish this was something HPT could work on for us, but with the E69 being such a limited use ECU, I have my doubts (I am pretty sure they are done messing with this ECU).
 

·
Registered
Joined
·
7,664 Posts
herb, I run an after market set up on my vintage race car, that utilizes a stock GM knock sensor. It hooks up to a gauge with an adjustable threshold. you run at steady speed and tune out the level of random noise you always get from an engine, until it disappears and then you watch for any future readings above that noise floor, as it probably indicates preignition or detonation.

I mention this not because you'd have any interest in converting an automatically monitored system to a manually controlled one, but rather to indicate that there must be a way to adjust sensitivity for the sensors.

See http://www.summitracing.com/int/parts/msd-8964/overview/

I use it in combination with a manually controlled advance/retard control in the cockpit.
 

·
Premium Member
Joined
·
389 Posts
Discussion Starter · #13 ·
herb, I run an after market set up on my vintage race car, that utilizes a stock GM knock sensor. It hooks up to a gauge with an adjustable threshold. you run at steady speed and tune out the level of random noise you always get from an engine, until it disappears and then you watch for any future readings above that noise floor, as it probably indicates preignition or detonation.

I mention this not because you'd have any interest in converting an automatically monitored system to a manually controlled one, but rather to indicate that there must be a way to adjust sensitivity for the sensors.

See http://www.summitracing.com/int/parts/msd-8964/overview/

I use it in combination with a manually controlled advance/retard control in the cockpit.
Right on Bill. That is some cool stuff brother. Nice use of available technology. John and I are working through the "false knock" phenomenon. He is without doubt, the most competent LNF tuner in existence. That includes anywhere within this solar system, and I am guessing within this galaxy and quadrant of the universe! - probably all the way out to Vega!...But seriously, his mind works a little differently that perhaps the average tuner dealing with this very difficult DI engine. To give you an example (and I hope I am not giving away any trade secrets or proprietary information), when most see KR, they would instantly try and retard timing and richen up the fuel. He has gone the other way with my engine - He has advanced the timing and leaned her out! Logs are showing positive results with KR. Trying to do this with stock pistons could be hazardous, but with forged slugs, his theories and application of them make perfect sense. I'll leave it at that for now until the final product has been produced. Sometimes you just have to look at things differently to get the desired results.

When are you coming to the Island next?
 

·
Registered
Joined
·
7,664 Posts
When are you coming to the Island next?
My only short trips this year will be to Kamloops in my 3.4 V6 powered 1956 rebodied MG and to the Kappa event this summer in Oregon. I'd love to see your car once it is all finished some time!!
 

·
Premium Member
Joined
·
961 Posts
I have heard of guys using g orings behind the knock sensor, therefore dampening the knock sensors ability .

But don't know enough about it to offer any advice
 
1 - 15 of 15 Posts
Top