RedHardSupra

iTrader: (1)

I get knock in the entire range of map/rpm. i have conservative timing tables (22degs@WOT). I also have a small cam and dual springs, which might be making noise. The whole car is on fairly stiff suspension, so just about everything else can be making noise, as it's pretty much a rattle box when just driving around.
Yesterday I had about half a talk of gas (93oct) so I went out and put another 8.6gallons of 100octane (76 brand 'competition') to verify how real this knock is.
for the first few mins there was no knock, but then it went back to it's normal, randomly knocking self.
That means it's false knock, and not 'too much timing' related, correct?

So what are the usual suspects as far as false knock is concerned? I don't have pulleys, stock manifolds, and i just put on LS6 intake (knock occured before that too, so it's not that). other than that it's rather mild setup.

any ideas?

God Forgives I Dont

Here is a few causes that I know of:

Exhaust hitting the car or driveshaft (I know from personal experience)
Really noisy valvetrain setup
Bad transmition or motor mounts
Lean
Too much timing
Hot intake air?

Other than those I'm not sure what causes it. I'm working on getting rid of my KR now. So far a good portion of my KR went away when I took my car back to the mufler shop and had them work on my Y pipe and move my Cat away from the driveshaft. I guess my next step is to back the timing off a bit in the area's i'm getting KR.
Good luck and when you get it figured out let me know what it was please! CYA

RedHardSupra

iTrader: (1)

trans mount is <1yr old, engine mounts are probably original.
conservative timing, VE/MAF tuned well, and it's <45F out there, so it's not hot air.
sounds mechanical, doesn't it? so what do i do now? start running around my engine with a stethoscope like a mad scientist?

God Forgives I Dont

LOL I don't know enough to help you out. Hell I've picked up a good bit of my current tuning methods from you and other here so I'm sure you know more than me as it is.

Just out of curiosity you should put your stock tune back on for long enough to get a new scan from it and see if the KR is exactly the same or better/worse. Not sure exactly what that will tell you but might rule out the tune all together if it is the same KR on the stocker. Just a thought.

RedHardSupra

iTrader: (1)

few graphs of LTFS(avg) and Retard(max)

all except the graph marked '93octane' are on half tank of 93/100octane

does this look like false knock?

LTFT's



retard

white2001s10

iTrader: (4)

Try adding 10% across the board on fuel and see what effect it has on the knock.

When you test it with the engine dead cold, do you still get knock? more/less?

Are you testing under boost? and have you tried testing with less boost?

RedHardSupra

iTrader: (1)

unfortunatelly, no boost
i don't know about how it acts when it's totally cold, i tend to watch it live, but not save logs, so i don't know about that one.
for the highway runs i even pulled another 1deg timing live across the board, and it didn't change squat.

God Forgives I Dont

Hmmm... Thats a lot KR spread out all over. It looks like its not real spikey but more so just starts and then kind of stays for a longer legnth of time. Mine always spikes up and then goes away fairly quickly. Don't have a clue for you man.

But your LTFT looks good to go!

white2001s10

iTrader: (4)

It does sound a lot like false knock given the evidence so far, but I'd test a little more first.
If your AFR is where it should be, and you're at zero boost with only 22* timing advance, then I don't see how it's possible unless caused by a hotspot in the chamber.
Are you going to end up disabling the knock sensor anyway?

RedHardSupra

iTrader: (1)

i'd prefer not to disable the knock sensor, just in case of actual knock showing up. I am testing for max power/2-3hrs at full blast at HPDE events, not just some 12sec drag blasts. I need it to be reliable when temperatures climb and hold, a perfectly operational KS would be helpful.

jimmyblue

iTrader: (11)

There are sensitivity params you can play with in
HPTuners, I saw. Don't know about their details.

You may want to go lower the attack and raise the
decay rate so that the actual knock events become
more clear. Because the KR is very persistent you may
be seeing retard in cells that actually have no knock
events themselves, but are getting smeared with the
residual retard tail. If you really have bad ping this
would make the protection a bit less effective but I
have been running around with 9.xx on the fast attack
rate and (I think) 0.2 for the decays. This means a blip
will bleed off in about 5 frames and won't be putting
any instant 4 degrees on you either. Makes the true
ping detects a lot clearer.

Once you've tuned it up you might want to go a little
conservative for the street/track but I run around this
way full time. I'm not boosted or anything, though.

RedHardSupra

iTrader: (1)

Jimmy:
are you talking about engine->spark retard->base retard vs rpm and recovery rate vs rpm?

for attack i get 24-26-28-32-34-44 sort of values
for recovery i got 0.015-0.016-0.013 (these are going from low to high rpm)

is that normal or am i seriously off? anyone cares to post their values?

RedHardSupra

iTrader: (1)

ok, i checked out a bunch of different bins from horist's site and they all look the same (except 98 fbody) so what's screwy with mine?

jimmyblue

iTrader: (11)

Not screwy, that's just the way GM set it up. Cover it up,
hard and fast, and who cares if it takes 10 seconds
to go away.

Where GM put base retard 26-44 degrees (this used to be
done by a "fast attack rate" but HPTuners decided later
that it was supposed to be called something else and be in
different units) I have 9.011 across the board. This means
my spark retard value jumps 1/3-1/4 the abrupt amount
per ping-detect. Like where you guys see bang!, 4 degrees
I will see it ratchet up 1, 1.5, 2 ... if I'm in persistent ping.
But one little sneeze on upshift doesn't slam me to -4
right off.

Where GM put 0.015-0.013 in the recovery table I currently
have 0.25 making my KR "go away" about 15X faster.

So where you'd be retarded 4 degrees and maybe miss all the
subsequent ping detects, but also carry 4, 3.9, 3.8, 3.7.... slow
fade across a few thousand RPM span, I see little KR blips only
close-in to the actual point of the problem. Then I go and step
on them with spark retard or VE bumpage or whatever seems
to be most pertinent to the evident cause.

Of course some of it -could- be false knock as in sensors picking
up something besides detonation. But the other "false" is really
the bogus persistence of retard, not the knock detect itself.

God Forgives I Dont

thats some interesting info there Jimmy! I'm going to play around with doing that on mine as well and see If it works for my setup. Thanks 4 Info!

RedHardSupra

iTrader: (1)

i did a quick scan on the way home (not much rpms, engine barely warm) with jimmy's settings, and i still get knock (and i'm still on that 93/100 mix). any particular patterns i should look for in the logs once i get up close and personal with the logs? are there any extra pids i should scan for?

i'm starting to think that the only way to go now is to go through all retard tables, and zero them out one by one, and see which one makes the retard disappear, then at least we know what it acts upon. does that sound right? what if the knock is a combination of things? ugh, i don't even wanna think about the complications...
any ideas?

jimmyblue

iTrader: (11)

Here is a screen shot of the settings I was talking about

http://home.cfl.rr.com/jimmyblue/KR_...R_settings.gif

You may have to save it & view in an image editor, it seems
to display kind of poorly in MSIE but it's a clean image.

You might want to check the misfire-current counts just to be
sure you don't have a single cylinder lean problem (like one
injector being weaker than the rest). False KR should persist
even if you over-fuel the motor, I think false knock detects
will tend to be at certain resonances / RPM and if you see
blips at (say) 4000RPM while your spark and fuel remain on
an even keel then suspect something in the valvetrain, an
exhaust buzzing sheet metal, etc. ringing the head. KR at
tip-in and gear changes indicates the speed density side
being too lean, at the point where the blip appears; an O2
"hole" preceding the blip points to that.

If you see blips at consistent RPM under steady-TPS,
steady-spark, steady-fuel operation then might suspect
a knock sensor sensitivity or true false knock (heh) and
that would be mechanical noise (valvetrain or exhaust
buzzing) or sensor sensitivity or a combination. I think
there are sensitivity params now, of some sort.

RedHardSupra

iTrader: (1)

whoa, i had 0's and then -3's for 'max kr vs rpm' that's quite a bit different from your 10..20 range could've that been it? i'm gonna try to test it later, my car is about to go under the knife for all the parts i stockpiled over the winter anyway, so i'm gonna have it looked over for loose heatshields and whatnot that could bang and make noise.
i hope this is it....

jimmyblue

iTrader: (11)

Those "Max KR" values are the lowest advance, that any
retard function is allowed to pull back to. The name is just
not right. I use these settings to limit the authority of TM.
If KR is pushing you back that far, that's a whole lot of
knock or a whole lot of sensitivity.

You really have to arrive at a mental model of what the KR
is following. Does it show up at a fixed RPM (or several
separate ones)? Just at opening throttle? Can you make it
go away by dumping fuel on it? The circumstance points to
the cause, if you can see that more clearly with the tighter
KR attack/decay profile then should be able to track it down
to the source.

RedHardSupra

iTrader: (1)

this whole weekend i did a few scans with your new gain/attack settings and this is what i got:
these are all 'max' values, on avg they are get drowned out by 0's so they dont' show up as often, just demonstrating how random it is.






why is it that they always occur at the same spot? does it mean it's mechanical?
thanks,
Marcin