Low RPM, high map knock

[Superdave]

I have the set from my '04 3500 for sure, and i may have the set from my '05.. both looked the same and i think both were non flex-fuel engines however they did have the blue Orings. I'll post up part numbers later.

[LZeppelin513]

The engine, according to the yard, was from a 2005 malibu. I am out of town and do not have the Fiero with me but I will try to check next weekend. Does the plenum have to come off to get to the part number?

[LZeppelin513]

for both stock and wild 3500's i keep the advance above 4K between 26 and 28 degrees, below it's usually in the low 20's. I'd add a little more fuel to get it into the low 12 AFR range and bump the advance up. If it's still knocking then post up a datalog, something isn't right.

https://docs.google.com/open?id=0B5PFNNInFEsbSGZILW9IV3dDYm8

Here is a data log of 3 wot blasts. I highlighted the wot portions and also new knocks.

[Superdave]

Interesting...

I'd back off the AE a little to start with and see if that helps.

[LZeppelin513]

OK I was thinking the same. I haven't touched transients yet (and this is going to sound dumb) but Im not sure the best way to back it off... Do I just reduce each value in the "AE mult vs AE Pulse" Table? If so, I don't know how the "pulses" work.

With a slight blip of the throttle I need a wee bit more AE, with heavy throttle I need quite a bit less. Would this mean increasing mult. in low Delta TPS and decreasing it in high Delta TPS in the "AE mult vs D-TPS" table?

In that Log I included the timing advance. I think it is strange its knocking at all since I have pulled a bunch of timing.

[LZeppelin513]

double post

[LZeppelin513]

I put in premium in the fuel tank and wrapped the knock sensor with insulation and teflon tape on the threads. I am still getting consistent knock at low RPM and high map. I am also getting a couple knocks on just about every WOT run. So... I think it is false knock because it should have went away with the premium. I also held it at 3000 RPMs and 85 map and racked up knock counts and KR of 10 degrees and it still kept knocking. I would think with the KR of 10 degrees it would be enough spark reduction to stop knocking.

I also ran a can of seafoam through the intake, just in case there was some carbon deposits causing the knock.

Any ideas?

[robertisaar]

yeah, that's false knock.... now to figure out what path you want to go down to get rid of it.

[LZeppelin513]

What are my options? I took some scotch bright pad and cut it up to wrap the knock sensor, then I wrapped it all in self fusing rubber tape. I doubt I'll be able to noise insulate the outside any better than that.

I bought the knock sensor new a couple years ago. Im pretty sure it was a high quality AC delco, delphi or equivalent. It was for an early 90's 3.1.

[robertisaar]

couple of different routes to take... can attempt to make it less mechanically sensitive or less electrically sensitive.

electrically, you can modify the circuit via a resistor pair so that when the sensing element swings around, the knock filter doesn't see as much of it. a small(REALLY small) capacitor may do well for this too, but i don't think i've heard of anybody doing it. otherwise, you can attempt to desensitize the knock filter(not sure how to go about that ATM), or even build your own(or modify an existing method) knock sensing mechanism and have it ground a pin on the ECM connector(the 7730/7727 does support this capability). or even swap out the knock filter on the MEMCAL for one that could work better for your engine.

mechanically, different amounts of torque on the sensor into the block will effect it's sensitivity, as does the number and type of connections it's threaded to(some TGO guys have to mount their sensor into a brass 45* elbow), among a couple of other options.... thirdgen.org DIY PROM forum has a lot of threads about it, though some won't be directly applicable.

[pocket-rocket]

Did you torque it to spec? I've been told overtorquing will bury too much of the ks in the block making it too sensitive.

Sent from my HTC One X using Tapatalk 2

[3400-95-Modified]

my large issue of KR was solved by wrapping the Knock Sensor in windshield butyl rope, same goes for Roberts 89GT that I did. I think the noise was coming from the headers though.

[LZeppelin513]

I already had the sensor wrapped in a thick casing of rubber self fusing tape and was still getting the false knock.

I added 2 brass fittings, a 45* street elbow, and a .5 inch long strait adapter. I am no longer getting knock. It still works because I could hit the engine with a wrench and get knock counts/KR. I also got about 20 knocks on start up and shut off, so it is still working, but no more knock at low rpm/high load or wot!

Thanks for the advise, engine feels strong now that Im not getting a bunch of KR!

[robertisaar]

have you tried a small burst of unnecessarily high SA to verify that when real knock should be detected, that it will?

otherwise, good job, false KR is usually a really tricky thing to diagnose.

[LZeppelin513]

I didn't try that. Ill give that a shot this weekend. Would it be safe to just start upping the advance with the emulator while cruising at a normal rpm and load until i either hear knock or it detects knock? My engine is a fair amount loader than stock, should it be easy to hear when it starts pinging?