F54 Block Knock Sensor Threads

[duragg]

Anybody know off hand what the thread size and pitch for the knock sensor hole on an F54 block is?

Looks about like a 10mm or 12mm but unsure the thread pitch.

 

Gracias.

Tj

[Lazeum]

Most of the threads are fine thread on our cars. M10 & M12 are usually going with 1.25mm pitch

Edited August 24, 2011 by Lazeum

[duragg]

Thank you.

I am mounting a "Knocksense" knock sensor in that position.

However the Bosch sensor included has an 8mm ID and I need to neck down the block hole from 12 to 8.

 

Probably use some type of thread repair gizmo.

Tj

[Zmanco]

I'm running KnockSense with the 81-83 ZXT knock sensor. Just a thought: it might be simpler to source one of those.

[duragg]

Do you like it? (The Knocksense?)

 

I should use the Datsun sensor, but the guy already calibrated for this Bosch.

 

Did you have to do any tweaking?

Does it go off much.

 

TJ

[Zmanco]

I would encourage anyone running a turbo with more than 7-8 psi boost to run one, even if all it does is flash a light to warn you. I had what I thought was a conservative tune (done by ear on the street) but was surprised at how often there was detonation that I couldn't hear. At first I thought it was false triggers, and in fact some were, but when I retarded timing in those regions most of them went away. FYI, I have it feeding into MS2/Extra which uses it to pull timing.

 

A few thoughts:

 

- I had the same concern about tuning the module based on sensor but Boris confirmed that the module is tuned to the bore of the engine, not the sensor.

 

- I found that above 6500 rpm there was so much valve train noise that it was false triggering, even with the timing retarded by large amounts. I eventually set MS to ignore it above 6500.

 

- The MS/Extra logic is a bit flawed in that it pulls more timing than required based on how it polls the input for knock event. I hope they will fix this on the next release.

 

- On track days I have piece of mind that I'll be able to drive the car home because MS will pull timing as it detects detonation. The air intake and engine temps are all elevated and hence outside of the range for which I tuned. But I know that the worst thing that will happen is MS will pull a few more degrees of timing than is perhaps optimal resulting in some power loss. FYI, I recently pulled the head to swap cams and after 4 track days there was no evidence of detonation on the pistons or cylinder head.

[duragg]

Interesting input.

 

I am Normally aspirated with Triples and a Wideband and now this Knock Sensor.

The motor is 10.3:1 which they say is knock prone.

Phoenix air induction temps are a bit high.

 

Will get it installed this week and find out!

 

TJ

[HowlerMonkey]

Any of you guys worked with trying to attenuate the noise between the block and the knock sensor so the valvetrain noise doesn't trigger it?

 

I guess I need to do some capturing of the "sound" by various sensors and check the waveform under true knock and no knock conditions so maybe we can find out which sensor works best for the nissan L6.

 

The problem with knock sensing is that different sensors have different outputs determined by the knock frequency of the application for which they are tailored.

 

I just got done running 87 octane in a f54 with flattops and a p90 head with 8 pounds of boost (intercooled) and am finding no audible pings or knocks so it's possible my infiniti M30 ecu using a M30 knock sensor screwed into the f54 block knock sensor hole is actually working as it should........but I doubt it's optimum because I'm sure the vg30 sounds the sensor expects to hear are a little different in character.

 

One of the perks of having a dead silent exhaust system is being able to actually hear what's going on with the engine.

Edited August 25, 2011 by HowlerMonkey

[Zmanco]

Any of you guys worked with trying to attenuate the noise between the block and the knock sensor so the valvetrain noise doesn't trigger it?

I haven't tried that - how would you address that with the stock sensor?

 

I was assuming that Nissan had done some homework in this area already, but realize that might be wishful thinking.

[Lazeum]

I've thought about how to do it. I've already done this kind of work on brakes during my studies (NVH engineering)

 

We would need to take the signal, cut it into pieces, transform the piece of signal with Fourier transformation (to translate the signal from being "Noise amplitude in dB" = f (time) to "Noise amplitude in dB" = f (frequency) ) and get some measure average done to reduce backgroud noise (white noise)

According to engine RPM, you should be able to get noise made by valves with the appropriate frequency & harmonics.

 

All that would have to done with correct sampling frequency, correct filtering options, correct management of samples, etc.

 

No so easy to do if you want to do it right

[duragg]

I ended up using an M12 x 1.25 metric bolt, cut off the head on a lathe, bored a hole down the center, threaded for 1/4-20 threads and parted off a piece about .380" long then cut a slot in the top so I could screw that into the head and then mount the Bosch Knock Sensor into that adapter.

 

Works fine.

 

The KnockSense has some known sensitivity problems with most motors at higher RPMS and I still need to do some calibration.

Many say over 5000RPM the engine noise is too high and to disregard.

 

But it is cheap enough and the sensitivity can be adjusted to make it usable in certain instances.