Yes I've searched.

How do you get the stock NB2 1.8 knock sensor to give useable readings?
# 1-4 readouts all show solid readouts that follow RPM pretty well, indicating pretty much 1 unit per thousand RPM. They do not seem to indicate actual knock just vibration from motor?
Tried very quick throttle transitions but no tip in knock indicated?

Also having a problem getting MPH to log even though it is selected?

 

If there's no knock, there's nothing to detect. You sure you have knock?

 

That's part of the problem. I really cant hear any, but then I cant hear worth a dam anyway. The clutch makes some noise at low RPM but not nearly the engine knock I heard normally under the stock ECU and tune.

Really want to know if the stock sensor will record any, and if there is a way to set up a filter so if I see any I will know where and when to pull a few degrees.

The catch is I will not hear it on track at high RPM, and hope the stock sensor has enough resolution to be better than my ignorant ears.

Not sure why each of the 4 sensor readouts is a slightly different frequency and what the knock spikes would look like?

 

I've only got it to read .1-.5 volts. I switched to a GM sensor that reads a more usable 0-5 volts. Very easy to setup and monitor. I just an adapter kinda like TSE's oil feed adapter.

I'm sure there's better ones out there to use, that may have a better thread pitch. I know the through bolt ones would probably be ideal with all the threaded holes available on a Miata block.

https://www.summitracing.com/parts/a...hoCBQEQAvD_BwE

 

That log looks completely normal for an NB2 knock sensor. Max of 4-5 units for a stock motor. Built motors are a little noisier, high-boost setups are noisier still. I think my filtering ramps up to 16 units or so at max boost on Acamas. I set a manual scale of 0-25 on the knock graph which helps show spikes better. Usually #2 and #3 are noisier because they are closer to the sensor.

If any single cylinder indicates 50% higher than the other 3, then that's knock. It's usually not subtle in the graph, and it's usually much higher than a 50% spike.

IMO the NB2 sensor works well enough that I've never been inclined to spend time looking elsewhere.