Knock Sensor ECU Reaction

[Nismo_Freak]

How does the ECU react to a malfunctioning knock sensor code?

[corn322]

I'd imagine it just takes the timing way down, so there would be very little chance for knocking.

[Nismo_Freak]

I'd imagine it just takes the timing way down, so there would be very little chance for knocking.

Depends... some cars react differently.

Most Mazdas for example stick to their base timing curves and do not variate when the knock sensor goes dead.

[Q45tech]

Nissan generally retards the part throttle timing [when KS fail from overheating] more than the WOT timing, so cruise and mild acceleration seems sluggish [bucking at low rpms] around 55-65 mph pulling hills slight grades.

In Japan the EVERY 90 day service [dealer oil change and intensive look over] includes a diagnostic [Consult] where these things are caught.

[s13sr20chris]

In Japan the EVERY 90 day service [dealer oil change and intensive look over] includes a diagnostic [Consult] where these things are caught.

that is so cool.

[Eswift]

there is a different fuel and timing map that is used when a knock condition is detected.

at least on VG series for sure.

[s13sr20chris]

fuel too? this i did not know. i would not doubt it though.

[Eswift]

yep, one for the fuel and another for the spark. There are two maps for each, one for high octane (normal operation) and the other set is for low octane (knock detected, or safety boost mode for the Turbocharged verisions)

[Q45tech]

Why the gas mileage suffers, 10-15% extra fuel [besides the 5 degrees less advance] to cool the combustion down so less knock is possible.

[Eswift]

<cough> water injection <cough> lean it up <cough>

[s13sr20chris]

well now thats cool. in a spark knock sucks kinda way. i remember from the supercharged frontier new model class that there was a special fuel map for "low octane" situations. i assumed that since they were mentioning it, it must be new. it now makes sense that it was not new, but rather just a refresher. they made a big deal about how much slower an already slow vg33er equipped frontier is with zero boost, retarded spark, and too much fuel. that will sure enough keep any customer from using regular.

[reggiegsd]

I'll as the question here too.

Does anyone know the resistance the ECU sees when it looks at the knock sensor?

[s13sr20chris]

it really should not be looking for resistance. the ecm listens for a signal from the knock sensor as part of its operation. also it tests voltage drop across an internal resistor to see if the circuit is open or shorted.

[Eswift]

none

[s13sr20chris]

yeah, come to think of it since the knock sensor outputs a small signal any resistance would quickly become null as far as the ecm is concerned.

[Eswift]

exactly.

[MrFox]

Its a piezo(sp?)electric device - very similar to a camping stove lighter. It generates its own voltage from the knock vibration. Like a O2 sensor which generates its own voltage from oxygen differential, it doesn't require ECU power input.

[Eswift]

correct, and correct spelling.

[Q45tech]

The signal is filtered [audio bandpass filter [in the 5-8KHz range]].

The bore diameter is the gating item in how long the knock takes to reach the block and the resultant frequency spectrum of the knock............so the ecu does get confused by other worn engine sounds.

Also there is a timing gate so that the KS is only listened to when knock could occur. Say for 15 degrees every other 90 degrees and alternating banks................each bank is separately monitored and accumulated and decayed. Counts per secound.Light knock vs heavy knock stores less voltage on a capacitor which is bleed down with a parallel resistor........counts above a threshold activate a progressive retard circuit.

Knock occurs at less than 6 degrees ATDC [from just after ignition 15-25-30 BTDC].

You can use a high impedence [greater than 1 Megaohm] [calculate the reactance of a series capacitor] amplifier to listen to knocks but don't forget to shape the response with a 5KHz high pass filter.

[s13sr20chris]

Its a piezo(sp?)electric device - very similar to a camping stove lighter. It generates its own voltage from the knock vibration. Like a O2 sensor which generates its own voltage from oxygen differential, it doesn't require ECU power input.

right it does not need it. it does, however, get it. all input signals get .3 volts from the ecm just for open/short diagnostics. that is how the ecm checks stuff with only one wire. this is according to jim at the washington dc area nissan tech school. it may be specific to a few model years and may not cover all input signals, but thats what he told me.

edit: just to add to that, the .3 volts is the reason o2 sensors often rail at .3 when they fail.