Knock Sensor Retard Timing?

[Drakey]

Hi All,

Quick question. If i have a faulty knock sensor on my rb20det, Is there anyway it can retard timing at high revs?

Or even prevent the motor from advancing the timing at high rpm?

Thanks

[**RB2530**]

It would depend on what the fault is I think...I do know that if you disconnect the knock sensors the ECU detects this and automatically pulls timing out over the whole rpm range..So if knock sensor is open circuit, the same thing would happen...and open circuit sensor would be the most common failure mode..

[Sandeep]

retarding the timing on stock ecu would require a remap/ecu socket with a new chip written with your tune

are you actually pinging? you can bypass the sensor with a resistor (i think 1 mega ohm is what's used on the z32's) so that the fault code will no longer come up but it's not safe if you are actually pinging. most people who are running safe mods can get away with bypassing the sensor as stock fuel maps are rich.

check out these links for info on bypassing detonation sensor (z32 specific, might apply to r32 also)

http://www.pexcom.com.au/z32cms/e107_plugi...php?content.429

http://www.pexcom.com.au/z32cms/e107_plugi...php?content.305

[Drakey]

retarding the timing on stock ecu would require a remap/ecu socket with a new chip written with your tune

are you actually pinging? you can bypass the sensor with a resistor (i think 1 mega ohm is what's used on the z32's) so that the fault code will no longer come up but it's not safe if you are actually pinging. most people who are running safe mods can get away with bypassing the sensor as stock fuel maps are rich.

check out these links for info on bypassing detonation sensor (z32 specific, might apply to r32 also)

http://www.pexcom.com.au/z32cms/e107_plugi...php?content.429

http://www.pexcom.com.au/z32cms/e107_plugi...php?content.305

Nah no pinging... just lack of power in the upper area of rpm. Im thinking it might be a timing issue. And i was thinking if there was something wrong with the knock sensor then it might not be allowing the timing to advance thus causing lack of power and worse fuel economy.

seeing as though it is not pinging it may be worth by passing to see if it runs ok.

[Drakey]

and i quote this from the above link:

"A faulty Det Sensor or det sensor circuit causes the ECU to go into safety map (retarded timing and rich mixture= poor performance, bad fuel economy "

That exlains my problem down to a t.

Anyone have a knock sensor for sale?

[(--MeUnVmosT---)]

i have this same issue cuz my knock sensors were not wired in.. its always in safe mode, running full rich with off the wall timing. my Wideband was going FULL rich at high rpms. so im off to wire them in!!

this was good info.

my question!!

how can u wire these knock sensors directly to ecu? can i just wire 1-3 to 23 and 4-6 to 24 direct? is it a 1 signal wire with a ground?? any info would be cool. thanks!

[666DAN]

Depends on what the ECU is looking for, most of the Nissan ECU's will register a fault if the knock signal is <1v or >4v that's why I've never understood the resistor mod, as the ECU is looking for voltage NOT a load.

If the knock sensor is over sensitive you can wire a series of diode in-line and it will reduce the volage by about 0.3v per diode (what I had to do to my one to get it to ~2v at idle)

If you're open circuit most likely the sensor is dead and you need a new one

[**RB2530**]

Depends on what the ECU is looking for, most of the Nissan ECU's will register a fault if the knock signal is <1v or >4v that's why I've never understood the resistor mod, as the ECU is looking for voltage NOT a load.

If the knock sensor is over sensitive you can wire a series of diode in-line and it will reduce the volage by about 0.3v per diode (what I had to do to my one to get it to ~2v at idle)

If you're open circuit most likely the sensor is dead and you need a new one

I have researched this and done the mod using a resistor. It works......Basically the ECU provides a constant current across the knock sensor. As the resistance of the knock sensor changes, the voltage drop across it changes....V=I x R gents...If you search my profile I have done a thread on it..

cheers

[**RB2530**]

http://www.skylinesaustralia.com/forums/Di...;hl=rich+retard

[666DAN]

OK then it's the R34 that uses a different type of detection, the ECU recognizes voltage for the GTt

[**RB2530**]

OK then it's the R34 that uses a different type of detection, the ECU recognizes voltage for the GTt

The knock sensors would be passive sensors on all models, R32s R33s R34 R35 ...The ECU would be measuring voltage drop across the sensor due to the fact that it has a current passing through it...This is most likely how all car manufacturers do it... This is the best way to do it because it is less prone to signal noise and interference from other things happening in the engine bay...ignition system etc

[666DAN]

It's definately active on the R34, the workshop manual has the oscilloscope readings you should be getting from the sensor and goes into detail about the voltage range. Also any attempt to put a resistor accross the input throws a fault on consult straight away.

[**RB2530**]

It's definately active on the R34, the workshop manual has the oscilloscope readings you should be getting from the sensor and goes into detail about the voltage range. Also any attempt to put a resistor accross the input throws a fault on consult straight away.

Very interesting...When I did my mod, I had Consult connected and experimented with different resistor networks until the fault code dissappeared...

there would definately be a way of tricking it...how many wires go to the R34 sensor? There must be at least 2 or 3 if it is active as you say...a ground -ve, supply +ve and a signal....

There would be a specific voltage range across the sensor for baseline engine noise under "No-knock" condition. If you know what this voltage is, you could use a zenner diode to clamp the voltage under the Knock voltage...Then as far as the ECU is concerned it would still see normal engine noise signals, but never a knock signal...

Disclaimer: Disabling knock sensors is playing with fire if you dont know what you are doing!!!

Sorry just had to put this out there again as i dont want to be blamed for rooting anyones engine...

[666DAN]

Just one wire to the sensor. Between the sensor and ground I get 2.4v DC & 3.4v AC (remember my one is shagged anyway so voltage is off)

2.0v is the recommended reading at idle according to the manual, yeah I've got my one set about there with a diode and now my low/mid-range has reappeared, boost is holding top-end

[**RB2530**]

Thats really interesting. Just confirm that the knock sensor is not connect to the ECU when you are measuring this voltage...

2.4V is the RMS voltage of 3.4V AC...In the manual do they refer to frequency at all?

[666DAN]

Yes that's correct completely detached from the ECU, so it's definately up to no good on its own, no frequencies mentioned