2GR Knock Setup

[Aqmar]

I am having a challenge to determine the right frequency channel. I am with the 2GRFE engine using the Toyota original knock sensor on each bank. Everything is stock standard. 

I have to say that I am still a novice in doing this. Experimenting things slowly. 

Based on the bore calculation, I should be selecting a 6kHz Narrow Band. Under the 6 kHz, with a gain channel of 20, I don't seem to get any knock detection. I increase the gain channel to 50, and it gets sensitive. Am I doing it the right way?

Anyone here willing to share their experience determining the right setup for the 2GR knocks sensing? below is my data log for the knock setup

[Adamw]

The gain should not need to be that high, either there is a wiring issue, or it is a resonant frequency sensor and you are looking at the wrong frequency, or the sensor is something special, not useable.   I have never seen anything need a gain of more than about 5.  With a gain of 50 the poor signal to noise ratio will not give you anything useful.

[Aqmar]

Thanks, Adam,

How interesting.

The wiring seems fine. Maybe I am reading it wrongly. So I set the ignition timing conservatively and ran at 3 different channels: - 6kHz, 8kHz and 12 kHz. To ensure reduce any variables, the gain is set at 5. 

I drove the car but didn't push it above 4000rpm. From my own perspective, I don't see much difference in the reading between all the 3 different channels. 

I know the best way to determine the knock is by using the KnockBlock kit. Unfortunately, I don't have one around now and will need to get it then.

Therefore, before I proceed to use the KnockBlock, is there anyways I can at least determine the right frequency for the 2GR engine? Here are the logs reading based on 3 frequencies. Is this how the data should read?

https://drive.google.com/drive/folders/1-UpC2wb30mQfnsu7yySIzLzRyKJJoB-T?usp=sharing

 

 

[Hodgdon Extreme]

1 hour ago, Solace said:

is there anyways I can at least determine the right frequency for the 2GR engine?

It's a straightforward calculation:

(573000 / bore diameter in mm) / 1000 = knock frequency in kHz

For you, it should be 6 kHz for the first order of resonance, or 12 kHz for the second

The issue I see with your arrangement is your sensors are super duper quiet. It should normally only take a Gain Channel of 3-5 to see a Knock Level amplitude of 200-300 or so... Yours are 2-5, so about 100 times lower than they should be.

First thing I'd do is swap the sensors out for flat-response sensors. However, the Knockblock unit comes with a pair of them - so if you're planning to buy a Knockblock anyway, you'll get the better sensors. 

[Adamw]

Yeah, the sensor(s) is not outputting anything useful, or there is a wiring issue.  If these are two-wire sensors, try swapping the ground/signal wire, some 2 wire sensors have continuity between the mounting face and the ground pin so if wired incorrect polarity you short the signal wire direct to ground.  

[Aqmar]

On 7/4/2023 at 7:19 AM, Adamw said:

Yeah, the sensor(s) is not outputting anything useful, or there is a wiring issue.  If these are two-wire sensors, try swapping the ground/signal wire, some 2 wire sensors have continuity between the mounting face and the ground pin so if wired incorrect polarity you short the signal wire direct to ground.  

@Adamw

Thank you for pointing that out. You are absolutely right. It was the wiring of the ground and signal. Now that has been sorted out, and I have gotten a knock reading with only a gain of 2! 

@Hodgdon Extreme, based on your calculation, my resonance is 6kHz. I don't understand the objectives of having the 2nd order of resonance. How can I use that as a reference? 

 

However, guys, I have taken a log of the reading using three freq channels; 6kHz, 7kHz and 12kHz. How can I determine which is the right frequency to use based on this data log? Are that ways to determine the right frequency to use without using the standalone Knock sensor audio?

I have reduced the gain for cylinders #3 and #4 because, based on my data loggings, I observed that these two cylinders read above average from the rest of the cylinders. Maybe due to the location of the OEM knock sensor being nearer to these two cylinders.

 

https://drive.google.com/drive/folders/1-UpC2wb30mQfnsu7yySIzLzRyKJJoB-T?usp=sharing