Reducing knock sensor Voltage. What size resisitor?
01-10-2007, 04:52 PM
#1
Member
Thread Starter
Join Date: Feb 2003
Posts: 1,338
Likes: 0
Received 0 Likes
on
0 Posts
Reducing knock sensor Voltage. What size resisitor?
In an on going thread in the Prelude forum with people having problems with there cars hesitating, with there ECU momentarily (3-5 seconds or so) retards the timing and dumping a ton of fuel.
Here's the thread that I am talking about.
https://honda-tech.com/zero...age=1
now the theory that I need your guys help with
(this is cutting and pasting from my last post in that thread)
I've put a little more thought into this topic and this is my theory.
Alot of people are pointing towards knock sensor as the issue, and alot of people are screaming get a new one, but I think that the stock one is doing its job.
The knock sensor is simiply a chunk of peizo electric material which basicly converts shock/vibrations into current. When there is detonation, it cause a soundwave that triggers a large current spike in the knock sensor.
My question to many of you, how many of you kept your balancing shafts? I did not. This could be a reason why I am sometimes having this problem. The frequency is smaller so the peaks in the sine wave of vibrations are spaced further apart. As RPM's rise, they become close and closer together to the point where the the knock sensor sees it as one continuous peak.
The purpose of the balancing shafts is to cause a vibration that will cancel out the main vibrations cause from the crank and other moving parts (destructive interference) When functioning, these cause the amplitude of the vibrational waves to be smaller. Now because they are not functioning, these waves are back to their original size. Because it is at a lower RPM, as I said, these peaks are spaced further apart and because the amplitudes are bigger then what it was programmed from stock, on occasion, the amplitude of the the wave is large enough to trigger a spike in current high enough for the ECU to think that a knock has occurred.
So the knock sensor is NOT faulty, it is just oversensitive.
Now how do you solve this. The knock sensor sends low voltage to the ECU at lower RPMs and linearly increases the amount of current as RPMs climb. When its a steady increase, the ECU acts normal. When there is a very sharp jump in voltage, the ECU interprets this as a knock. Now, this is for all you EE's out there (I'm a materials engineering student with limited circuits background) what size resistor would bring the voltage levels down a bit.
Does anyone have an idea on the size of resistor would work for this application?
Here's the thread that I am talking about.
https://honda-tech.com/zero...age=1
now the theory that I need your guys help with
(this is cutting and pasting from my last post in that thread)
I've put a little more thought into this topic and this is my theory.
Alot of people are pointing towards knock sensor as the issue, and alot of people are screaming get a new one, but I think that the stock one is doing its job.
The knock sensor is simiply a chunk of peizo electric material which basicly converts shock/vibrations into current. When there is detonation, it cause a soundwave that triggers a large current spike in the knock sensor.
My question to many of you, how many of you kept your balancing shafts? I did not. This could be a reason why I am sometimes having this problem. The frequency is smaller so the peaks in the sine wave of vibrations are spaced further apart. As RPM's rise, they become close and closer together to the point where the the knock sensor sees it as one continuous peak.
The purpose of the balancing shafts is to cause a vibration that will cancel out the main vibrations cause from the crank and other moving parts (destructive interference) When functioning, these cause the amplitude of the vibrational waves to be smaller. Now because they are not functioning, these waves are back to their original size. Because it is at a lower RPM, as I said, these peaks are spaced further apart and because the amplitudes are bigger then what it was programmed from stock, on occasion, the amplitude of the the wave is large enough to trigger a spike in current high enough for the ECU to think that a knock has occurred.
So the knock sensor is NOT faulty, it is just oversensitive.
Now how do you solve this. The knock sensor sends low voltage to the ECU at lower RPMs and linearly increases the amount of current as RPMs climb. When its a steady increase, the ECU acts normal. When there is a very sharp jump in voltage, the ECU interprets this as a knock. Now, this is for all you EE's out there (I'm a materials engineering student with limited circuits background) what size resistor would bring the voltage levels down a bit.
Does anyone have an idea on the size of resistor would work for this application?
01-11-2007, 09:24 AM
#3
Honda-Tech Member
Join Date: Aug 2004
Location: Sterling Heights, MI, USA
Posts: 2,006
Likes: 0
Received 4 Likes
on
3 Posts
Re: Reducing knock sensor Voltage. What size resisitor? (asmallsol)
Without knowing the voltage range (I'm guessing 0-5v) and current, it's impossible to figure out the size of the resistor. You can try putting a 1k in series with the sensor and see if it helps.
01-11-2007, 02:00 PM
#5
Voltage on knock sensors stay the same all the time, they never variate. I usually wire in a 5 ohm resistor into chevy knock sensors to get them to pass OBD emissions with lumpy idle cams.
To find out the voltage it produces, measure the voltage coming out of the knock sensor on a normal running car at idle.
To find out the voltage it produces, measure the voltage coming out of the knock sensor on a normal running car at idle.
Thread
Thread Starter
Forum
Replies
Last Post
