!!!read Here!!!(why You Shouldn't Use An Lt4 Km)
#1
Registered User
Thread Starter
Join Date: Feb 2003
Location: NE philadelphia pa, FOX CHASE
Posts: 236
!!!read Here!!!(why You Shouldn't Use An Lt4 Km)
These are my notes as to why you shouldn't use an LT4 knock module on a car not designed for one. These notes come from two compelling emails I saved from the LT1edit mailing list. I claim nothing here a my original thoughts or work.
Wednesday, March 20, 2002 9:43 AM
[LT1edit] Why the correct knock sensor circuit resistance is important.
I can explain it in more familiar terms for us. We know the knock sensor is a piezoelectric sensor, basically a microphone. To visualize we can think of a microphone like a car/ home stereo speaker, except instead of broadcasting sound, it receives it. Our knock sensor is poorly optimized for a certain frequency range. So the knock module acts as a bandpass filter, meaning it lets certain frequencies through that indicate real knock (in this case prob around 5KHz) and filters out the rest. Back to the speaker example, anyone that has played with stereo knows about crossovers, which include bandpass filters (like the knock module.) Different speakers have different resistance (impendance) depending on how they're wired, car speakers are generally 4 Ohms and home speakers are usually 8 Ohms. A bandpass filter that has a low cutoff frequency of 5KHz on a 4 Ohm speakers will have a 2.5 KHz cutoff on a 8 Ohm speaker for example. So here we see why the resistance of the knock sensor circuit is crucial, if the circuit has the wrong resistance (wrong sensors, 1 instead of 2, etc) then the frequency range the knock module filters is completely different. This has been my answer why the LT4 knock module isn't optimum for a single OBDI setup, it's designed for 96 PCM, which has two sensors, different circuit resistance (3.8K Ohms versus 1.9K Ohms I believe.) If you've changed anything, just make sure the total circuit matches, don't mix and match or you'll get false reading, and now you know why.
Hope this makes sense,
Programmer
There is no possible way to "desensitize" a knock sensor (other than disconnecting it or shielding it from engine vibration). You can only alter the amount of timing retard that would normally occur in the software. This in NO WAY desensitizes a knock sensor. A different knock module filters more of the "noise" that could be interpreted as "false knock".
Wednesday, March 20, 2002 9:43 AM
[LT1edit] Why the correct knock sensor circuit resistance is important.
I can explain it in more familiar terms for us. We know the knock sensor is a piezoelectric sensor, basically a microphone. To visualize we can think of a microphone like a car/ home stereo speaker, except instead of broadcasting sound, it receives it. Our knock sensor is poorly optimized for a certain frequency range. So the knock module acts as a bandpass filter, meaning it lets certain frequencies through that indicate real knock (in this case prob around 5KHz) and filters out the rest. Back to the speaker example, anyone that has played with stereo knows about crossovers, which include bandpass filters (like the knock module.) Different speakers have different resistance (impendance) depending on how they're wired, car speakers are generally 4 Ohms and home speakers are usually 8 Ohms. A bandpass filter that has a low cutoff frequency of 5KHz on a 4 Ohm speakers will have a 2.5 KHz cutoff on a 8 Ohm speaker for example. So here we see why the resistance of the knock sensor circuit is crucial, if the circuit has the wrong resistance (wrong sensors, 1 instead of 2, etc) then the frequency range the knock module filters is completely different. This has been my answer why the LT4 knock module isn't optimum for a single OBDI setup, it's designed for 96 PCM, which has two sensors, different circuit resistance (3.8K Ohms versus 1.9K Ohms I believe.) If you've changed anything, just make sure the total circuit matches, don't mix and match or you'll get false reading, and now you know why.
Hope this makes sense,
Programmer
There is no possible way to "desensitize" a knock sensor (other than disconnecting it or shielding it from engine vibration). You can only alter the amount of timing retard that would normally occur in the software. This in NO WAY desensitizes a knock sensor. A different knock module filters more of the "noise" that could be interpreted as "false knock".
#2
Registered User
Thread Starter
Join Date: Feb 2003
Location: NE philadelphia pa, FOX CHASE
Posts: 236
I WOULD LIKE TO HERE FROM ANYONE THAT HAS ANY THING TO ADD TO THIS THAT KNOWS WHAT ALL THIS MEANS AND APLLYS TO LIKE WHEN DO U KNOW WHEN YOU REALLY NEED TO USE ONE LIKE WOULD I OR NO CAUSE I ONLY HAVE HEADERS?
#5
While it sounds well written, is it accurate? First off, my 96 only has one knock sensor so that bit of info right there puts the entire thing in question.
I've been using the LT4 knock module in my cars for a very long time. They serve their purpose for me. In fact, I once got a bad batch of gas... I would get 10 degrees of knock retard everytime I barely stepped on the gas. Filled up with good gas later on and all was back to normal. This little non-scientific test along with the "hammer test" show that noise/knock can still be detected when using a LT4 module in a OBDI PCM.
While I realize I am reducing the ability for my PCM to detect real knock, I do know that the ability for the PCM to detect pinging is still there... just "desensitized" so to speak. Or, in other technical terms ... it just won't detect real or even false knock quite as good as the stock knock module. But, since that is exactly what I want and I'm willing to accept the risks involved, the LT4 knock module serves ME well
I've been using the LT4 knock module in my cars for a very long time. They serve their purpose for me. In fact, I once got a bad batch of gas... I would get 10 degrees of knock retard everytime I barely stepped on the gas. Filled up with good gas later on and all was back to normal. This little non-scientific test along with the "hammer test" show that noise/knock can still be detected when using a LT4 module in a OBDI PCM.
While I realize I am reducing the ability for my PCM to detect real knock, I do know that the ability for the PCM to detect pinging is still there... just "desensitized" so to speak. Or, in other technical terms ... it just won't detect real or even false knock quite as good as the stock knock module. But, since that is exactly what I want and I'm willing to accept the risks involved, the LT4 knock module serves ME well
#7
I'm pretty sure that the only thing that cares about the knock sensor impedance is the terminal it's connected to at the PCM. Once the signal is inside the PCM, they would seperate the frequency curve that is riding on the bias voltage, and process that with the KM.
The second quote about desensitizing is irrelevant to the knock module. As was pointed out in the first post, the KM is simple a filter to reduce the sound signatures that are known to be associated with the LT4 valve train noise. No one is desnsitizing the KS with the KM, they are filtering it.
The point about the dual knock sensors on the Corvette LT1 is also questionable... each knock sensor on the Corvette connects to an independant circuit in the PCM and controls individual banks. It is not likely that the PCM sees the "combined" load of the two sensors.
Too many people have verified that there is little if any impact from the LT4KM on the knock retard function of the OBD-I PCM, except to filter "false" knock. I spent a lot of time verifiying that the LT4KM would respond properly to "true" knock situations, like poor quality fuel, and if performed just like the stock KM in that respect.
The second quote about desensitizing is irrelevant to the knock module. As was pointed out in the first post, the KM is simple a filter to reduce the sound signatures that are known to be associated with the LT4 valve train noise. No one is desnsitizing the KS with the KM, they are filtering it.
The point about the dual knock sensors on the Corvette LT1 is also questionable... each knock sensor on the Corvette connects to an independant circuit in the PCM and controls individual banks. It is not likely that the PCM sees the "combined" load of the two sensors.
Too many people have verified that there is little if any impact from the LT4KM on the knock retard function of the OBD-I PCM, except to filter "false" knock. I spent a lot of time verifiying that the LT4KM would respond properly to "true" knock situations, like poor quality fuel, and if performed just like the stock KM in that respect.
#12
JMHO:
Since I only mod the 93 a LT4 KM isn't in my future So, I'm sorta non-biased.
BUT,
If GM would install these on the cars they sold to the public, and they have to service the cars later, I can't see where they would see it is a "bad' thing.
but then again they thought the opti was a good idea too
Since I only mod the 93 a LT4 KM isn't in my future So, I'm sorta non-biased.
BUT,
If GM would install these on the cars they sold to the public, and they have to service the cars later, I can't see where they would see it is a "bad' thing.
but then again they thought the opti was a good idea too
#13
My 1994 Z28 M6 has 1 knock sensor. I too plan on changing it to the LT4 sensor and module, as my knock retard pegs at 10 deg even with racing gas (Increased valvetrain noise from headers?).
My 1996 Impala SS has 2 knock sensors.
Not sure about the Vettes.
As the desensitizing, the knock sensor could get you closer to detonation, has anyone tried to install a second LT4 knock sensor on the driver's side block to better capture that cylinder bank?
My 1996 Impala SS has 2 knock sensors.
Not sure about the Vettes.
As the desensitizing, the knock sensor could get you closer to detonation, has anyone tried to install a second LT4 knock sensor on the driver's side block to better capture that cylinder bank?
#14
Re: !!!read Here!!!(why You Shouldn't Use An Lt4 Km)
Originally posted by Dmx4ever3
Programmer
There is no possible way to "desensitize" a knock sensor (other than disconnecting it or shielding it from engine vibration). You can only alter the amount of timing retard that would normally occur in the software. This in NO WAY desensitizes a knock sensor. A different knock module filters more of the "noise" that could be interpreted as "false knock".
Programmer
There is no possible way to "desensitize" a knock sensor (other than disconnecting it or shielding it from engine vibration). You can only alter the amount of timing retard that would normally occur in the software. This in NO WAY desensitizes a knock sensor. A different knock module filters more of the "noise" that could be interpreted as "false knock".