Tuning Guru's: timing when street tuning?
01-14-2004, 01:54 PM
#11
1.0 BAR
Join Date: Apr 2003
Posts: 288
Re:Tuning Guru's: timing when street tuning?
you would be surprised at how much of a difference on AFRs your timing can make. Don't underestimate this.
You really need accurate datalogging of RPM/MAP/wideband AFR/EGT. If you can know all of those variables at once, you can accurately tune the vehicle. Tall order...
You really need accurate datalogging of RPM/MAP/wideband AFR/EGT. If you can know all of those variables at once, you can accurately tune the vehicle. Tall order...
01-14-2004, 02:54 PM
#12
3.0 BAR
Join Date: Apr 2003
Posts: 2,260
Re:Tuning Guru's: timing when street tuning?
Originally Posted by blundar
you would be surprised at how much of a difference on AFRs your timing can make. Don't underestimate this.
Hackish: Can you throw a little mini-how to, or point me in the direction of trying that out? Sounds like an interesting idea. I would like to try this out.
01-14-2004, 05:59 PM
#13
1.0 BAR
Join Date: May 2003
Posts: 267
Re:Tuning Guru's: timing when street tuning?
DIY Detonation Stethoscope
http://www.autospeed.com/cms/article.html?&A=0353
on ebay, search for "spy ear". you can get them for about 7 dollars shipped.
http://www.autospeed.com/cms/article.html?&A=0353
on ebay, search for "spy ear". you can get them for about 7 dollars shipped.
01-14-2004, 06:31 PM
#14
0.0 BAR
Join Date: Jun 2003
Posts: 0
Re:Tuning Guru's: timing when street tuning?
That stethoscope is definitely a good idea but wouldn't the voltage from the knock sensor tell you the same thing or is that detonation to faint for the knock sensor?
Does Turboedit display/monitor knock voltage? I'm new to the complete stand alone systems so I'm trying to learn how to not kill my Civic. My friend is in the Navy so he let me borrow his EMS for a long while so I'm trying to learn how to use it, my eyes are bleeding from reading all the PDF files and all.
P.S. Mad props on Turboedit, that is truely badass
Does Turboedit display/monitor knock voltage? I'm new to the complete stand alone systems so I'm trying to learn how to not kill my Civic. My friend is in the Navy so he let me borrow his EMS for a long while so I'm trying to learn how to use it, my eyes are bleeding from reading all the PDF files and all.
P.S. Mad props on Turboedit, that is truely badass
01-14-2004, 07:22 PM
#18
0.0 BAR
Join Date: Jun 2003
Posts: 0
Re:Tuning Guru's: timing when street tuning?
I didn't know there wasn't a knock sensor, the EMS is for a p28. I read about the stethoscope and I really would like to try it but I didn't know if it would be more accurate to use the voltage
01-14-2004, 07:38 PM
#19
1.0 BAR
Join Date: Apr 2003
Posts: 288
Re:Tuning Guru's: timing when street tuning?
knock sensor voltage is snake oil. False security.
Not a simple explanation, but...
Knock sensors are essentially a microphone.
Microphones produce voltage in response to a pressure wave.
We call repeated pressure waves "sound." The amount of force the pressure wave carries normally registers to our hearing as "volume." The number of pressure waves / second that register normally registers to us as "pitch."
Changes in RPM generally create a change in "pitch." Knocking is a very loud event, and generaly produces a change in "volume." Also, knock occurs at a different time than most engine noise, so it disturbs the rhythm of the motor.
Voltage coming out of a knock sensor is NOT directly proportional to either "volume" or "pitch" but rather a combination of the two.
The reason that monitoring only sensor voltage is somewhat oversimplified and an overall poor knock detection strategy is that it is a very coarse measurement in the sense that individual ignition events are not being analyzed as much as the overall engine noise - voltage is more of an average of the noise*frequency in a period of time than anything else. At higher RPMs, knock sensor voltage will increase due to engine noise causing the difference between ambient noise threshold and a knock event to be small enough that subtle preignition events are missed.
Measuring the knock sensor voltage will detect extreme detonation events, as the spike produced by a LOUD ping will register clearly on the voltage output. Measuring the knock sensor voltage will not detect more subtle preignition events as they get mixed in with ambient engine noise too much. Knock sensor voltage is much more useful and relaible for LOW RPM knock detection than high RPM.
Knock detection is an area of current research in the engineering community. Current advanced approaches generally use DSPs coupled to the ECU to do spectrum analysis of the noise coming out of a knock sensor, focusing on the compression stroke for each cylinder so that as much engine noise can be ignored totaly solely based on the time at which it occurs. Ion sensing systems, genetic algorthims and non-Fourier transform DSP solutions have much promise.
I don't know how the J+S Safeguard devices work, but I imagine that they are somewhat more advanced than a simple voltage threshold / analog filter design.
Not a simple explanation, but...
Knock sensors are essentially a microphone.
Microphones produce voltage in response to a pressure wave.
We call repeated pressure waves "sound." The amount of force the pressure wave carries normally registers to our hearing as "volume." The number of pressure waves / second that register normally registers to us as "pitch."
Changes in RPM generally create a change in "pitch." Knocking is a very loud event, and generaly produces a change in "volume." Also, knock occurs at a different time than most engine noise, so it disturbs the rhythm of the motor.
Voltage coming out of a knock sensor is NOT directly proportional to either "volume" or "pitch" but rather a combination of the two.
The reason that monitoring only sensor voltage is somewhat oversimplified and an overall poor knock detection strategy is that it is a very coarse measurement in the sense that individual ignition events are not being analyzed as much as the overall engine noise - voltage is more of an average of the noise*frequency in a period of time than anything else. At higher RPMs, knock sensor voltage will increase due to engine noise causing the difference between ambient noise threshold and a knock event to be small enough that subtle preignition events are missed.
Measuring the knock sensor voltage will detect extreme detonation events, as the spike produced by a LOUD ping will register clearly on the voltage output. Measuring the knock sensor voltage will not detect more subtle preignition events as they get mixed in with ambient engine noise too much. Knock sensor voltage is much more useful and relaible for LOW RPM knock detection than high RPM.
Knock detection is an area of current research in the engineering community. Current advanced approaches generally use DSPs coupled to the ECU to do spectrum analysis of the noise coming out of a knock sensor, focusing on the compression stroke for each cylinder so that as much engine noise can be ignored totaly solely based on the time at which it occurs. Ion sensing systems, genetic algorthims and non-Fourier transform DSP solutions have much promise.
I don't know how the J+S Safeguard devices work, but I imagine that they are somewhat more advanced than a simple voltage threshold / analog filter design.
