Page 21 of 34 FirstFirst ... 11161718192021222324252631 ... LastLast
Results 301 to 315 of 509

Thread: 1997 F-Body ECM

  1. #301
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    To finish off regarding the input pre-amp/switch:

    This is no pre-amp at all. The part turns out to be a dual 4:1 analog mux with a common part number of 74hc4052. The two select inputs permit a single sensor to be selected and the analog forwarded to the knock filter module. With the PRU you can also select both channels at once. In this case the output sums the two inputs. AC input in the standard (left/right) configuration is attenuated to half the input value. Plan to update the schematic, it will remove a bunch of ?s

    -Tom

  2. #302
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Here is the revised schematic for the TSide PCM knock filter front end. Very happy to understand this circuit... in the case of Fbody, the whole mux arrangement could be replaced by a couple of resistors. Circuit seems to be used by Y&Bbody.
    Attached Files Attached Files

  3. #303
    Fuel Injected!
    Join Date
    Sep 2012
    Location
    Huntsville, AL
    Posts
    237
    Any chance you could make the font on that bigger? Great work!

  4. #304
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Quote Originally Posted by sherlock9c1 View Post
    Any chance you could make the font on that bigger? Great work!
    I tried a few things, but this free CAD package doesn't have a function to resize all the text. Short of editing the library and changing each text there is no easy path. I also tried to "print" the schematic to Dsize (that is what I use) PDF. There is no Dsize in the pull down. Not sure if I can help here, zooming in does give readable text in Acrobat...

  5. #305
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Looking at the knock module this morning. Started to characterize the workings of the analog level out (pin 2). I have built a test set with the Olds V6 module 16212229. This includes the connection to my sound card I wrote about the other day.

    I think starting with an 8KHZ tone makes sense, so I started there. The output with no input signal starts at about 0.25V offset. As the input level is increased from 0 to 25mV, I see a linear increase in the output to about 0.5V. From there through 750mV p-p input I see the analog level out step up. The increments are not linear. The last step is about 0.5V.

    It looks to me as if the knock module contains circuits that filter the input, pass it to an a/d converter, buffer the output with an op-amp. I can see the analog level go through about 24 steps. Probably some sort of r-2r ladder a/d.

    Knowing this all, I will follow the level out to understand how the Eside uses the result. Slow, but I am finally getting some where.

    -Tom

  6. #306
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,470
    Finally some breakthrough with testing. I am sure you will catch up quickly with the current test setup. I am not sure but it the knock module will be tuned for precise frequence range. All other is considered noise and is filtered. Changing the input frequency and the signal level will give more info, how it works.

    If you have a spare knock sensor you can thread it to a block of steel and tap it a hammer. It will definitely produce some signal that can be recorded and than playbacked later to knock filter.

    How hard will be to record sound of the knock sensor. I have some good quality sound recording equipment for mics that can be used.

  7. #307
    Fuel Injected!
    Join Date
    Jul 2019
    Location
    Orange, CA
    Posts
    757
    I'm sure you've probably seen this before, but here was some simulator testing another user did on MEMCAL knock modules back in 2011. These are pretty similar to ours, so maybe it's helpful in some way.

    https://pcmhacking.net/forums/viewtopic.php?f=11&t=1820
    1990 Corvette (Manual)
    1994 Corvette (Automatic)
    1995 Corvette (Manual)

  8. #308
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Quote Originally Posted by kur4o View Post
    Finally some breakthrough with testing. I am sure you will catch up quickly with the current test setup. I am not sure but it the knock module will be tuned for precise frequence range. All other is considered noise and is filtered. Changing the input frequency and the signal level will give more info, how it works.
    Yes, I think you are right. Also, I think that the knock is measured relative to the noise that the engine produces. Could be that additional noise added in "non=knock" frequency ranges would attenuate the effects of false knock. Time will tell.

    Quote Originally Posted by kur4o View Post
    If you have a spare knock sensor you can thread it to a block of steel and tap it a hammer. It will definitely produce some signal that can be recorded and than playbacked later to knock filter.
    A steel block would ring/vibrate at a frequency depending on it's shape and mass. Might not be the frequency the knock filter is looking for. With the PC sound card (or some other sine generator) we can create a pure sine wave with ease. It's frequency can be changed with a touch of the keyboard. I will need to learn more about PC sound to create noise and add a measured quantity of knock frequency sound to it. Damn another project...

    Quote Originally Posted by kur4o View Post
    How hard will be to record sound of the knock sensor. I have some good quality sound recording equipment for mics that can be used.
    Good quality sound recording equipment is probably a little out of my depth. I would never use a microphone other than a contact mic. In fact, that is what the knock sensor is, so why not just use that? The frequency response of the knock sensor is probably not flat... likely tuned to peak in the range that knock produces. These are all just guesses on my part.

    If you would try an "easy" recording method, I believe the schematic below should give you what you want (and I want too). Requires a audio jack with pigtail of the TRRS style. Pick off the S (sheath) and R (ring next to it). The R is ground and the S is microphone. Set a laptop up with sound recorder or some better software and test it out. CAUTION: I have not tired this.

    Once working, it would be nice to get some clean running at various RPM. 20 seconds of each is more than enough. Along with that, adding some bad gas and taking a run or two to catch knock would be the icing on the cake.

    Here is the schematic if you want to try... I forgot to add a value to the cap, I am suggesting a 4.7uf (2 - 10 should be OK) with a breakdown voltage of at least 10v (working voltage is 5v+-)

    EDIT: Error in schematic now fixed. YIPES

    KnockCorder.jpg
    Attached Images Attached Images
    Last edited by Tom H; 11-20-2020 at 11:38 PM.

  9. #309
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Quote Originally Posted by NomakeWan View Post
    I'm sure you've probably seen this before, but here was some simulator testing another user did on MEMCAL knock modules back in 2011. These are pretty similar to ours, so maybe it's helpful in some way.

    https://pcmhacking.net/forums/viewtopic.php?f=11&t=1820

    I am guilty of not going back over what has already been done. Thank you for the link...I will go back and re-read

  10. #310
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,470
    If I wire knock sensor output wire straigth to the recording device will it work. The device have built-in preamp and is very sensitive so low level signal won`t be an issue.
    I know the pcm supplies 5v to knock sensor. Is this only for checking presence of the knock sensor or is needed for the knock sensor output.

    I was looking at some tables for newer pcm that can control the knock input by software and there wer some khz mentioned there. Something between 5 and 8khz. I will look further and see what other settings are in there or some more hints of how the knock settings is controlled by software.

  11. #311
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Quote Originally Posted by kur4o View Post
    If I wire knock sensor output wire straigth to the recording device will it work. The device have built-in preamp and is very sensitive so low level signal won`t be an issue.
    I know the pcm supplies 5v to knock sensor. Is this only for checking presence of the knock sensor or is needed for the knock sensor output.

    I was looking at some tables for newer pcm that can control the knock input by software and there wer some khz mentioned there. Something between 5 and 8khz. I will look further and see what other settings are in there or some more hints of how the knock settings is controlled by software.
    I think the 5v bias is needed to center the input ~half way between the analog +10V and ground. Signal W/O bias will cause latchup of the knock chip. I suggest you AC couple the signal as in my diagram. The higher the impedance of the equipment, the smaller cap you can get away with. If you have high$ equipment that was intended for recording studio, the front end of the pre-amp might not be protected well. You might consider a pair of diodes to prevent the ac coupled signal from extending more than a drop +ve or -ve to ground. Laptops all have good protection in the front end, not so much a problem there. I think the knock sensor produces less than 1V p-p. Perhaps you could measure the AC with a good DVM?

    I think the 06/7 cars had a different way of checking for presence/working order of the knock system. I will get to that later...

    I find it easy to trigger the knock module. Interesting that it triggers and stays down for a time, then there is a time out if the signal is sustained. I need to be able to modulate the input signal. I don't know enough about the sound card to be able to write a function generator. Over the next while I will look around at example code to see if I can figure a way. Other than that, purchase of some test equipment(?)

  12. #312
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,470
    The device is set for condenser mics and can take a line input of music instruments. Not sure if it have some protection but will play it safe adding some components based on the graph. Worst case I will plug it in the laptop.

    I read that the knock sensor uses some piezocrystal that generates sound wave. Isn`t that what is used in turntables to get sound.

    I will disconnect the knock sensor from pcm and run the wire straight to the audio recording unit. I just remembered that I have a portable audio recorder that can take mic input. I might try that first to get some engine sound wave noise.

  13. #313
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Quote Originally Posted by kur4o View Post
    I read that the knock sensor uses some piezocrystal that generates sound wave. Isn`t that what is used in turntables to get sound.

    I will disconnect the knock sensor from pcm and run the wire straight to the audio recording unit. I just remembered that I have a portable audio recorder that can take mic input. I might try that first to get some engine sound wave noise.
    I only know enough about piezo electric devices to be a danger to myself. Piezo was used in very early turntables to convert needle movement to electrical signal. The force used to move the crystal was the issue, causing wear on they vinyl. Turntables all moved to moving coil type pickups. When piezo is used as a microphone, the bending of the crystal creates the electrical in the same manner. When a voltage is imposed on the crystal, it bends. I am not sure if the bias our PCMs put on the crystal is used to tune the response. Can anyone tell us this?? Interested to know the DC resistance of the knock sensor if you could measure it.

    I think disconnecting the knock sensor might put your ride at risk. For myself, it would be best if you just use a small aluminum electrolytic cap in line to isolate the DC and record the AC. Can't wait to see how you recordings sound... and their effect on the knock sensor filter module.

    I am working on finding a way to characterize the modules. Much to read, much to do. -Tom

  14. #314
    Fuel Injected!
    Join Date
    Jan 2019
    Location
    Canada
    Posts
    477
    Something of a false start working through the function of the knock filter. I tried to use the PC as a "sound source" for testing. To date, this hasn't worked well. I tested first using the standard windows Beep() function. This is very limited and while I could produce tones that would activate the noise channel and knock outputs characterizing the module will need a better source. I tried using MIDI and was able to make both level and frequency changes. This is also limited because it won't produce sine waves, or at least media player won't. I will give the PC one last try before moving on to another way to produce sound. I am looking into the API windows uses to produce sound but it may be too complicated for me... in which case I need to build/buy equipment.

    I have confirmed the pinout given earlier and the basic function. The analog output (which I now refer to as noise channel) is used to detect functioning of the knock system from sensor through wiring to knock module. I believe that the filtered input is converted to an envelope ( it is AM and this step is just demodulation). The knock filter chip looks for rapid increases in the envelope to detect knock. More work is needed here.

    In terms of tune, it is likely that if an engine is bored oversize the resonant frequency would change. I am still unsure of what factors affect the resonant frequency. Here are my thoughts of possible candidates, more than one may be involved

    - Cylinder swept volume
    - Cylinder swept volume in the knock window phase of rotation
    - Mass of the iron that creates the cylinder
    - Resonant frequency of the block's iron
    - ???

    I believe that the resonant frequency that the knock module responds to can be changed by altering resistor values on the knock module. I have not yet determined the type of filter that is involved.
    The second tune-able characteristic may be the delta level of the envelope. If confirmed, this may also be tuned by altering resistance on the module.

    Ideal for testing this would be a dual channel function generator with an input to permit amplitude modulation of one output. Let's see what I can turn up on Kijiji...
    Last edited by Tom H; 12-07-2020 at 03:50 PM.

  15. #315
    Fuel Injected!
    Join Date
    Jul 2019
    Location
    Orange, CA
    Posts
    757
    For knock filter theory, perhaps this document from 1995 can shed some light?

    EDIT: I've also discovered most of the meanings behind the spreadsheet I had posted earlier in the thread: http://www.gearhead-efi.com/Fuel-Inj...ll=1#post85161

    I found a Delphi patent for a programmable knock filter here: https://patentimages.storage.googlea.../US5594649.pdf

    The phrasing in this patent document perfectly jives with the spreadsheet. As such, the descriptions and units are as follows:

    SBST = Software-based Self-test? Internal GM test routine?

    FIRST STAGE (Bandpass Filter)
    FREQ = Frequency (kHz)
    GAIN = Gain (multiplier)
    Q = Q Factor

    COMP GAIN = Compressor Gain (ratio of input power to output power beyond threshold)

    LPCF = Low-pass Filter Cutoff Frequency (Hz)

    NOISE CHANNEL
    GAIN = Gain (multiplier)
    OFFSET = DC Offset Voltage (mV)
    ITC = Increasing Time Constant (ms)
    DTC = Decreasing Time Constant (ms)
    IGN = ???
    AOT = ???

    ID RES K Ohms = Resistance value used to ID filter on ID pin (kOhms)
    Attached Files Attached Files
    Last edited by NomakeWan; 12-08-2020 at 12:49 PM.
    1990 Corvette (Manual)
    1994 Corvette (Automatic)
    1995 Corvette (Manual)

Similar Threads

  1. 94-95 LT1 $EE Y-body vs. F/B-body PCM differences
    By johnny_b in forum GM EFI Systems
    Replies: 5
    Last Post: 01-15-2023, 02:41 PM
  2. Tuner Pro XDF 1999-2000 F Body + Y Body
    By john h in forum OBDII Tuning
    Replies: 33
    Last Post: 02-02-2020, 11:12 PM
  3. Replies: 31
    Last Post: 09-20-2018, 06:00 AM
  4. F-body engine install to B-body
    By serge_an in forum GM EFI Systems
    Replies: 4
    Last Post: 09-22-2016, 02:51 PM
  5. 95 F-body Fuel Pump with 95 B-Body Engine/Tank
    By EPS3 in forum GM EFI Systems
    Replies: 7
    Last Post: 09-19-2016, 02:40 PM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •