Jump to content
SAU Community

Recommended Posts


Oh, so the voltage being displayed is via a cheap volt meter wired in?

my suggestion, and I could be wrong and don't want to offend you, is that may be wired into the wrong pins. Don't take that as a negative or having a go, sometimes going back to the start / basics is what has to happen. The whole 2.4 volts thing has confused everyone except you, because you have seen it and did it. Everyone else's knowledge of a knock sensor doesn't match your experience, and that is why you are getting a bit cranky..

But a knock sensor is a simple thing, understood by many, that doesn't put out a voltage. So something is up, which has impacted this thread in a big way. Can you confirm what pins you connected to?

 

Also, proper terms seriously matter, you can't just use different terminology that may be conflicting or not make sense then tell people to just accept it and move on.

 

  • Like 1

Thank you for being decent this post ben.

I will check the pins again and will check the wires for any previous modifications.

Yes terminology. Im not great with it and does me no favours when explaining to people who know the exact terms.
Why i stated i wont be using it and that you guys are smart enough to decifer or re-correct me.

I do remember reading about it back in the day.
Just cant remember enough.

Ill post back up the pins there on.

You need an oscilloscope not a voltmetre to determine knock, a knock sensor behaves very similar to a microphone.

Like I said in a previous post, if you test the voltage on the knock sensor pin without the sensor connected you'll detect a voltage when using a stock ECU.

  • Like 1

Ok the thing is dosepipe. I know that.
I just couldnt get across that i cant and havent - and i have volts.
I remember reading that it was useless but id already wires it up.

The wires seem correct pins 23 and 24.
Off white almost cream wires as in the pic.

Geebs i havent dove in there for a while hecticly taped and messy. Needs a tidy up.
Ahahaha dont get sus i know theres wires everywhere i did most of em , and i just broke one with the tape lol. Easy fix.

Former owner did safc as standard install with gtr injectors when i bought it back in 2013. Ok for e85 hahaha.
There is also some hacking from b4 2.

What is thar brown plug with all the wires. Is it to do with the wire shielding?20190304_190925.jpeg20190304_193549.jpeg20190304_185301.jpeg20190304_191005.jpeg

This is the most pointless exercise ever, and i love it. 

I couldn't have even dreamed of anyone trying to measure knock with a jaycar multimeter.

Also love your 20/20 hindsight knowledge. Everything anyone posts you 'already knew'.

Good luck figuring it out!

  • Like 5
  • Haha 1

holy f**k Slap you did not take my advice.

the knock sensor is a microphone. it generates an AC wave. that is the sound of the motor. taller wave = louder. spacing of the waves = frequency. its not a DC voltage. a sound wave can have many many many frequencies. it is continuous, not discreet. you only care about noises in the frequency range the engine vibrates at when it knocks.

a decent ecu will filter that noise to a particular "band" of frequencies (say 6-7khz). the peak Amplitude of that signal is your knock level. it can be in volts, it can be in ADC counts. if your engine is not knocking it will be quiet in this range. LOW VOLTAGE. if it IS knocking it will be loud in this range HIGH VOLTAGE.

a shit ecu will not filter the frequency, which is ok if your sensor is resonant to your engine knock frequency. but for a wideband sensor.

the voltage you are reading will be the PEAK voltage seen by the ecu, and probably have a peak/hold circuit that gets reset after a delay. now for this to be useful, you need to know:

is your sensor wideband or resonant?

does your ecu do bandpass filtering on that signal?

the voltage you are looking at is an indicator of the VOLUME of the noise the knock sensor (microphone) is picking up. that is NOT usefull if you have the following:

wideband sensor with no bandpass filter (this will be all the engine noise)

this IS useful under the following:

wideband sensor with bandpass (this is knock frequency only)

resonant sensor with or without bandpass (if sensor is resonant to the frequency your engine knocks at)

if you want to know how LOUD the signal is, you can use a peak/hold multimeter. this will just tell you maximum volume, it could be at 100hz or 15khz. it won't tell you about knock.

if you want to know what FREQUENCY the knock is occuring at you will need a scope. you need both the frequency and amplitude of the signal to work out if its actually likely to be knock or just loud noises at frequencies you don't care about.

 

to understand what is happening you really need to plug the knock sensor into a usb soundcard and use some spectrogram/ real time fourier (frequency) graph software. something like this: https://digh.com/product/deka/

 

some free software I use is this which works great.

http://www.sdradio.eu/weaksignals/spectran.html

 

 

 

 

Edited by burn4005
  • Like 1

* Waits for "I know this* Response then reply asking about why there is voltage there *

I am not sure how this is taking so long to understand, especially with some smart people giving some smart answers.
I feel like you have backed yourself into a corner and now just want to attack everyone saying your wrong rather then admitting that you made a mistake and you now have been given some knowledge.

If you really want to get some information out of the knock sensors, something like the below may help:

Ebay Oscilloscope

  • Like 1
Quote

So you are saying the voltage is manipulated to bend in amplitude according to vibration causing it to become a frequency. Not changing volts.

Would the voltage displayed be the threshhold in wich the frequency works at.

why does the voltage drop on e85? Less noise? Less frequency needed? Magically? or im full of shit and it doesnt wich makes no sense to my eyes that have witnessed it ?

Anolog is as simple as a boost switch or bleed valve.
And i know it needs a digital tune to make the most. Geebz nearly any mod manualy controled or manipulated is anolog tuning.
The anolog gauges are just a unique trend . 

this is one of the dumbest f**king things I've ever read.

the only thing that might make sense here is you are actually Japanese and are running all your answers through google translate into english. 

the reason you are being attacked is you have been unable to convey any coherent concept whatsoever but are still somehow attempting to claim the intellectual high ground here.

Edited by burn4005
  • Like 1
  • Haha 1
24 minutes ago, burn4005 said:

the only thing that might make sense here is you are actually Japanese and are running all your answers through google translate into english.

Maybe you should run it back and forth through google translate between Japanese and English until it starts to make sense.

  • Haha 1
holy f**k Slap you did not take my advice.
the knock sensor is a microphone. it generates an AC wave. that is the sound of the motor. taller wave = louder. spacing of the waves = frequency. its not a DC voltage. a sound wave can have many many many frequencies. it is continuous, not discreet. you only care about noises in the frequency range the engine vibrates at when it knocks.
a decent ecu will filter that noise to a particular "band" of frequencies (say 6-7khz). the peak Amplitude of that signal is your knock level. it can be in volts, it can be in ADC counts. if your engine is not knocking it will be quiet in this range. LOW VOLTAGE. if it IS knocking it will be loud in this range HIGH VOLTAGE.
a shit ecu will not filter the frequency, which is ok if your sensor is resonant to your engine knock frequency. but for a wideband sensor.
the voltage you are reading will be the PEAK voltage seen by the ecu, and probably have a peak/hold circuit that gets reset after a delay. now for this to be useful, you need to know:
is your sensor wideband or resonant?
does your ecu do bandpass filtering on that signal?
the voltage you are looking at is an indicator of the VOLUME of the noise the knock sensor (microphone) is picking up. that is NOT usefull if you have the following:
wideband sensor with no bandpass filter (this will be all the engine noise)
this IS useful under the following:
wideband sensor with bandpass (this is knock frequency only)
resonant sensor with or without bandpass (if sensor is resonant to the frequency your engine knocks at)
if you want to know how LOUD the signal is, you can use a peak/hold multimeter. this will just tell you maximum volume, it could be at 100hz or 15khz. it won't tell you about knock.
if you want to know what FREQUENCY the knock is occuring at you will need a scope. you need both the frequency and amplitude of the signal to work out if its actually likely to be knock or just loud noises at frequencies you don't care about.
 
to understand what is happening you really need to plug the knock sensor into a usb soundcard and use some spectrogram/ real time fourier (frequency) graph software. something like this: https://digh.com/product/deka/
 
some free software I use is this which works great.
http://www.sdradio.eu/weaksignals/spectran.html
 
 
 
 
Voltage explained!

And the rest of you trolls spittin shid like your skyline preists can go lik an arse.

You couldnt answer it and thats that . You went on about everything else making me confused as shid.
And maybe when ur intelectual abilities extend beyond the bible of tuning and mechanics you might actualy understand what i said , where i was confused , why it was said, and why a lot of what you have posted is erelevent to not only the point but to me. Because i dont dyno tune or have sufficient consult equiptment.
As i stated i know it isnt usefull and asked why i see it. Wich atleast was answered by someone with intelect and wisdom.
Not a load of shid about its impossibleness.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



  • Similar Content

  • Latest Posts

    • Well, that's kinda the point. The calipers might interfere with the inside of the barrels 16" rims are only about 14" inside the barrels, which is ~350mm, and 334mm rotors only leave about 8mm outboard for the caliper before you get to 350, And.... that;s not gunna be enough. If the rims have a larger ID than that, you might sneak it in. I'd be putting a measuring stick inside the wheel and eyeballing the extra required for the caliper outboard of the rotor before committing to bolting it all on.
    • OK, so again it has been a bit of a break but it was around researching what had been done since I didn't have access to Neil's records and not everything is obvious without pulling stuff apart. Happily the guy who assembled the engine had kept reasonable records, so we now know the final spec is: Bottom end: Standard block and crank Ross 86.5mm forgies, 9:1 compression Spool forged rods Standard main bolts Oil pump Spool billet gears in standard housing Aeroflow extended and baffled sump Head Freshly rebuilt standard head with new 80lb valve springs Mild porting/port match Head oil feed restrictor VCT disabled Tighe 805C reground cams (255 duration, 8.93 lift)  Adjustable cam gears on inlet/exhaust Standard head bolts, gasket not confirmed but assumed MLS External 555cc Nismo injectors Z32 AFM Bosch 023 Intank fuel pump Garret 2871 (factory housings and manifold) Hypertune FFP plenum with standard throttle   Time to book in a trip to Unigroup
    • I forgot about my shiny new plates!
    • Well, apparently they do fit, however this wont be a problem if not because the car will be stationary while i do the suspension work. I was just going to use the 16's to roll the old girl around if I needed to. I just need to get the E90 back on the road first. Yes! I'm a believer! 🙌 So, I contacted them because the site kinda sucks and I was really confused about what I'd need. They put together a package for me and because I was spraying all the seat surfaces and not doing spot fixes I decided not to send them a headrest to colour match, I just used their colour on file (and it was spot on).  I got some heavy duty cleaner, 1L of colour, a small bottle of dye hardener and a small bottle of the dye top coat. I also got a spray gun as I needed a larger nozzle than the gun I had and it was only $40 extra. From memory the total was ~$450 ish. Its not cheap but the result is awesome. They did add repair bits and pieces to the quote originally and the cost came down significantly when I said I didn't need any repair products. I did it over a weekend. The only issues I had were my own; I forgot to mix the hardener into the dye two coats but I had enough dye for 2 more coats with the hardener. I also just used up all the dye because why not and i rushed the last coat which gave me some runs. Thankfully the runs are under the headrests. The gun pattern wasn't great, very round and would have been better if it was a line. It made it a little tricky to get consistent coverage and I think having done the extra coats probably helped conceal any coverage issues. I contacted them again a few months later so I could get our X5 done (who the f**k thought white leather was a good idea for a family car?!) and they said they had some training to do in Sydney and I could get a reduced rate on the leather fix in the X5 if I let them demo their product on our car. So I agreed. When I took Bec in the E39 to pick it up, I showed them the job I'd done in my car and they were all (students included) really impressed. Note that they said the runs I created could be fixed easily at the time with a brush or an air compressor gun. So, now with the two cars done I can absolutely recommend Colourlock.  I'll take pics of both interiors and create a new thread.
    • Power is fed to the ECU when the ignition switch is switched to IGN, at terminal 58. That same wire also connects to the ECCS relay to provide both the coil power and the contact side. When the ECU sees power at 58 it switches 16 to earth, which pulls the ECCS relay on, which feeds main power into the ECU and also to a bunch of other things. None of this is directly involved in the fuel pump - it just has to happen first. The ECU will pull terminal 18 to earth when it wants the fuel pump to run. This allows the fuel pump relay to pull in, which switches power on into the rest of the fuel pump control equipment. The fuel pump control regulator is controlled from terminal 104 on the ECU and is switched high or low depending on whether the ECU thinks the pump needs to run high or low. (I don't know which way around that is, and it really doesn't matter right now). The fuel pump control reg is really just a resistor that controls how the power through the pump goes to earth. Either straight to earth, or via the resistor. This part doesn't matter much to us today. The power to the fuel pump relay comes from one of the switched wires from the IGN switch and fusebox that is not shown off to the left of this page. That power runs the fuel pump relay coil and a number of other engine peripherals. Those peripherals don't really matter. All that matters is that there should be power available at the relay when the key is in the right position. At least - I think it's switched. If it's not switched, then power will be there all the time. Either way, if you don't have power there when you need it (ie, key on) then it won't work. The input-output switching side of the relay gains its power from a line similar (but not the same as) the one that feeds the ECU. SO I presume that is switched. Again, if there is not power there when you need it, then you have to look upstream. And... the upshot of all that? There is no "ground" at the fuel pump relay. Where you say: and say that pin 1 Black/Pink is ground, that is not true. The ECU trigger is AF73, is black/pink, and is the "ground". When the ECU says it is. The Blue/White wire is the "constant" 12V to power the relay's coil. And when I say "constant", I mean it may well only be on when the key is on. As I said above. So, when the ECU says not to be running the pump (which is any time after about 3s of switching on, with no crank signal or engine speed yet), then you should see 12V at both 1 and 2. Because the 12V will be all the way up to the ECU terminal 18, waiting to be switched to ground. When the ECU switches the fuel pump on, then AF73 should go to ~0V, having been switched to ground and the voltage drop now occurring over the relay coil. 3 & 5 are easy. 5 is the other "constant" 12V, that may or may not be constant but will very much want to be there when the key is on. Same as above. 3 goes to the pump. There should never be 12V visible at 3 unless the relay is pulled in. As to where the immobiliser might have been spliced into all this.... It will either have to be on wire AF70 or AF71, whichever is most accessible near the alarm. Given that all those wires run from the engine bay fusebox or the ECU, via the driver's area to the rear of the car, it could really be either. AF70 will be the same colour from the appropriate fuse all the way to the pump. If it has been cut and is dangling, you should be able to see that  in that area somewhere. Same with AF71.   You really should be able to force the pump to run. Just jump 12V onto AF72 and it should go. That will prove that the pump itself is willing to go along with you when you sort out the upstream. You really should be able to force the fuel pump relay on. Just short AF73 to earth when the key is on. If the pump runs, then the relay is fine, and all the power up to both inputs on the relay is fine. If it doesn't run (and given that you checked the relay itself actually works) then one or both of AF70 and AF71 are not bringing power to the game.
×
×
  • Create New...