Author Topic: splicing knock sensor  (Read 11489 times)

Offline jrtuning_ben

  • Jr. Member
  • **
  • Posts: 44
  • BHP: 2
splicing knock sensor
« on: January 03, 2008, 05:58:05 am »
i have recently bought a a ks3 knock detector/alarm. am i able to use the one knock sensor (a lovely new bosch one) for both the ks3 AND the vems? or would i ideally need two sensors?

Offline rob@vems.co.uk

  • Hero Member
  • *****
  • Posts: 3115
  • BHP: 49
    • VEMS Forum
Re: splicing knock sensor
« Reply #1 on: January 03, 2008, 08:23:06 am »
Tricky one that, on one side of the coin you have the fact that the sensor is located in the correct position to pick up detonation noise across all four cylinders (as decided by the Nissan designers) - which means you'd be best placed to use that sensor.
On the other side I can't be sure that sharing the sensor won't degrade the signal.  As the knock chip we use is passive with a high impedence I'd think it would be okay - but that would be for my use on my car where I'm happy with the risks.

Offline jrtuning_ben

  • Jr. Member
  • **
  • Posts: 44
  • BHP: 2
Re: splicing knock sensor
« Reply #2 on: January 03, 2008, 08:48:17 pm »
ok dude, ill splice them both onto the same connector and see how it goes  :)

i take it you cant stack the sensors? ???

Offline rob@vems.co.uk

  • Hero Member
  • *****
  • Posts: 3115
  • BHP: 49
    • VEMS Forum
Re: splicing knock sensor
« Reply #3 on: January 03, 2008, 09:16:37 pm »
I seriously doubt it.

First off I'd start with the KS3 on its own, then when you're reaching a stable state of tune you'd want to add the complication of the knock sensor system on the VEMS.

Offline jrtuning_ben

  • Jr. Member
  • **
  • Posts: 44
  • BHP: 2
Re: splicing knock sensor
« Reply #4 on: January 03, 2008, 11:04:01 pm »
ok mate :)

unsure yet if ill map it myself...fancy having a go but im struggling to understand it all :( leaving off complications to start with sounds like a good idea lol

Offline rob@vems.co.uk

  • Hero Member
  • *****
  • Posts: 3115
  • BHP: 49
    • VEMS Forum
Re: splicing knock sensor
« Reply #5 on: January 03, 2008, 11:26:38 pm »
Yeah, its a steep learning curve, but as you get the ecu setup and progress through the installation you'll begin to see why certain features are there and understand why you need them and how to use them.

Its best to try and eat an Elephant a bite at a time ;)

Rob

Offline cliffb75

  • Global Moderator
  • Full Member
  • *****
  • Posts: 167
  • BHP: 10
Re: splicing knock sensor
« Reply #6 on: January 05, 2008, 08:38:47 pm »
Tricky one that, on one side of the coin you have the fact that the sensor is located in the correct position to pick up detonation noise across all four cylinders (as decided by the Nissan designers) - which means you'd be best placed to use that sensor.
On the other side I can't be sure that sharing the sensor won't degrade the signal.  As the knock chip we use is passive with a high impedence I'd think it would be okay - but that would be for my use on my car where I'm happy with the risks.

I'd pretty much agree with that. Your biggest problem is likely to be with the signal being dragged down, hence signal to noise ratio getting worse. However, if you're able to adjust both systems, then you should be able to compensate for it.

ok dude, ill splice them both onto the same connector and see how it goes  :)

i take it you cant stack the sensors? ???

Stacking the sensors is an interesting idea. At first you'd say its a no no, but thinking about it, as long as there is a bit of space between them (a washer) so that they don't foul each other, they might work OK. I guess the problem would be potential lateral vibration of the bolt. But thats likely (but not guaranteed) to be at a different frequency, so may not be an issue if you've chosen correctly.

Offline Jorgen

  • Global Moderator
  • Jr. Member
  • *****
  • Posts: 59
  • BHP: 500
Re: splicing knock sensor
« Reply #7 on: January 06, 2008, 08:50:00 pm »
Hi,

The knock sensor ground in our ECU is connected to the sensor ground inside the ECU and it is possible to create a ground loop or to inject ground noise in the ECU through the knock ground. This can affect other sensors, even the important trigger sensors.

If I were to use one sensor for both boxes i would try to use a capacitor in series with the knock ground, it should not affect the signal quality and it will prevent any noise from affecting other sensors connected to the ECU.

Jörgen Karlsson
Gothenburg

Offline jrtuning_ben

  • Jr. Member
  • **
  • Posts: 44
  • BHP: 2
Re: splicing knock sensor
« Reply #8 on: January 06, 2008, 10:16:50 pm »
the ks3 gain can be adjusted so there isnt a problem there - is the same possible on the vems? (i havent got mine yet so cant check :()

can the stacking thing be checked? id try but ive no idea what im doing :( from what youre saying cliff it sounds feasible

Offline cliffb75

  • Global Moderator
  • Full Member
  • *****
  • Posts: 167
  • BHP: 10
Re: splicing knock sensor
« Reply #9 on: January 07, 2008, 09:23:25 pm »
the ks3 gain can be adjusted so there isnt a problem there - is the same possible on the vems? (i havent got mine yet so cant check :()

can the stacking thing be checked? id try but ive no idea what im doing :( from what youre saying cliff it sounds feasible

Sure it can - stack them up and connect the 2 sensors to a scope - first run the engine up in a no knock condition (safe ignition) and check that there is a sensible base signal - nothing that looks like det. Then get the thing detting and check the signals - if they are the same, and you can see the det , you have a winner. :) Anyone got an engine and scope we can try it on??

For info, when using a scope for looking at knock sensors, you will be able to tell the difference betwen base engine noise and det by the regularity. Base noise happens every cycle, so in term sof time its a regular pttern, whereas det is semi-random. We used an anaolgue scope with an infinately adjustable time base, and I then tweak that so that I end up with the signal just slowly scrolling past when the engine is running at a fixed speed. Its then very easy to see the normal noise pattern, and the det 'spikes' which appear additionally from time to time.