Knock Sensor Installation Location

Great webinar! I wish I was able to catch it live. I have three questions:

1. Is there any reason you cannot simply piggyback the audio knock sensor on to the existing ECU knock sensor using a longer bolt? Is the loss of surface area connection to the block that significant that it would affect the output signal more than say selecting a sub optimal location over this option?

2. You mentioned that it's not satisfactory to electrically parallel your audio amplifier/ signal processor with an existing ECU knock sensor. However knock piezo sensor impedances tend to be in the 1-4 Kohm region so an audio amplifier with orders of magnitude higher input impedance (100K to 1M ohm) should not effect the signal voltage going into the OEM or aftermarket ECU by anything significant if you "shared" the signal. Have you noticed the signal degrade noticeably when you have tried this approach? I would have thought that the ECUs would utilise some form of adaptive thresholding/automatic gain control so that the system is relatively immune to changes in absolute signal level changes given how much they would need to adapt to the changing audio outputs an engine would exhibit as it deteriorates over time.

3. I've seen some people tune by disabling the closed loop knock control so that they can actually hear the knock without the ECU intervening and pulling timing and adding fuel. This way they can tune without the ECU making it harder to actually tune up to knock limited MBT. They wish to tune as close to optimally in open loop before turning on the feedback/ECU protection. So in the case where you have the closed loop knock control turned off anyway, couldn't you just remove the ECU knock sensor and substitute the audio one in its place for the open loop tuning phase and then just revert back to the ECU sensor when you turn the closed loop knock control on again? Or is tuning this way just an overall bad idea?

Hi David, glad you enjoyed the webinar! I've answered your questions below:

1. this is actually a topic I meant to discuss and it slipped my mind. I'll start by stating this is not something I've done so I can only comment from a theoretical approach. I'd think that this could be effective however I am a little worried about the loss of surface contact between the sensor and the block. The base of the sensor is quite a large area where as the top where the bolt tightens down is much smaller in diameter and this is the area that would contact the second sensor. Ultimately it's worth testing to see how well it does work.

2. I've personally tired sharing the signal on a factory ECU many years ago and straight away ended up with a DTC. I can't say it'll be problematic with every application but I haven't rushed back to trying this since. For a professional tuner tapping into the OE signal is also more time consuming than adding an additional sensor just for tuning. If you're tuning just your own car then this technique may make more sense.

3. It's important top understand that knock control is reactive, not predictive which means knock must occur in order for the ECU to do something about it. For this reason it's not necessary to turn knock control off during tuning. While yes, you could replace the stock sensor during tuning if the knock control was disabled, the reason i like to run an additional sensor is because we can then validate the knock control system is working. For an aftermarket ECU this is an essential step in setting the knock control frequency and threshold too.