Project NSX: Part 19 – Designing a Crank Position Sensor and Knock Control

NSX Custom Hall Effect Crank Position Sensor and Tone Wheel We design a custom tone ring and hall-effect sensor mount for our Motec M150 to more accurately record the crankshaft position and control ignition timing at high RPM.  We also upgrade to a wideband knock sensor for a better failsafe against knock.

NSX Crank Position SensorOur custom crank position sensor journey started from the desire to improve an existing design that mounts a crank position sensor onto the oil pump housing.

NSX Billet Crank Position Sensor CADOur goal was to design a one-piece billet sensor mount that would better fit our custom dry sump oiling system.

NSX Crank Position sensor vs 3D printed test pieceThe original mount (left) was made from bent metal and a machined sensor holder.  We 3D printed test pieces to refine our design.

NSX billet crank position sensor machining.Once we were happy with our design, we had it machined.

NSX crank position sensor designMany revisions resulted in the final design that we were happy with.  However, we had a change of plans and decided to create a crank position sensor that mounted a hall-effect sensor to read the teeth on a tone ring that bolts to the front of an ATI pulley.

8 comments

  1. Back in my OEM engine calibration days, we referred to the time period where knock is likely to occur as the gate period. This can be determined by the crank angle and firing order. Knock should only typically occur during a small range of crank rotation for any given cylinder. We utilized 4 and 8 channel oscilloscopes to view and measure the raw knock sensor output signal, and then match it up against what we heard. A good microphone and a quality set of headphones go a long way. If a ‘knock’ waveform was detected for a particular cylinder, but was outside the gate period, it was largely ignored. We also calibrated the ECU to ignore these signals so it does not interfere with the knock control learning logic. Knock control system calibration on a modern Japanese V8 engine with 4 knock sensors could take 4-6 months. So basically, you’d run the engine at medium to high loads, advance the timing till a determined level of knock was achieved, and then analyze the data. After 4-6 months of knock testing, we’d do a HP check, confirm numbers, and then disassemble/inspect the engine. It was usually one of the last calibrations performed as any mechanical change or addition to the engine would change the harmonics, and require re-testing.

    FYI, most OEM engines are calibrated to run at a certain/safe level of knock. It’s perfectly fine in most situations. You’d be surprised at how durable most engines are… well the one’s I worked on anyway. 🙂

    1. I also have some past excursions in engine dev; we supplemented knock sensors with individual cylinder pressure sensors (also with an O-scope during dev). These were big engines though, and I think I remember the pressure sensors were a permanent part of the control system feedback in the same way as the knock sensors. Crank angle gating is definitely a normal part of knock strategy.

  2. These are the articles I hope to see anytime I click on MotoIQ, Thanks!!
    Hey one of the best project Supra articles has the wrong pictures in it- it’s an article about the cylinder head, but with all bodywork photos!!!

    Mike I also wanted your opinion on Bosch 4.9 wide band sensor placement in a turbocharged 4-500 horse flex fuel engine-
    Thanks guys sorry not fully on topic

Leave a Reply

Your email address will not be published. Required fields are marked *

*
*