Tuesday 23 May 2017


I can't even flirt with the idea of pretending I'm normal, but I'm sure that this isn't what normal people do with an afternoon off work:

That's an old oscilloscope, two speedometers and a multimeter on a workmate connected to a kit car which is on axle stands running at 120mph. Between the seat and the wheel arch you can just see the bamboo which operates the throttle.

This is the temporary connection onto the speed sensor.

My speedo had gone wrong in a very odd way and I found it hard to find the fault. My brother (when he's not drinking rum, stealing treasure and hitting people with belaying pins) is an aircraft engineer and he's much better at fault-finding than me. I told him about the ridiculous amount of time and effort I spent on the speedo fault. "There's an inductive sensor pointing at a toothed wheel on the driveshaft, there's the wiring up to the speedo, and there's the instrument itself. I've got a spare speedo and that showed the same fault. I tested the sensor and that's OK so there must be a fault with the loom"... My brother interjected "Or the driveshaft" and in a matter of moments had understood a fault that it took me a week to find. It seems that the magnetic reluctance is off-centre and the speed signal shifts phase with every turn of the wheel.

Richard "aconcentric reluctance" B


  1. Damn you guys are some nerds!

    1. Not as nerdy as when I program an Arduino to do a fast-fourier-transform on the speed sensor signal, average the frequency of the fundamental over 250ms or so and then run a nice clean oscillator at that frequency to feed into the speedo. That's fix it. And cement my reputation.