Trigger wheels and rev limits

Our trigger wheel testbench allows us to spin wheels up to speeds far in excess of those that would ever be seen in the real world.  Larger trigger wheels present an extremely strong signal so here we test our smallest wheels.  As we will see, even in a "worst case" scenario of a small wheel that has been poorly mounted the ECU has no problems decoding the signal.  Our experience is provided that the wheel is mounted centrally and the sensor mounted 0.5-1mm problems are rare and those that we do see are usually due to wiring or interference issues.

Here is a 3 inch trigger wheel spinning at a shade over 10000 rpm, the missing tooth can clearly be seen in the waveform towards the left of the plot.

The same wheel, this time at 26000 rpm.  Here the slight eccentricity in the way that this wheel has been mounted on its mandrel is visible - the peaks on the right of the missing tooth are a shade larger than the ones on the left.

As counter-intuitive as it sounds the toughest thing for a sensor to pick up is not the high speeds but the really slow speeds, this time the wheel is spinning at just 18rpm, the signal is still there but where it was generating 100 volts peak to peak whilst spinning at 10000rpm at this slow speed it's just 3 volts.  As we will see this is still plenty for the ECU.

The signal from the sensor is processed by the ECU into a digital signal.  Each ECU approaches this differently but one of the cheapest and most simple ways is to use an off the shelf LM1815 chip.  Most ECUs use something rather more sophisticated than the LM1815 so it represents an excellent test. 

The trace below is a signal from a wheel spinning at 10rpm after it has been  processed by the chip.  Each spike is a tooth and the missing tooth is clearly visible on the left hand side of the image.

The LM1815 has a fixed duration output pulse so at 26000rpm the spikes are much broader in relation to the gaps between them but as we can see even a $2 chip like the LM1815 has no problem with the signal at this speed.

Perhaps the ultimate test of sensors and wheels is our tiny 2.5" wheel.  Just to make it even more difficult for the chip in the cases below we have mounted the wheel as badly off center as we would ever like to see in the real world.  At a dawdling 10rpm the LM1815 has quite remarkably been able to decode the wheel.

At 11000rpm things get easier.  Notice how the waveform of the missing tooth differs from the one at the very top of the page, it looks as though the tooth and the adjacent missing tooth have been blurred into one.  This happens because with teeth just 2.5mm wide the sensor can see more than one tooth at a time, the signal is still strong so decoding the signal is not a problem, this effect can be minimised by running the sensor closer to the trigger wheel.

With our test rig cranked up to maximum we can get up to speeds in excess of 28000rpm with the 2.5" wheel:

As we can see the chip is still managing to decode the signal but this is about as far as we can push is on a a wheel this small.