How does a wireless cadence sensor work

[Fleming]

Does anyone know the ins and outs on how these sensors work? Specifically the reed switch based units with a magnet on the crank and how their battery systems work?

The reason I ask is the company I work for builds automated machines and we have a need for a compact wireless sensor to monitor rotation. We couldnt find a industrial sensor small enough but being a cyclist I had the idea to use a cadence sensor. The battery life was a concern so we went and bought one and set up a test with a motor and the sensor and have had it running at 120rpm 24/7 for around 3 weeks now. We wrote a basic app on a PC to log the the signal and battery life. We are a little confused as the battery life the sensor reports seems to fluctuate a lot. It sometimes drops as low as 37% but then goes back to 100% later. Mostly it sits between 80% and 100%.

The sensor uses a standard CR2032 coin cell that isnt rechargeable. My theory is the sensor has a coil that generates a current each time the magnet passes and this charges a small internal battery. During use the sensor uses this battery and this is the battery level that is reported. The CR2023 battery is only there to power the sensor during sleep mode when not in use. Can anyone confirm this? The sensor we bought is a Giant 2 in one unit.

[Pieter-za]

Won't an accelerometer based speed sensor be better?  (Or an MR Sensor)

The V2 Garmin Speed Sensor can also cache measurements to a certain extend. 

Edited September 11, 2023 by Pieter-za

[Fleming]

22 minutes ago, Pieter-za said:

Won't an accelerometer based speed sensor be better?  (Or an MR Sensor)

The V2 Garmin Speed Sensor can also cache measurements to a certain extend. 

The machine is such that the shaft we want to monitor is mounted to a assembly that rotates around a second shaft. As such the sensor will be rotating around a different center to that of the shaft we want to measure. Im pretty sure a accelerometer based sensor would get confused.

Im pretty happy with the performance of the reed switch sensor, just confused by the battery life it reports. 

[RobertWhitehead]

It works without wires . 

 

But seriously: Does the sensor not work on the same principle as a inverter? Hear my logic, an inverter reads the demand and allocates a portion of the battery power to the required load but if the load is reduced i.e. the kettle has stopped boiling it then releases the allocated power and displays the actual value? Am I giving too much credit and or making the sensor smarter than it actually is . 

[Fleming]

20 hours ago, RobertWhitehead said:

It works without wires . 

 

But seriously: Does the sensor not work on the same principle as a inverter? Hear my logic, an inverter reads the demand and allocates a portion of the battery power to the required load but if the load is reduced i.e. the kettle has stopped boiling it then releases the allocated power and displays the actual value? Am I giving too much credit and or making the sensor smarter than it actually is . 

I honestly have no idea. Im a mechanical person, my knowledge of such things is quite limited but I wouldnt expect the level to reach 100% after so much use.

[The Ouzo]

are you not perhaps logging the wrong metric ?

[RobertWhitehead]

6 minutes ago, Fleming said:

I honestly have no idea. Im a mechanical person, my knowledge of such things is quite limited but I wouldnt expect the level to reach 100% after so much use.

I used to have a cadence sensor (wahoo) on my indoor trainer which is similar in the sense that it would measure one thing and that things battery (also CR2032) would last for months on end 

[Fleming]

34 minutes ago, Fleming said:

I honestly have no idea. Im a mechanical person, my knowledge of such things is quite limited but I wouldnt expect the level to reach 100% after so much use.

Its possible. I didnt write the app myself. Ive spoken to the guy who wrote the app for me and he is confidant he is doing it correctly. The sensor only reports 3 things, wheel rotation, crank rotation and battery life.

[The Ouzo]

1 minute ago, Fleming said:

Its possible. I didnt write the app myself. Ive spoken to the guy who wrote the app for me and he is confidant he is doing it correctly. The sensor only reports 3 things, wheel rotation, crank rotation and battery life.

if its a cadence sensor then how is it reporting wheel rotation ?

[Fleming]

Its a 2 in 1 sensor. We just ignoring the wheel sensor.

https://online-bicycle-shop.co.za/products/giantant2in1speedcadencesensor

[TheoG]

23 hours ago, Fleming said:

Does anyone know the ins and outs on how these sensors work? Specifically the reed switch based units with a magnet on the crank and how their battery systems work?

The reason I ask is the company I work for builds automated machines and we have a need for a compact wireless sensor to monitor rotation. We couldnt find a industrial sensor small enough but being a cyclist I had the idea to use a cadence sensor. The battery life was a concern so we went and bought one and set up a test with a motor and the sensor and have had it running at 120rpm 24/7 for around 3 weeks now. We wrote a basic app on a PC to log the the signal and battery life. We are a little confused as the battery life the sensor reports seems to fluctuate a lot. It sometimes drops as low as 37% but then goes back to 100% later. Mostly it sits between 80% and 100%.

The sensor uses a standard CR2032 coin cell that isnt rechargeable. My theory is the sensor has a coil that generates a current each time the magnet passes and this charges a small internal battery. During use the sensor uses this battery and this is the battery level that is reported. The CR2023 battery is only there to power the sensor during sleep mode when not in use. Can anyone confirm this? The sensor we bought is a Giant 2 in one unit.

I'm using Polar cadence sensors for a long time now and they eventually do go flat.  I dont think there is any charging benefit from the magnet passing the sensor, it would be very cool if it was the case though.

The batteries could then potentially last "for ever".  Maybe someone with electronic background can shed some light on this for us.

[DieselnDust]

 A reed switch sensor works but allocating a portion of the voltage to the switch circuit. Everytime the magnet closes the circuit it registers a voltage and the duration that it took to rise from 0.0V to the measured voltage an drop off again. As rotational speed rises the time the circuit is closed decreases and that is reported as an increase in RPM. You need to give it some data through like a radius to the surfaces where the angular velocity is changing.

[Shebeen]

4 hours ago, Fleming said:

Its a 2 in 1 sensor. We just ignoring the wheel sensor.

https://online-bicycle-shop.co.za/products/giantant2in1speedcadencesensor

https://hotspotcycles.co.za/product/giant-2-1-digital-sensor/

 

 

So there are 2 sensors, one transmitter.

the BLE/ANT+ rf radio will be powered by the CR2023 battery.

There will be no charging/energy trasfer from the rotating magnet.

 

 

The battery level fluctation might be confused as you are using this continuously?

If you're concerned about long term battery life, then measure the current draw directly on the battery. pretty sure you could get a datasheet on cr2032 performance.

[Shebeen]

1 hour ago, DieselnDust said:

 A reed switch sensor works but allocating a portion of the voltage to the switch circuit. Everytime the magnet closes the circuit it registers a voltage and the duration that it took to rise from 0.0V to the measured voltage an drop off again. As rotational speed rises the time the circuit is closed decreases and that is reported as an increase in RPM. You need to give it some data through like a radius to the surfaces where the angular velocity is changing.

Not likely. It will just register a time between the trigger of the rising (or falling) edge of what is a very low duty cycle square wave signal. the computer will do the maths for the rest (which maybe in this little box as shown)

[DieselnDust]

46 minutes ago, Shebeen said:

Not likely. It will just register a time between the trigger of the rising (or falling) edge of what is a very low duty cycle square wave signal. the computer will do the maths for the rest (which maybe in this little box as shown)

well yes its pretty obvious the little computer does the math, the switch itself is just opening and closing to register a voltage (a change in signal) 

There's a reason why i added the word "Sensor" after "reed switch"

simple bike computers works this way. There is no battery in the switch , it runs everything off the 3V CR2032 in the head unit.

More expensive units may use a hall effect sensor that will require constant current where the magnet interrupts the electrical field and here magnetic shielding is required to isolate the circuit from outside unintended influences. These are typically found on higher speed systems like the tachometer for a 2-stroke engine