Pedal sensor

[Nigelk]

Hi

Can anyone please tell me how the pedal sensor works on the bikes, is it just a simple
switch, ie would just joining the wires complete the circuit or is it a signal that is send back to the controller when activated.

Thanks

 

[neptune]

OK, I will jump in and expect correction if I am wrong. Basically it is a hall sensor [google it ]. Usually there are 3 wires , a positive 5 volt supply, a negative and a signal wire. There is a rotating disc on the bottom bracket axle with a number of magnets on it. As each magnet passes the sensor, the signal wire goes 5 volts positive.
I have always said that it would be possible to build a small circuit with a 555 timer chip to simulate this if required. There are other ways, such as fitting a magnet disc to your non-motor wheel, and putting the sensor next to it .

 

[flecc]

Adding to what Neptune has posted, the bike's controller has to receive a stream of 5 volt pulses to indicate pedalling before it delivers current to the motor. No pulses or continuous current won't do the trick.

 

[Nigelk]

ok many thanks for the reply's, most helpful.

 

[danfoto]

So if all the pedelec sensor does is sense rotation of the crank axle and thereby allow the controller to calculate cadence, how does the controller of my Agattu know how much assistance to allow the motor to provide in any given circumstances?

 

[flecc]

That's not only a pedelec sensor in your power unit Danfoto. Instead of just sensing the rotation in that way, your pedal effort is transmitted through an amorphous alloy sleeve. That causes a magnetic disturbance in surrounding coils and the level of effort determines the degree of disturbance. The resulting feedback to the controller tells it how hard you are working. You can see all the details of your pedalshaft assembly with a description of how it works on my repairs page for the old unit. Scroll two thirds down the page:

Motor unit repairs
.

 

[danfoto]

Ahah! All is now clear, so thank you sir (again!).

 

[laroche]

With respect to cadence sensors, I'm still confused about what happens when the frequency of the pulses exceeds the minimum rate for the motor to start. How, if at all, does the motor react to any further increase in frequency and any subsequent reduction?

 

[gray198]

Flecc,

Are there no limits to the depth of your knowledge

a much impressed gray

 

[flecc]

With respect to cadence sensors, I'm still confused about what happens when the frequency of the pulses exceeds the minimum rate for the motor to start. How, if at all, does the motor react to any further increase in frequency and any subsequent reduction?

Many don't, they just determine motion. Some can vary the motor responses according to the sensor pulse rate, but it's the controller that does that using pulse timing, not the motor. The controller is then able to deliver the current to the motor phases in a desired pattern.

 

[flecc]

Flecc,

Are there no limits to the depth of your knowledge

a much impressed gray

As with everyone Gray, immense limits and you should not be too impressed.

I always say that for every one thing I know, the number of things I don't know is almost infinite.

 

[neptune]

Further to my previous post re fooling the pedelec sensor. How many magnets does a typical magnet disk have , and can anyone estimate the minimum pulse rate necessary to trigger the controller to switch on the motor? I am thinking of a very low to tech way to fool that sensor. Imagine a small electromagnet, possibly the coil from a 12 volt relay. Place it very close to the sensor, and drive it with a flip-flop oscillator based on 2 transistors, or a 555 chip . Power it with a couple of torch batteries. And suddenly, you become the nephew of Robert. In other words Bob`s your uncle.

 

[flecc]

Indeed, easily done, though even less complex would be to omit the electromagnet and just feed the 5 volt pulses directly from the timer to the controller.

 

[mike killay]

Indeed, easily done, though even less complex would be to omit the electromagnet and just feed the 5 volt pulses directly from the timer to the controller.

And that way I imagine, the new EU laws will be defeated if ever throttles become outlawed.

 

[flecc]

This would only keep the motor running, not control it's power though. To do that would need a way of defeating whatever torque sensor is fitted. The type of rear in-frame torque sensor that's becoming popular now could be turned into a throttle. It would need a lever or twistgrip that used leverage advantage to apply variable pressure to the transducer, quite easy to achieve.