- As we all know the Pedal Assist System (PAS) sensor only produces an "I am pedalling" signal when pedalling forward but not when turning the cranks backwards (are there excptions to this?).
- I am idley curious as to how the sensor works out the difference between forward and backwards crank rotation - anyone?
PAS Sensor direction of rotation question
- Thread starter Bikes4two
- Start date
The sensor has a hall sensor in it, which will switch from off to on when the correctly orientated magnetic field passes it in the right direction. It will also switch the same when the opposite magnetic field passes it in the opposite direction. Wrong field direction or wrong movement direction won't switch it. A double hall sensor will only switch in one movement direction and one field direction.
The little chips we call 'hall sensors' have rather more inside them than just the sensing element, so they can be made sensitive to specific magnetic field changes.
Thanks for your input folks. Directed by your responses I've done a little bit of digging about Dual Hall Sensors, semi conductor devices that can have two Hall plates in the same chip and to quote from one source (LINK)
The rotation speed of a magnetic pole wheel can directly be calculated with the second sensor output signal. This signal is triggered by each polarity change of the magnetic field. "
There's also a YT vid demonstrating this principle HERE, specifically from the 5 min point onwards and the presenter gives the waveform pattern below which supports the written description above.
---------------
"Detecting the direction and speed of rotation of a magnetic pole wheel requires two separate Hall plates. Due to the distance between the two plates, they detect slightly different signals at any designated time. This is termed the phase difference. During a rotation direction change, the phase difference changes polarity. The vertical dual-Hall sensor detects this change and delivers a corresponding signal. The rotation speed of a magnetic pole wheel can directly be calculated with the second sensor output signal. This signal is triggered by each polarity change of the magnetic field. "
There's also a YT vid demonstrating this principle HERE, specifically from the 5 min point onwards and the presenter gives the waveform pattern below which supports the written description above.
We have seen the issue many times on the forum when an incorrectly fitted PAS is fitted and the user experinces back pedal operation due to the wrong orientation of the magnetic pole.
Last edited:
I'm pretty sure that's BS. AFAIK, my bike has a single hall sensor and it functions perfectly only in the forward direction. We get more problems on this forum with dual hall sensor pedal sensors not working than single ones.
The dual magnet sensor only prevents you from fitting it on the wrong side of the bike or the magnet disc the wrong way up. If you fit a single hall sensor on the wrong side of the bike or with the disc the wrong way up, it'll work only backwards, but it can always be fixed by flipping the disc, so you can choose which side of the bike to fit it. A dual hall one only works on its designated side of the bike. It's an an expensive way to solve a problem that has a simple no cost solution.
Last edited:
If I had nothing better to do, which will happen in the future but is not the case right now, I would set up a test to measure the pulse width coming from the sensor and plot it against wheel rpm.
If the pulse width stayed constant, I would infer a greater level of internal shenanigans than if pulse width was proportional to rpm.
If the pulse width stayed constant, I would infer a greater level of internal shenanigans than if pulse width was proportional to rpm.
Pedal Assist Systems - Learn
An in-depth technical look at PAS and Torque sensors. Have a look at our <a href="/getting-started/pas-options.html">PAS Options page</a> for a more simplified overview if you're trying to decide on which to use with your conversion kit.
gives a brief overview of various PAS systems
And I'm pretty sure that you're wrong about this, but when I questioned it before you were dismissive, to say the least. I still maintain that detecting direction requires two hall effect devices, albeit they can be integrated into a single chip.
I'm with you on the dual hall sensor explanation @jarnold, and in the link to the technical piece the single chip was 2 hall elements.
As for the greater reporting of dual hall sensors issues vs single ones, I doubt the average person would now the difference. Further, I can well imagine that manufactures just use the term 'hall sensor' or even just 'PAS sensor' irrespective as to whether the device is dual or single.
Until someone can come up with a technical explanation of how a single hall elements can sense direction of rotation, I'll stick with the dual hall explanations given in the links in my post.
As for the greater reporting of dual hall sensors issues vs single ones, I doubt the average person would now the difference. Further, I can well imagine that manufactures just use the term 'hall sensor' or even just 'PAS sensor' irrespective as to whether the device is dual or single.
Until someone can come up with a technical explanation of how a single hall elements can sense direction of rotation, I'll stick with the dual hall explanations given in the links in my post.
Here's the p;roof. I dismantled one. Dual halls have 4 or 6 legs. Learn from the master, not some irrelevant shite on the internet::
Last edited:
This forum is no exception to that of course. So how does a 3-legged device determine the direction of roation, that's what I'd like to know?
And taking a more simplistic approach, I have yet to find a 3-legged i.c. that offers dual direction detection, have you as then I might be convinced?
Mass produced stuff aims for minimum cost, so a specialist IC dedicated to a specific purpose is unlikely to have an unused pin, and the controller is not going to bother accommodating a function that can easily be performed externally.
The datasheet is for a 4 pin device with separate outputs for speed and direction, and specifically mentions two hall plates.
I suspect that some of the PAS sensor chips are like this but omit the direction pin and internally decide whether to toggle the speed pin depending on the detected direction.
The datasheet is for a 4 pin device with separate outputs for speed and direction, and specifically mentions two hall plates.
I suspect that some of the PAS sensor chips are like this but omit the direction pin and internally decide whether to toggle the speed pin depending on the detected direction.
When I raised this thread I was curious to know
In simple terms (and happy to be corrected):
Hi @matthewslack, looking at that data sheet was useful in helping me to understand, in principle at least, how a Hall Sensor with two Hall Plates can determine the direction of rotation of a magnetic ring like that that we see on ebikes.
In simple terms (and happy to be corrected):
- A single Hall Plate when subjected to a magnetic field (as in the rotating PAS Sensor magnet) produces a Hall Voltage that peaks as the magnet rotates past it. The single Hall Plate working on its own cannot know the direction of the rotation.
- However, using two Hall Plates (in the same chip) along side each other and in the same plane, (let's call them 'X' and 'Y') can determine the direction of rotation because one plate will see it's Hall Voltage peak before the other. E.g. lets say for clockwise rotation 'X' peaks before 'Y'.
- Conversely of course for an anti-clockwise rotation 'Y' peaks before 'X'
- The difference in time between the voltage peaks is referred to as the phase difference* and is easily detected and signal processed within the same chip and of course it is easy to generated a signal output only for when the rotation is in one particular direction.
- So a 3-pin device, Gnd, Vcc, signal and what from a marketing perspective what might we call such a device - Dual Hall Plate Sensor, Duel Hall Sensor or just a plain old Hall Sensor? I suspect it would get called all three, depending on the audience.
Last edited:
I was never relying on some "irrelevant shite on the internet". Just on my understanding of how Hall effect devices work and 50 odd years working with electronics.
This doesn't prove anything. Only three pins are required for a direction detecting chip that internally incorporates a dual hall sensor. And that internal dual hall sensor is required for the logic inside to work so that the signal will be present only for one direction of rotation. It is not possible to do that with a single hall sensor and a magnetic ring with uniformly spaced magnets no matter what their orientations are.
Related Articles
-
MTF Enterprises announces acquisition of EMU Electric Bikes- Started by: Pedelecs
-
Wisper 806T folding bike wins Which? ‘Best Buy’- Started by: Pedelecs
-
Sustrans calls for protected cycle lanes- Started by: Pedelecs
-
Amazon launch their first UK e-cargo micromobility hub- Started by: Pedelecs
