How does a simple, 1-wire, PAS sensor differentiates between CW and CCW rotation?

[FBminis]

Please note the attached photos, that's the kind sensor I'm referring too.

I've read the info in the following website but still have this doubt: https://ebikes.ca/learn/pedal-assist.html

I don't understand how those discs can make a hall sensor (with a Schmitt trigger) generate a square pulse when you rotate CW but no pulse when rotate CCW.

1) how do these magnets face the sensor? Always same pole or N-S-N-S-N-S?

2) I understand some hall sensor (like in pc cooling fans), can generate two pulse, 90º out of phase. This would be a 4-wire PAS sensor and the controller could easily interpret the sequence of pulses and determine the direction of rotation. But how do these 3-wire PAS system not produce a signal when you pedal backwards?

 

[E-HP]

I've read the info in the following website but still have this doubt: https://ebikes.ca/learn/pedal-assist.html

What is the doubt part?

 

[JackFlorey]

I don't understand how those discs can make a hall sensor (with a Schmitt trigger) generate a square pulse when you rotate CW but no pulse when rotate CCW.

By using two Hall sensors internally.

 

[amberwolf]

Please note the attached photos, that's the kind sensor I'm referring too.

I've read the info in the following website but still have this doubt: https://ebikes.ca/learn/pedal-assist.html

I don't understand how those discs can make a hall sensor (with a Schmitt trigger) generate a square pulse when you rotate CW but no pulse when rotate CCW.

1) how do these magnets face the sensor? Always same pole or N-S-N-S-N-S?

2) I understand some hall sensor (like in pc cooling fans), can generate two pulse, 90º out of phase. This would be a 4-wire PAS sensor and the controller could easily interpret the sequence of pulses and determine the direction of rotation. But how do these 3-wire PAS system not produce a signal when you pedal backwards?

There should be additional electronics in them to process the signal, rather than just a bare hall sensor. I haven't opened one up to see what's in there (went off to look for one to do this when writing this, but the box with PAS parts eluded me), but I expect there could be two hall sensors and a comparator of some type in some of them, or similar logic that allows sensing when a magnet passes one sensor vs when it passes another, whcih could be compared and if it's one way it passes the signal and the toher it does not.

 

[E-HP]

I don't understand how those discs can make a hall sensor (with a Schmitt trigger) generate a square pulse when you rotate CW but no pulse when rotate CCW.

By using two Hall sensors internally.

The KT one piece PAS has dual hall sensors. https://www.amazon.com/Electric-Accessories-Magnets-Connector-Conversion/dp/B09VNMG2CR/ref=sr_1_2?crid=3UNXOHJR66I5C&keywords=kt+dual+hall&qid=1654658070&s=sporting-goods&sprefix=kt+dual+hall%2Csporting%2C195&sr=1-2
I suspect the Grin all in one unit also is dual hall, but that's not listed in their description.

 

[FBminis]

4-Wire PAS system rely on dual plate HAL sensors like those in PC cooling fans. These differential field sensors have two outputs that emit two square wave signals which are out of phase, as two consecutive magnets with with different pole orientation pass by the sensor. One of these ICs: https://content.instructables.com/ORIG/FO0/D0FY/GZAD16Y9/FO0D0FYGZAD16Y9.pdf
The output from these ICs can be processed as a quadrature sensor by the controller, this way it becomes trivial to identify the direction of rotation.

Question about 3-wire, single signal output PAS sensors:
1) Do all the 12 magnets have the same pole facing out? If not, what's the order?

 

[E-HP]

1) Do all the 12 magnets have the same pole facing out? If not, what's the order?

I believe this is the case, all facing one direction. That's why an odd number of magnets is supported (e.g. 5 magnet discs). I added more neodinium magnets while experimenting, and found that at a point, too many magnets caused interference between them, which is why the ones with more magnets are also larger in diameter (5 magnet vs 12 magnet discs).

 

[FBminis]

1) Do all the 12 magnets have the same pole facing out? If not, what's the order?

I believe this is the case, all facing one direction.

This is why I don't understand how can the controller tell if we are pedaling forward or backwards.

If someone could test with another magnet they have around by passing through the disc's magnets and note if they repel or attract, that would help solve the mystery.

Found some interesting links about the subject
https://www.pedelecforum.de/forum/index.php?threads/funktionsweise-digitaler-pas-sensoren-details-fuer-interessierte.9539/
https://youtu.be/S4mNT4YW9k4?list=PL1C9WWfQsA70gqrmIoS-QggVKscYgITYv&t=271
https://www.digchip.com/datasheets/download_datasheet.php?id=359998&part-number=HAL50X