KT36 Controller and PAS sensor

[MLarsen]

Hi Pedelecs

I trying to change my stock controller and display to a KT36/48WSRTD-SJT02L2 together with a KT-LCD8H Display

The controller is a dual mode controller and thus able to handle Hub that does not have Haa effect sensors in the HUB

I have also previously tried installing a brainpower controller but that did not support motors without Hall effect sensors so i bought a KT36... controller

But... I am going nuts as i still cant get the HUB motor to run while pedalling :-(. THe Hub is turning around when i'm pressing Downarrow and going into the Walk mode and the display shows 6km/h

I'm using the PAS senser already installed in the bicycle which works absolutely fine with the stock controller. So why the heck cant I get it to work with the KT controller???

I have verified that the PAS sensor is outputting pulses of 5V and have ensured that the pinning should be correct.

Any tricks to emulate PAS sensor or any other methods to search for the fault?

Hope to some great input and idears

Br
Morten K

 

[AGS]

Try pedalling backwards. If that works then set C1 to 5, 6 or 7.

 

[MLarsen]

Try pedalling backwards. If that works then set C1 to 5, 6 or 7.

Thanks for the advice... I have tried to backpadel but not tested it thoughroughly and consistently in all settings
Btw I have a 12 Magnet PAS sensor. Could thet cause any trouble?

 

[Nealh]

Show us a pic of the PAS sensor , I had a 12 pole PAS that gave me issues years ago so opted for a 10 pole set up and never had an issue again.

 

[MLarsen]

Hi All

Thanks for your replies... I have now taken the time to investigate further into to PAS sensor...

It seems like the PAS sensor is a typical PAS sensor as it is providing a Signal when pedaling forward and nothing when pedaling in reverse, so far so good. But I wonder whether the High period of the PAS sensor Signal is to short?

the only reference i can find to a explanation of PAS sensor signal is this: https://ebikes.ca/learn/pedal-assist.html

But there are not references to min and max time period for High and low part of the of the signal

Do any of you know of souch specification or have made similar measurements on your PAS sensor so we can compare my signal to another reference?

Thanks in advance for your help

BR
Morten Klitgaard

 

[MLarsen]

Hi All

I have now fixed the problem which was an issue with the PAS sensor. however not an issue with the PAS sensor itself but rather the connection between PAS Sensor and KT36 Controller

PAS sensor are utilizing another wire coding than KT36 controller and to make matters worse they both utilize a Male connector so I messed up in the wiring scheme and erroneously connected sensor output to GND

KT controller wiring colors: Brown = 5V / Black = GND / Yellow = Sensor input (Squared signal)
PAS sensor wiring colors: Red = 5V / Black = GND / Blue = Sensor output (Squared signal)
DIY Adapter cable: Red = 5V / Blue = GND / White = Sensor signal

I simply got confused by the colors and thus connected Blue to blue... once correctly connected it all worked fine...

I actually came to the situation where is mounted a Analog Discovery Logic analyzer and signal generator to my KT36 Controller (and also the old one) to modulating a signal in terms of frequency, ON/OFF periods (timing) and voltage levels to see how "robust" the controller was against variation in the PAS sensor signaling

And the answer to this is: It very robust as it accepts a wide range in terms of voltage level, timing and frequency

- Accepted Voltage level: 3.0V to 5.2V
- Accepted Frequency: 2-1000Hz (1000 pulses pr min = 83 rev/min with a 12 magnet encoder) It was not that particular about this so there is some uncertainty about this number
- Accepted Duty cycle: 1% - 99% (Duty cycle is relation between signal is High and Low, a signal with 1% duty cycle the time time of signal is high is 1% of the total period and 99% of the time low)

So in general one should be able to emulate the PAS signal by just quickly pulsating the signal manually by i.e. "tapping"/touching the signal wire on to the 5V supply line 10-20 times.

Any how thanks for your support even though i solved it my self ;-)

Br
Morten Klitgaard