PAS Sensor KT V 12L

[Kermit]

Hi All,

A little bit of introduction. I have a 250W Mxus, 36V hub motor, mounted rear with cassette. I am using the KT controller with LCD 6 display. Also I have mounted PAS sensor, KT V 12 L – 6 Magnets, 12 signals.

I have fitted all together and at first glance it worked fine. However I have noticed one annoying thing with my PAS.

When I pedal evenly at a certain speed, PAS is giving the signal to the motor, speed is constant and you can feel the assistance from the motor. Then when I want to accelerate, PAS stops working for a while, speed on the LCD drops down significantly, motor is not working. When I reach a certain higher speed and starts to pedal evenly again, everything is back to normal, PAS gives a signal to motor and you can feel assistance again.

It looks like the controller is not getting the signal from PAS/controller is not giving a signal to motor when there is a sudden change in cadence. Literally it seems like I need to match my pedaling to the sensor and the sensor does not react on my pedaling which, what I think it should look like.

It might be a setting issue however there’s so much possibilities with P and C parameters that any suggestion or a clue would be appreciated.

LCD manual
http://www.szktdz.com/en/news_show.php?article_id=542

PAS
http://www.szktdz.com/en/news_show.php?article_id=392

Did you hear anything similar?

 

[Benjahmin]

Can you be sure that the splines of the pas are gripping the crank shaft fully? It sounds as if the pas stops rotating with change of cadence. Some people have had a problem with that type of sensor when the crank arm is too tight up against the sensor body, so slightly squashing it. Filing a couple of mm off the inside face of the crank arm seems to cure it.

 

[Deleted member 4366]

Yes, In think it's more likely a rotation problem than a settings one. Can you see a clear gap between the sensor and the pedal arm?

 

[Kermit]

I admit this was an issue at the beginning, there was a difference less than the mm between the gap and sensor width so I filed a crank arm a bit (it was enough to take only the coating). I thought it was ok since at constant speed, with even pedaling under no load it was working fine.


I’ll try to file it more to get more clearance.

 

[Nealh]

If the sensor is gripped nicely on the bb axle and rotates without slipping then try two settings,
C1 values 5, 6 or 7 & C14 pas tuning strength values 1 -3 with 2 often the default setting.

 

[Kermit]

Bigger clearance does the job. It works now as a charm. Thank you all for a suggestion!

 

[Kermit]

Hi,
I was wondering what kind of magnets are used in these kind of sensors? Are there a neodyme magnets or others?
The other question is the direction of magnetic field. In KTV12 it is on along the radius, towards the sensor, right?

 

[Kermit]

I wanted to share some thoughts here. I wanted to upgrade my KTV 12 signal, 6 magnets PAS sensor to something like 24 signal, 12 magnets. There is a room for additional magnets already made in plastic wheel. I have bought some random neodyme cylindrical magnets, with 3mm radius and 2 mm height. They fit perfectly. I’ve double checked the right polarization. And it worked, however not in the way I’ve expected. I think, the sensor is more sensitive but what is a negative thing here the PAS worked also with backward movement of the crank. I’ve played with some C parameter settings with no success. With disappointment I have removed additional magnets.

I don’t know, maybe magnets were too strong, or the controller is not programmed to such a number of signals have no idea.

 

[TJS109]

Hi ,
I had a kit with one of these sensors and replaced it with a 12 magnet one, so I don't think it is the number of magnets that is the problem. The sensor I used was much bigger hence bigger spacing between the magnets. What I think might be happening is that the controller uses the mark-space ratio of the signal to work out the forward / backward rotation and that the increased number of magnets is changing to something like 1:1 so loosing the rotation information. If you have access to a scope you should be able to see it.