Bike speedometer sensor signal

Hello everyone.
I am trying to fix an ebike which have an internal speed sensor, which share the white wire with a temperature sensor (thermistor).
From ebike forum I have learned, that the voltage reading represent the temperature, and frequency of "zero, low" pulses in this reading represent the wheel speed.

There is some kind of electrical problem, I suspect with the thermistor and the voltage is out of range - it creats error message.

So if the speedometer signal is not connected (or not strong enough) the bike only allows some partial motor speed which shouldn't be exceed .

One solution is to build an external wheel revolutions sensor - I have access to +5v and GND, the component is called "unipolar hall sensor" with 10k ohm pull up resistor, and of course spoke magnet.

In this way the signal would be more "natural", more similar to the original - frequency is increasing with the speed.

I have read some post here, a lot of people mentioned that it should be all secured and tight , that is also correct.

So some people offered just to provided a signal in electronic way, that would be easier for home built, but I guess it would be constant frequency signal

The question is , will the ebike controller "tell the difference" ?
Because when there is no signal at all , the controller still know the speed from another digital sensors (3 position sensors that are used for the motor operation).

I think that even with a fixed frequency, the second option is easier to built (physically)
And I could try some different frequencies,
Maximum is 180 rpm.

Unless the controller is really comparing the speed sensor with the other 3 on a linear basis.

So I would be very grateful for any tips or info.
Its not about hacking the speed, I'm just looking for a solution.


Is replacing the current speed sensor not possible ?

Thanks you for your reply!
It requires opening the motor with star shaped tools, and I'm not sure if I would find replaceable components inside.
You are right, I'm trying to find a service lab also.

Maybe inside I would have access to the resistor, or even to resistor wires - I think the speed sensor is ok, just the thermistor is faulty.

I have tested it with voltmeter, and it responses to the wheel turning.

I even twice manged to fix it all -
One time by adding resistance
Another time by changing the voltage.
When I changed it from 0.5 to 0.8 it all worked good.

Is there a way I could amplify a signal that is now 0.08 volts to at least 0.8 volts ?
That would be good too.

I'm trying to avoid opening the motor...

You really need to replace the part

1 Like

Torx T25

Inside a hub motor you have phase winding and three Hall sensors which are positionong feedback. The BLDC controller or the LCD calculates speed using pulses from the Hall sensors

Another wau to calculate speed is to put a magnet on a spoke and a hall sensor pickup on a stay and count RPM, then multiply by wheel circumference.

The thermistor is there for a reason, so to "replace" it with injected voltage...

Thanks for your reply!
So I have checked with multimeter, there is an internal component which bring the voltage reading to zero, once in a revolution. (When air ventil pointing down)
I think it's called unipolar hall sensor.
Anyway it's a different signal from the 3 "conventional" hall position sensors, you know they are called ABC or uvw = they represent the angle of the motor like a binary count 0-7, and the special combination changes every 60 or 120 degrees...
So either this 4th sensor is damaged, or either a thermistor connected to it is.
I find it very interesting, the way the ebike controller reads data...

Thank you for the reply.
Do you mean I should not cancel the thermistor warning, should not bypass it ?
Well if it will work, I Will have to watch the temperature myself, to stop once in while and feel the temperature.

Definitely not, unless your really sure you know why its there, you don't want the motor bursting into flames whilst your riding the bike presumably ?

I would have thought an eBike forum would be a better source of information of why the thermistor is there and the practicalities of whether if fitted it can or should be bypassed.

1 Like

That's entirely up to you. Is it there just as a diversion, or is there a purpose? But seeing the continuance it seems to me you made up your mind.

No you won't. You'll be very busy "Weeeeeeeee" and forget all about it. :upside_down_face:

1 Like

True.

Hay everyone. A little electronics...
I want to convert the


input signal :
High=80 mv. Low =40 mv DC=97% max freq = 3hz

The output should be normal 5v and 0 wave.
So I can create a V ref with 120 ohm voltage divider with 10 k. 5v will make VRef=60 mv.
Do you think it will work with LM 358 ?
"Non inverting comparator"
How can I simulate it on PC

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.