Bringing the Sunstar S03 mid drive motors back from the dead

Electrifying the Dummy part 1

So a few weeks ago we received an email from Antony and Celia who had a very interesting proposal.

They had 4 broken SunStar S03 Mid drive motors that they had been donated. They all looked to be in good physical condition but none of the electronics worked and all had not shown any sign of life. Did we want them and could we get any life out of them?

Always up for a challenge we agreed that yes we would indeed be very interest and if we could get more than one working would be happy to help Antony and Celia in electrifying there Surly ‘Big Dummy’ Cargo bike with one.

Anthony had already done some background investigation into the 24V motors, and he had one controller board de-potted and come up against a brick wall. The likelihood is that some of the mosfets were blown and it might be possible to replace them, but as there is very little information on the electronics used in these old but well build drives a different approach might be the best option. Anthony had done what I would have to be honest and purchased a cheap 24/36v generic ebike controller with the hope to wire this up and get some life out of the unit. But this is where things got a little complex. He could wire up the motor phase cables but whatever combination he tried he just got a ’03 error’ on the screen.

So the driver clearly has three high power phase windings for the motor and connecting these was not tricky, however that’s not all that’s needed to get the motor to turn, or more accurately to tell the controller how to turn the motor.

Antony contacted us at this point to see if we could help them figure this out.

So I took the 4 motors and started to look in more detail at the wiring.

Now almost all ebike motors I have worked with use hall sensors within the motor stator to detect the speed and direction of rotation. Its these hall sensors that tell the driver, when the motor is running, and simply pulse a single voltage every time a magnet passes them, so a hall sensor needs supply voltage, a ground and gives an output signal. Without this vital feedback the controller would very quickly refuse to output any power. In fact its so quick you don’t really even notice the motor move before it stops. So I was confident the missing link here was just how to wire up the hall sensors.

Easy!, well not quite…

You see the hall sensor cable on the controller side was logical and the pinout was as follows…

Red – 5v
Green – Phase 1
Blue – Phase 2
Yellow – Phase 3
Black – Gnd

Simple enough. but at the motor end I had 6 pins not 5 and the colours were…

Brown, Red ,Orange ,Yellow ,Green ,Black


So what could this pinout be, what’s the extra pin for and which one is it.

Well thankfully after removing the motor and gaining access to the Hall sensor printed circuit board I could clearly see the tracks and work out the pinout.

So colour coding the pins you can see the heavier tracks top and bottom highlighted as Brown and Black look to be logical power and ground rails. and brown and black would be logical colours to use, however my first thought was that red would be 5v power 😉

Ok so lets assume those are the power rails what’s left?

Well you can see the next three wires , red, orange and yellow go to the three hall sensors via a inline resistor and they are even labelled R1, R2 and R3. I think its a safe guess these are our three Hall Sensors!

Ok so what’s the remaining green wire?

Well if we follow that track it goes right over to the far left of the board and hidden under the blob of white silicon you can just see it says ‘TH1’ could this maybe be a thermistor? sure enough the other side of the thermistor goes to ground with a filter capacitor to the 5v rail.

So I think we have found our pinout!

Motor side – Driver side

Brown – Red 5v
Red – Green Phase 1 GREEN
Orange – Blue Phase 2 BLUE
Yellow – Yellow Phase 3 YELLOW
Green – TH1 Temp sensor
Black – Black GND

Bingo, cable it up and we no longer get an error and the motor now rotates!

I could even measure the resistance of the green wire pin as the motor heated up, we may even be able to calibrate a thermometer to this value to get a temperature reading.

After this proof of concept on the bench, and showing Anthony and Celia the good news, we ordered a cable to better connect the motor to the controller and used a separate controller housing on the bike to make it a tidy solution that would be able to cope with some wet weather!

Anthony helps me solder up the new cable when it arrived.

and fitted the motor onto his bike for the initial throttle tests.

Shortly afterwards, he wired up the pedal assist sensor

Hope you like pedalling fast Anthony!

We would like to test the motor on 36v as the power generated is not as much as it could be and with such a heavy bike when fully loaded and the steep hills of Calderdale every bit would help!

That thermistor may come in very handy if we go down that route to keep track of how hot the motor is getting.

So keep posted for future updates on our experiments with this motor.


Hi, I am one of the trustees of Empowered people. As well as looking after this website, I am also an ebike builder, modifier, hacker, and general bike mechanic. I have been involved with the Charity since 2015 and participate in as many rides as I can with my wife Catherine on the 'Bike Train' if you look around the website you are sure to find me in the background somewhere.

11 comments on “Bringing the Sunstar S03 mid drive motors back from the dead

  1. i just picked up one of these motors but the problem is i cant find a chainring for it? do the bafang ones fit ?


  2. Thomas Schilling

    I have the same problem. Great you documented your way through it. One last question is about the pedal assist sensor. Did you use a new one or does the sunstar s03 has its own. If it has its own, where do i find the cables and to what connection pin i connect them to the controller?
    Thanks and best regards from germany!


    • Neil Perry @ Empowered People

      Hi Thomas. As we were using an external controller. We used a throttle and PAS sensor connected to that new controller. The sunstar does have a built in torque sensor and appears to be standard 0v,signal,5v pinout so with a suitable controller that has support for a torque sensor this could maybe also be utilised. But we have not been able to try that yet.( I hope to try it at some point)


    • Neil Perry @ Empowered People

      I have to say that from our experience I would say the sunstar would run better on 36v , 24v does seam under powered compared to other mid drive motor s we have used.


      • Thomas Schilling

        Thank you Neil for your quick response. I appreciate that. Whenever a personal project is getting stucked and i have no clue how to come to an solution, this really makes me feel uncomfortable. My family has to suffer from my bad mood :-). So the like to thank you too. On the other hand i feel like god, if it works. Unfortunately most of the time i feel like a small earthworm.
        I already have the 36v version. It is the S03+ . The S03 and earlier versions are recommended with 24V
        Just to check that i got you right. I ordered a standard China-Controller for ebikes with connection for a PAS-System (and throttle). Installing this PAS-system from the controller should make the sunstar motor unit work even without a throttle? I am not familiar with these components, but as far as i investigated the internet belonging to PAS-Systems, it is similar to the torque 0V – Signal – 5V system, right? So there should be a chance, that the torque unit of the sunstar will work with this controller?!
        Thank you and best regards from germany.


      • Neil Perry @ Empowered People

        Hi Thomas,

        I am sure by the end you will feel king of the mountain and not doing laps with the earth worms 😉

        Yes the sunstar motor should run fine on 36V, almost all e-bike motors run fine on double there ‘normal’ voltage in fact they will draw less Current at the same power output. the main concern with over voltage is not to generate too much heat so keep an eye on that. I have used cheap Digital Aquarium thermometers in the past to monitor the heat inside the motor housing e.g

        I have another sunstar s03 with a dead driver that I may experiment with on 36v and 48v to test how it performs, if I do I will write this up in the blog.

        I have not yet tested the torque sensor port in the sunstar motor but I agree that it should be 0v,signal(0-5v),+5V and should wire direct into any controller with a torque input cable. I think the main issue will be calibrating the level the signal gives out at different levels of torque. For example it may be over or under sensitive. With some controllers ( like the tongsheng TSDZ2) you have the option to run very configurable opensource firmware, and those to have options to calibrate the torque sensor at different levels of weight on the pedal. But you may find with the simple cheaper Chinese controllers the mapping is basic. worth trying anyway.
        It may also be possible to do the mapping with an external small circuit board maybe even a micro controller like Arduino.

        Please keep us updated on your project always interested to hear what others have done with different motors.

        All the best
        Mit freundlichen Grüßen



  3. Finding info on these sunstar systems is hard! Maybe someone can help, I have the original 24v Sunstar S03+. Had a battery problem so replaced it with a Greencell 24v battery. Problem is the new battery has only positive and negative connections. I cut the old 4 pin connector off and soldered the +- cable on to the new battery. So now I only run on emerceny or lowest setting. Since there is no communication cable on im thinking I want to jumper it so it always runs on the middle setting. Is this possible? best regards


    • Neil Perry @ Empowered People

      Hi Mike, you can jumper the other two pins on the battery cable to bring the driver alive. I also discovered that. But I don’t think you would be able to get the controller to work without a display to communicate with it. Most displays for ebikes run a simple serial communication between the display and motor controller. I don’t know for sure if this is the case with the sunstar. But I suspect so as the LEDs do respond to error codes depending on different faults which I would expect to be done with simple communication. Also the display does contain a chip which would also suggest it’s more complex than a simple extension of switches etc. I haven’t ever looked into it further as we don’t have any working controllers, never have we were donated a box of ‘dead’ motors and so decided to bring one back to life by using an external controller and it’s own screen etc.


  4. Danielle

    Hi there, did you ever try run it on 36V? I have a Nihola with an S03 and a 24V battery. Would love to switch to a 36V, but unsure if I can just do this, or need to modify something?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: