No announcement yet.

FAQ - Throttles and debugging throttles.

  • Filter
  • Time
  • Show
Clear All
new posts

    FAQ - Throttles and debugging throttles.

    There is a lot of good information on electric bike throttles. Not much of it has been written from the perspective of a builder trying to diagnose throttle issues. As a result, throttles can seem like black magic.

    The good news is that for compatibility all electric bike throttles work about the same way. They are basic three wire sensors. They have two reference wires, the ground wire and the +5 volt wire. The third wire is a signal wire which varies ground and +5 volts depending on the position of the throttle.

    Some throttles assemblies also include an on-off switch and voltage meter. This system runs on a separate two wire pair at ground and full battery voltage. When the on-off switch is open the controller is off.
    Getting back to the throttle. A bit of history helps to understand why throttles work the way they do. Electric bike throttles evolved from the ‘gas’ pedal on electric golf carts and forklifts. These systems used a fixed resistor and a variable resistor set up as voltage divider to enable the driver to adjust the speed of the vehicle by moving the pedal.

    Click image for larger version

Name:	VoltageDividerSchematic.jpg
Views:	2569
Size:	26.4 KB
ID:	29363

    Moving the contact on the variable resistor causes the signal voltage to vary between ~0v and ~5v. Because of voltage losses in the fixed resistor one never got quite the full 5v range. This worked well enough for over fifty years.

    There were a couple of problems with the system:
    1. Variable resistors are exposed to the environment. The winding on the fixed side must be in physical contact with the wiper on the moving side. Water and other contaminants can work their way into the system. This reduces reliability.
    2. Accurate variable resistors are rather expensive to manufacture. Electric bikers want good throttle control at a reasonable price.
    3. It is hard to isolate the system from mechanical shock. Since the two part of the system must be in physical contact they must be durable enough to withstand the mechanical stress electric bikes put on their handlebars.

    To work around these issues throttle designers started replacing the variable resistors with hall sensors. Hall sensors work by detecting the distance between a sensor and a magnet. This allows designers to fully enclose the sensor in the stationary side of the throttle and fully enclose the magnet in the moving side of the throttle.

    __PIC__ hall throttle

    To retain compatibility with older throttles, they continued to use the same 5 Volt three-wire standard to communicate with the controller.

    The first challenge with diagnosing a dodgy throttle is to figure out if the problem is with the throttle or with the controller. The easiest way to diagnose a dodgy throttle is to just the replace throttle in question with one that you know works. It doesn’t take long to accumulated a box of spare throttles.

    As ykick pointed out in a related thread, throttle testing on the bike can be unsafe at best. I decided to rig up a little tester that one can use on a work bench.

    Click image for larger version

Name:	1.jpg
Views:	2496
Size:	280.8 KB
ID:	29364

    Throttles need 5 Volts so a little USB wall wart can become the power source. To make this reusable and practice my crimping skill, I crimped jst-sm connectors to the ground and 5 volt wires.

    Click image for larger version

Name:	2.jpg
Views:	2462
Size:	131.7 KB
ID:	29365

    I just crimped a third green wire to use to measure the signal voltage.

    Click image for larger version

Name:	3.jpg
Views:	2502
Size:	183.4 KB
ID:	29366

    Finally, I inserted the crimped connectors into a 3 pin jst-sm socket, making sure the ground, 5 V, and signal wires corresponded with the wires on the throttle.

    Here is a bench test of a throttle.

    Click image for larger version

Name:	4.jpg
Views:	2462
Size:	344.5 KB
ID:	29367

    In the off position the voltage reading was 0.85V

    Click image for larger version

Name:	5.jpg
Views:	2436
Size:	351.7 KB
ID:	29368

    At full throttle the voltage was 4.52.

    It seems I am having a throttle problem today I went to go for a ride and noticed it at been raining as soon as I got on my bike I turn the power on the throttle and I believe the bike immediately wanted to take off without twisting the throttle so I powered it off and pulled the brake got the wheel to stop spinning I then turn the power back on and had more throttle sticking issues. Today the throttle stuck stopped pause stuck wide open about 5 or 6 times I made sure the bike was parked in a covered area all day coming home from work it seem like the problem had went away this throttle is brand-new only a couple weeks old I think this is the second time it has done this. I'm pretty sure it was raining the first time that's happened as well. How common are these throttle shorts or throttle sticking with these e-bikes I have a durable thin sheet of rubber laying over my throttle that stays in place to prevent water seeping into it evidently it was not doing its job but what I noticed today is sometimes when I went to activate the throttle it would not engage the motor until I spun the wheel a quarter-turn kind of like a Pas system. The controller is pedal assist compatible but not hooked up. I have not noticed this quarter turn Behavior out of the tire necessary for motor engagement until today . Would it be more likely I had water in my throttle or would it be more likely that I had water in my PAs connector or is this just a guessing game and possibly the controller as well. Will this rain shorted throttle cause the controller to fry. I bought one of those e-bike LCD controller system testers on eBay but have not received it yet. My guess is if I can reproduce the throttle sticking and unplug the throttle from the controller this will confirm controller good or bad.
    Last edited by JamieWlcox; 05-19-2019, 03:15 AM.


      One of the nice things about keeping the old throttle output/sensor input design, is that if the ebike controller and display head don't provide a true cruise control, you can build your own and bypass the throttle as needed. I simply designed a small circuit that puts an selectable voltage on the sensor lead for the speed I want to maintain (on level ground). In the case of one of my bikes, I just selected a single speed and set a single voltage to maintain a decent "cruising" speed of about 15mph/24kph. Sure helps avoid a cramped hand on long rides. (the return spring on that throttle is STRONG. If I ever have to open it up to fix that throttle I'm also going to adjust the spring torsion)

      Extra Credit:
      Of course, setting a fixed voltage won't maintain a constant speed as the battery is depleted. It will gradually slow down a bit over time. But, it's close enough for jazz. I didn't think designing and making a circuit with speed feedback was worth the time but I bet an Arduino or similar would be a reasonable approach. (and if one really wants novelty or unique method of speed measurement, could place a playing card in the spokes and use a microphone to pickup the frequency of its "clacking" and just set the Arduino to adjust the throttle sensor lead voltage to maintain that frequency.....cue drum roll and cymbol clash)
      Last edited by jGecko; 09-15-2020, 12:12 PM.


        Originally posted by jGecko View Post
        I simply designed a small circuit that puts an selectable voltage on the sensor lead for the speed I want to maintain
        I'm interested in this setpoint solution... would you mind posting a diagram and instructions?


        See my completed Magic Pie V5 rear hub motor E-Bike build HERE.