Announcement

Collapse
No announcement yet.

Guide to Hall Sensor Throttle operation, testing, and modification.

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

    Guide to Hall Sensor Throttle operation, testing, and modification.

    ***** 8-16-2018 Welcome to the Hall Sensor Throttle Thread! If there is something you'd like to add, correct, needs better explanation, or have a question about... feel free to add a Post or P.M. me! Please no comments... Thank you! -T.C. *****

    Introduction and Safety Precautions...

    Having my newly minted Magic Pie v5 kit start jerking at a stand still like a top fuel dragster coming up to the start line with no input on the throttle. Got me instantly interested in the magical world of hall sensor throttle operation. Also seeing owners with issues of front dead band, jerky and wheelie starts, and need to reduce top speeds or troubleshoot throttles fueled my interest more. Hopefully the following will provide useful information for you.
    Oh, and my problem was a bad hall sensor ground connection. See post #44 here...
    Information here will be mostly presented as to the way it works with the Pie, as it is what I own, maintain, and test on. But will certainly be very similar to other controllers.

    The throttle has a very important role, that to make the motor go. Even more important is the function of the brake lever's E-Brake switches, and the bikes regular mechanical braking system. You must make absolutely sure that you can disable the motor electrically by the use of switches actuated by the braking levers that disable the motors electrical drive circuitry, and be able to stop the bike safely mechanically before any test rides after working on the throttle. Test first on the bike stand checking both start, run, safety cut-out(s), and stopping, until everything works perfectly.

    Know how it works? Go right to... TESTING.
    Or for more advanced...MODIFICATIONS.
    See also... CONTROLLER OPERATION.


    Types of Hall Sensor Throttles...

    Full twist and half twist throttles. Which refers to the length of the handle that twists, as opposed to the actual angle that it rotates. Which by the way is a little under 68 degrees. Thumb throttle. Which also rotates about 68 degrees. And foot actuated throttles. None of which should be confused with a resistor type throttle which works with a potentiometer verses the Halls electronics.

    Hall Sensor Throttle...

    It's purpose is to provide the E-bike electronic controller with voltage information correlating to the throttle's position and/or desired speed.
    This voltage output varies typically from .8 vdc to 3.5 vdc with the input voltage at 4.3 vdc, and at a very low amperage, typically around 6 mA.











    Hall Sensor Operation...in particular Type SS49E.









    With a constant DC voltage input, the hall sensor will output a variable linear DC voltage in proportion to how much magnetic force (gauss) to which it is exposed. The magnetic force may be positive, neutral, or negative depending on throttle or magnet position. And may be produced by one (typically) or more magnets. If a magnetic field is not present or 0 gauss, output will be at 2.5 vdc. Where it will be pulled down to 1 vdc with negative gauss, or pulled up to 4 vdc with positive gauss. (With an input voltage of 5.0 vdc.)

    Note: If your sensor voltage output is reversed from what you want. You can reverse it by flipping the magnet, or magnets, end to end reversing the polarity. Or by flipping over the sensor housing so that the part stamp is facing away from the magnet!

    Note that the hall sensor is a Honeywell Type SS49E, which is different from those used for motor position hall sensors or brake reed switches. They have no moving parts to wear out, and are reliable.

    For more specific details see SS39ET/SS49E/SS59ET Series Linear Hall-effect Sensor ICs Datasheet here...








    Controller...

    Provides the regulated 5 vdc input voltage for sensor operation. And receives the variable output voltage from the sensor to determine power output to the motor accordingly. The controller will also lock out motor operation if throttle sensor input is higher than 3.8 vdc (shorted output). Note some throttle manufactures may put a voltage drop resistor between 5 vdc+ and ground for the specific reason of making sure full voltage output stays below the controller lockout voltage. Lower than .4 vdc (open output, error code #2), or somewhere in the middle on power up (stuck throttle), which will automatically reset when the throttle is returned to the full off position. All to prevent unwanted or dangerous operation. These safeties are certainly welcome, but may get in the way of throttle troubleshooting... be aware.
    Last edited by Tommycat; 3 weeks ago. Reason: Updated information...
    See my E-Bike build in progress HERE.

    #2

    Throttle Testing...

    One must start out with the hazards involved. I recommend working on your throttle under test bench conditions using a battery powered voltage (5+vdc) supply.Where you'll be comfortable, have better accuracy, and don't have to worry about the motor or wheel starting up. If not possible, or when you have to see the controller input/output. Be very wary of motor start-up and operation. Keeping the bike position steady and the wheel securely held off the ground.

    The electrical hazard at the hall sensor itself could be considered low, as you'll be working with 5vdc at about 6mA. But of course be aware of any higher voltages close by. On my Pie, full battery power is in the same harness connector and in the throttle housing. When in doubt as to what goes where... open them up and take a look! See disassembly instructions farther below...










    From here on out for wiring reference only, the 5 vdc + input wire will be RED. The 0vdc ground wire will be BLACK. And the sense wire will be GREEN.




    You must determine and verify which wire color goes to which.

    Input positive voltage may also be referred to as 5 vdc even tho it typically varies between 4.25 and 5 vdc or so.

    As any electronic component could be, the hall sensor can be damaged. Mechanically as in broken terminal legs or excessive hard vibration. And electrically by over voltage input, over current (shorts), or reversed polarity... all to be avoided!

    With the low current of the throttle circuitry. Erratic operation from the throttle wiring getting WET is a known issue. They are not water proof unless specified and if full battery voltage is in same housing, may get shorted.


    Throttle Wiring...

    The hall sensor throttle requires three wires to operate. Red for 5+vdc, Black for 0 vdc-battery ground, and Green for sensor output.

    The most important thing to get right is knowing which wire are which. Wiring colors vary widely between manufactures and can not be trusted. So always verify!

    If your fortunate to have a wiring or connector pin-out diagram, use it! On the examples below or any other diagram, be sure and notice which side of the connector you are on. Male (with the PINS) , or female. As the opposite side pin-outs will be mirrored. Very easy to mix up, be careful here. Then check to be sure.









    If unknown, you can start at the hall sensor itself, see the input and output terminals on the diagram below. And confirm visually which wire colors are connect to each leg. They should be well soldered and properly insulated to prevent shorting. The beveled side of the sensor, with the part number stamped on it will face towards the magnet. If you need to disassemble your throttle for wire colors and/or further troubleshooting... see throttle disassembly below.






    Controller Wiring...

    Note: the controller 5 vdc regulator is not very powerful and is only rated to provide around 100mA or even less.(I've seen down to 40 mA) Do not short out the supply wires. It will survive a brief short, which I unfortunately can attest too. But probably not for an extended period, which I decline to verify...
    Verify battery power to controller.Some controllers require an activation (key switch or button...) or "on" signal to it before energizing 5+vdc supply.
    Have a map?...Good use it

    If not start by checking the three wires for the 5+vdc regulated output power.


    Mark the positive wire as 5 vdc +(RED), Then using a DMM with high resolution and keeping the positive probe there...check the other two wires with the Black test lead, both which will provide a grounding path, but the sensor wire has a little resistance in it. So true o vdc battery negative wire will have the highest voltage potential. Mark it Black. With the sensor wire, just a hair less voltage potential. Mark it Green. Example: In my testing, from 5+vdc to true ground meter read 5.07 vdc. to sensor wire read 5.06 vdc.

    If your controller has more that 3 wires going to the throttle, odds are one of them has full battery + power going to an LED or some type of voltage display, or perhaps a power/key switch. Keep this full battery voltage away from the wiring used by the Hall Sensor Throttle!

    Once you have all the wires sorted, we can get to testing. The first thing you ought to know by now is if your controller is providing 5 vdc ( on the Pie, more around 4.3 vdc).
    If not, it's a problem relating to the controller. Check to see if the controller is activated and powered on and if any of the 5+vdc supply wires are being shorted.
    Having 5+ vdc available, continue on to the throttle checks.

    Having an external and open connector such as a JST, will make things easy, as you can just back-probe the connections. See this great guide for general testing.








    Pie owners have it a bit more difficult with the weather proof HIGO connectors. I made up a set of jumper wires that have a female connector on one end, and a male pin on the other, with a bit of insulation missing in the middle to make electrical connections for testing with the throttle fully assembled.











    Throttle Output and Profiling...

    After correct input voltage and proper ground is verified, sensor output voltages are now taken. At throttle off or minimum position, output voltage should read about .84 vdc. At WOT (wide open throttle) full on, voltage should read around 4.25 vdc. Don't get to hung up on exact readings as they very a bit. If magnet is missing or removed, output should be about 2.5 vdc.

    If you have a certain specific issue that you would like to identify and resolve, a throttle output profile is a great way to visually see the problem. Throttle dead zones, erratic operation, and wheelie starts to name a few can be plotted and looked at for fixes.

    First tape off and mark your throttle's twisting motion into 10 equal segments, 0% to 100% (AKA, WOT for wide open throttle) operation in 10% increments.









    Then use this form to fill out the voltage output readings at each graduated step.











    A typical well working plot will look like this...





    You should have a nice straight line between the starting and ending "hooks".
    Voltage required for maximum desired motor speed, and amperage draw may also come in handy. From this it can be more easily determined what the problem is with your throttle, and possible solutions to make it better. See Modifications in the next chapter.
    O.K. so now you see that a problem exists and need to get access to the internal components, such as the magnet, hall sensor, wiring and connections of your throttle to check out.


    Throttle disassembly...

    Thumb throttle is easy to see and describe. First push in on the three retaining tabs that hold the assembly together, a type of three finger salute will do. Being careful of the return spring which is under tension. Mark as required to get it back properly in place during reassembly.









    Remove housing retaining screw, and all should be accessible now.














    Full twist throttle is a bit more difficult to get apart as you need to reach up thru the middle of the housing (throttle off the handlebars) to unlock the same type locking tabs.

    I found that a tool kit like this is a must. Try bending one or two tabs inward, while applying prying separating force to housing. For more information on this look HERE!













    After access is gained, check to see that the magnet is clean and secure in it's proper position. And that the hall sensor is in it's proper position ( 1-2 mm from magnet face, beveled side with part stamp facing magnet) and sitting snug in it's holder. Verify that it's wiring is sound, properly connected and insulated from shorting.


    Bottom line...

    5+ vdc in to .8 to 3.4 vdc linear out depending on throttle position.
    Last edited by Tommycat; 3 weeks ago. Reason: Information update...
    See my E-Bike build in progress HERE.

    Comment


      #3
      Hall Sensor Throttle Modifications...

      Please note that a lot of controllers have the capabilities thru programming to modify operational parameters that can achieve results as obtained below, some even better. Checking your controller documentation first is highly recommended.


      Introduction: By now you should have a good understanding of how the hall sensor throttle works, it's power in and power out. If you don't read the sections above.
      In this section I'll discuss ways to manipulate the output, ether electrically or mechanically if you have the capabilities and desire to do so... Thereby changing it's operating characteristics.

      I have both bench tested first, then field tested the below modifications with both a Luna full twist throttle. And a generic thumb throttle, both of which use SS49E hall sensors. So I am confident that they actually work.

      That said, not all controllers or throttles are the same so
      please take precautions as you see fit before experimenting. Brake cutouts should always be used.As an added precaution I ran the 5 vdc throttle supply thru my throttle latch/on switch (which is usually used as a light switch) to use as a manual kill switch. Handy to disable the throttle output while working on the bikes electrics preventing unwanted accidental motor start up from bumping the throttle... And always check operation first in the shop with the wheel up, to make sure everything is perfect.

      As a hall sensor is relatively inexpensive I believe the first goal is to protect the 5 vdc power supply coming from the controller. Which is not easily replaced, and the cost for a new controller unappealing.

      For this reason while testing I use 3 multimeters... One meter in series with the positive power 5 vdc supply to observe amperage draw. (Keep below 10mA.) One meter reading supplied voltage, and one meter reading output voltage. A truly cautious modder may want to install an inline fuse of about 20mA, if there is such a thing. I never saw an amp draw over 10mA, even with multiple transistors. Typically for the modifications below, amp draw will actually be reduced. Of course all bets are off if you accidently short the incoming voltage which I did once. Luckily my meter put up an alarming racket which prevented any damage. Proceed at your own risk...


      Remember... No excessive voltage, amperage, or reversed polarity. And with the low circuit ampacity, high quality soldered connections are a must.

      Bench testing in action... I like to use a 5 vdc rechargeable battery supply.







      Field testing in action...










      #1 Starting Voltage/ Dead Zone Elimination Mod...

      Will help if you have to turn your throttle too much before your motor starts (front end dead band). Or to much of a jerk at start. Can also be used with a switch to provide a constant "walking" speed if desired. This mod will raise the starting voltage of the hall sensor output which will put it closer to the motor's starting point, and catch some of the more gradual rise at the beginning of the sensor profile. Note: this mod will not change the final (WOT) voltage output. More important is that the loss of sensor ground will result in full voltage output. AKA- FULL MOTOR OUTPUT. See precautions above! Leave enough room between starting voltage and motor start for voltage fluctuations due to temperature changes or an accidental slight bump of the throttle.





      Modification results profiled.











      #2 Ending Voltage/ Maximum Speed Mod...

      Will provide you with a means to regulate top end speed. Good for the grand kids bikes or to keep under mandated speed regulations...













      #3 Double Dipper Mod...

      Just a combination of the first two... easier to understand?







      #4 Delay Voltage/ Smooth Acceleration Mod...

      Also known as the anti-wheelie fix. :-) This mod is used to smooth out the acceleration curve by providing a delayed, gradual, and consistent timed on voltage output. With the usual off voltage drop or deceleration. Using the starting voltage mod with the extra electronic components works well. This mod requires more "tuning" of the pots due to the differences in your controller's internal resistances. Increasing the capacitor value will increase the time delay, but may also increase motor off time. This mod may decrease top speed a bit.



      Your goal with TP1 is to increase the fill-up time of the capacitor, more resistance longer delay. But not reduce final output voltage.
      Your goal with TP2 is to keep the starting voltage in desired range. More resistance, higher starting voltage.











      #5 Mechanical Mod...

      Electronics not your thing? There are still possibilities in the mechanical realm. But it will probably take more patience and testing to get just the right results.
      Manipulation of the magnets gauss to sensor will be tricky to say the least. A lot of trial and error is to be expected...
      Easiest, and probably most reliable that would start you out, is the change of gap between sensor and magnet. This would require that you modify the sensor holder of the throttle assembly. Typically the gap would be ~2mm, but say increasing the gap to 3mm would result in the change as described in the graph below...






      The loss of over all gauss will increase the starting voltage and decrease the ending voltage With the midpoint remaining the same. Be wary of not getting the starting voltage too high or to close to the motors starting point.

      Other experiments can be tried using Mu-metal to block magnetic field, or steel to modify it. A two magnet throttle would offer the most opportunity for manipulation.
      Last edited by Tommycat; 4 weeks ago.
      See my E-Bike build in progress HERE.

      Comment


        #4
        Verifying Controller Operation...

        O.K. So now you know your throttle is toast and/or you just want to test your controller to see if it will power the motor without a typical hall sensor throttle input.

        After properly securing bike/wheel for testing... here are three possibilities, ranked in order of my preferred preference. Just looking for an easy mid level or lower, motor speed actuation.

        Pie owners be wary of full battery + power in throttle harness! Re-check those pin-outs! See wiring diagram above. 5+vdc pin #1, Sensor pin #2, Ground pin #3. Keep separation with insulation. Do not short.

        Note: During testing on start-up the controller will set a loss of throttle error (#2) as you must always start with no input (no throttle input safety). Then apply sensor voltage input for a couple seconds, then remove for a second, then reapply.
        Well O.K., you could wire everything up. Turn on the power, then remove sensor input for a second, and then reapply. Be prepared to start/stop the input sensor voltage.

        Throttle (test inputs) on the left, controller connections are on the right.

        #1 Use a 3 vdc battery to produce the sensor voltage input. Pros... No danger of accidently shorting 5+vdc power as it's not used. Puts out a nice mid level speed voltage. Can use just a small, compact, "button" type battery. Safe, simple, direct, and smooth.




        #2 One 15 K ohm resistor is series with the 5+vdc supply. Pros... Single part and easy wiring (if any) required. Good mid level speed voltage.




        #3 Two resistors in a voltage splitting configuration. Pros... Good mid level speed voltage. Very accurate, reliable, and steady. Closest to typical current draw.




        Had read that a diode could be used in series with the 5+vdc power... tested. Could get it to work with extra resistor to lower high voltage output. IMO not worth the effort.
        Last edited by Tommycat; 09-03-2018, 09:39 AM.
        See my E-Bike build in progress HERE.

        Comment

        Working...
        X