No announcement yet.

Need help on 36v scooter

  • Filter
  • Time
  • Show
Clear All
new posts

    Ya, the hall seems to ‘drain’ voltage after throttle release. Drops most of course from 3.62v, but there is a small but long gradual drop to the 1mv. Not like it’s a clean drop to 1mv. Not sure if that makes sense.
    Ill try to get 2 more hands to help me.
    Have not tested from hall negative, but previous test useing board negative showed no voltage change at all until over half way throttle. I’m going to assume it’s the same useing hall negative. Would this happen with your suspect of low starting voltage? I mean it seems nothing at all is going in, low or otherwise, when you pull the throttle until half way .


      Well it could be possible that the board is imposing a parasitic drain on the output of the hall sensor. Something like that would probably show up if you where able to check the current draw of the sensor. ( Which is typically ~6.5mA.) I'd still do all voltage testing directly at the sensor legs. To eliminate any bad connection possibilities. Does the voltage increase instantaneously into the good zone (1-4vdc) with the throttle twisted even a little? And what voltage are you reading when the throttle is half open at the point when the motor starts? And the small drain drop... is it starting from ~.8vdc?

      This may help to keep track of things... sounds like you'd have to add drain drop in tho.

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


        Ok, Was having a lot of problems getting reliable readings. My test equipment maybe having issues. And everything is covered in Vaseline or whatever, makes it slippery and not an easier, LOL. These are solid numbers:

        0%/aprox 50%(throttle kickin)/100%
        0.863v / 1.5v / 3.62v

        so there is 0.863v at rest, and it does go up incrementally as you pull throttle, then kicks in at about 1.5v, and continues until a max of 3.62v

        the drain thing may be an anomaly of my tester. It’s never reading 0v, always a slight value drift when unconnected.


          Now those readings are lot better! If I understand correctly the motor is kicking in at 50% throttle with a reading 1.5vdc. The voltage seems reasonable, but at the 50% position, that's obviously wrong.
          I would expect a reading just shy of 2.2vdc at 50%... Recheck by removing the magnet/throttle assembly totally away from the sensor and see if it hits that reading. If it does, it would indicate that the magnet to sensor positioning is skewed or a possible magnet issue. If it remains low... I would first try to carefully clean all of the slick stuff off from the sensors electronic connections and traces wondering if it could be conducting some electricity in an adverse way?

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


            Ok ill Disassemble the magnet when I get a chance. How can I apply throttle if I’ve removed the magnet?

            Regarding cleaning, does the entire circuit board need to be cleaned, or just the hall sensor area? Would you spray this down with alcohol? Could this cause a short, and do I need to disconnect power again to clean it?
            Last edited by TheNiceGuy; 03-15-2019, 05:16 AM.


              No need to apply the throttle... with the magnet removed (still installed in the finger pull lever assembly-not saying to separate at this time...), and the sensor under no influence of magnetism (gauss) The output of the hall sensor should be reading the middle of the voltage provided range.(~2.2vdc) That's the voltage we're checking on. The motor may or may not start depending on if the controller has throttle protection, be wary.

              Start with that first.

              Only if the middle voltage is still low, mainly clean the sensor and the board around the pins. Always with the power off, when working on the board or it's wiring. Admittedly this is a long shot. As the substance may be just non-dielectric silicone grease from the throttle lever assembly, but worth a try. Seems alcohol would work nicely. Making sure everything is nice and dry before reactivation.
              See my completed Magic Pie V5 rear hub motor E-Bike build HERE.


                Ok! 2.32v with no magnet. And no Writing of any kind on the sensor.



                  A quick recap.

                  Vcc in: 4.3vdc.
                  Closed: .86vdc.
                  Mid with magnet:1.5vdc, Mid no magnet: 2.32vdc.
                  Open: 3.62vdc.
                  Motor starts @ 1.5 vdc.
                  Motor @ 100%: ???vdc

                  With the hall sensor's operation at the 3 points measured with the magnet not present are perfect. Why is the mid point voltage output dropping so low with the magnet installed? Ether the sensor itself is skewing the voltage output inconsistently from negative to positive gauss. Seems unlikely. Or the magnet has unequal magnetic field strengths. Perhaps a bit more likely. I can't say which of these is the answer at this time.

                  More data points would really be helpful to keep track of progress as different remedies are tried. 10 would be great for a proper profile, but 5 would be ok. See Throttle Output and Profiling a little down in the section linked here...
                  Don't worry about plotting them, I will do that as I'd also like to add this information into the thread, as I've never seen an output skewed like this before. (with a hall sensor throttle)

                  This is where after getting the base voltage readings I would start.

                  Verify at which sensor output level the motor is running at 100%.

                  Is the sensor seeing the full sweep of the magnet face? If not making it fully to the left end, a magnet position adjustment would be helpful but difficult.

                  CLEAN the magnet, and hall sensor face and legs as we discussed previously. Then re-check.

                  An easy first change would be to move the sensor face away from the magnet 1-2 mm being careful not to damage the legs, sensor, or connections This will increase the starting voltage or at least the point where the voltage starts to rise. ( need more data points! :-O) Causing the motor starting point to move closer to the front end. See how it works, take readings. Note: this will also decrease the WOT voltage output. Hence the need to see where the 100% motor operation at what voltage is required. (top end dead zone)

                  If good progress is made. Making sure that the legs do not short together. Perhaps a slight twist of the sensor face pointing it more tword the left end of the magnet would help

                  If uncomfortable or seeing that damage may result from bending the sensor legs for repositioning. (you got eyes on it). See below.

                  Some other thoughts...
                  Install a stop on the throttle lever so it can't open all the way. This will eliminate the dead zone, but decrease over-all throttle lever travel.
                  Try another magnet if availible.

                  Lastly there is always the electrical mod. of raising the starting voltage. But would be very tricky to pull off in your case.(no wiring so to speak) And with lack of data could not recommend at this time.

                  With the extra testing perhaps it can be determined if the sensor is at fault?

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


                    Ok, I’ll map it off and make a video also. Just to clarify, mapping is only for the motor engaging range?
                    Edit: ok!
                    i’m not sure how useful this will be, given my caveman resources. But I used a ruler and marked off the usable throttle range which was 8 mm (very roughly center of hall to center of hall). I then (roughly!) marked off 2 mm increments, which gave us a total of five points, first being motor barely engaging, last seems to hit full speed (v):
                    these values were tested on throttle menu #2 speed, the most used.
                    Im sure my 2mm markings are off though!
                    I cleaned all the magnet/sensor, And unscrewed the circuitboard to take a look behind it. I had previously tried various repositions of hall (even touching magnet), but no noticeable difference.
                    I have a pile of new halls sitting here now, so I could possible replace it.
                    One thing of obvious concern is the original solder job is very sloppy of the five wires to the board. The red and white wires actually possibly appear to have a solder touch at one point, and the red wire also appears to be touching a case screw ( although it just screws into plastic with no apparent shorting anywhere).
                    Heres a quick Video:
                    Last edited by TheNiceGuy; 03-18-2019, 07:05 AM.


                      Writing down the hall sensor's voltage outputs at exact 10% intervals from totally off to full on would give an accurate picture of the throttle's performance and possible output problems.

                      In your instance, if the plot is flat until it starts to rise right before the motor starts. It would indicate a magnet problem. But if it has an un-linear, hyperbole type output from the get go. Would begin to suspect the sensor.

                      Disappointing that the change of sensor position didn't help at all. :-( But physical noticeable difference is a long way from actual accurate voltage measurements...
                      See my completed Magic Pie V5 rear hub motor E-Bike build HERE.


                        Ok, I did it wrong then. To get more accurate measurements, I’ll have to make some kind of pre-measured tape to stick on the throttle mechanism. There also appears to be about a 2 mm gap at the top end of the magnet to the hall sensor Edge, while the starting and seems to match almost perfectly. I also noticed the magnet arc moves closer to the hall at the center and farther away at the ends.


                          I feel bad with these measurements, as my marking off on the throttle is not perfect. But I made 10 increments, which gives 11 test points:
                          0.87;0.9;0.98;1.25;1.68;2.22;2.85;3.2;3.5;3.54;3.6 17


                            So looking at the plotted output below.(thank you very much! helps me visualize for sure)

                            Easily shows that there is no flat spot at the beginning, and a fairly linear rise when it gets going. What it does show is an oversaturation of magnetic gauss at the front and back end!
                            Which backs up my recommendation of increasing the distance between the magnet and sensor face. Don't worry about keeping it parallel at this time. Just carefully get some more gap in there and retest.
                            I would try to get the starting voltage output up to about 1.25vdc by adding gap if possible. Which may change depending on does the different speed setting effect the voltage output exactly?

                            Good work! Accurate reading steps and measurements are a plus... the rise of the magnet in the middle in this application I don't think too much of an issue. But I appreciate the observation. Sloppy soldering can be problematic, but with these readings I'm confident your on the right path. (sensor/magnet verses board issues) Please follow up with readings taken after adding gap...
                            See my completed Magic Pie V5 rear hub motor E-Bike build HERE.


                              I think we’ve got this. I’m mucking around with gap/attitude and it’s almost fixed. I’ll post back later.
                              edit: video = 10000 words. Wondering if at this point I should just install a mechanical 3 mm stopgap in the throttle. I think it’s quite usable at this point.
                              Last edited by TheNiceGuy; 03-19-2019, 06:49 AM.


                                Hey, just a heads up. If you make a new post or comment, it will set a flag letting the forum know there is a new post to the thread. Just editing a post does not... hence the delay in response. ;-)

                                It looks much better, but once again not a replacement for good hard voltage numbers. (could use them for my thread...hint, hint.) Are you still getting full top end out of the motor? Curious that if you where to hold the throttle at the past WOT point in the video. Does the motor go to half speed and stay there? As I think it should...

                                If your good with losing just a bit of throttle rotation by adding a small stop, by all means do so! It may be time to quit while your ahead! :-) Way to hang in there. Job well done!

                                P.S. It's still a bit of a mystery to me if the throttle information is getting to the controller thru communication, or you now can see a single wire with variable throttle voltage on it?
                                Can you clear that up?

                                Last edited by Tommycat; 03-23-2019, 06:44 AM. Reason: Added P.S.
                                See my completed Magic Pie V5 rear hub motor E-Bike build HERE.