BionX, want help connecting up this motor

donn said:
lithail said:
This is what I am thinking.

No pedals?

From here, that looks like a really steep head angle, even more than your pictured trike. I don't know if it even matters for a trike, but you decide. See Bicycle Man's paragraph on "Delta" trikes for some examples - https://bicycleman.com/about-recumbent-trikes/.

It's just an example. This was my first try using free cad. its not to scale I was just trying to give a visual of my build.
 
lithail said:
This is what I am thinking.

trike design.jpg

I once built a trailer very much like the back end of that; it worked ok, but I found a better way to use large diameter (26" in this case) wheels, that lets you have a lower deck for better CoG, more space back there, etc. Also, since the wheels are *above* the tubing, you can take a wheel off to work on it without tipping the trike on it's side or jacking it up (helpful if it's loaded up on the roadside).

My Raine Trike design started with a simillar rear end of dual bike frames, but changed to the present design that's proven to work on a big trailer (that can haul a piano):

https://endless-sphere.com/forums/viewtopic.php?f=2&t=80951
Basic frame:
file.php


Mostly finished, seat not bolted on yet
file.php



works, although I'm working on a way to run the pedal chain up to the front wheel without a twisting chain (like the CycleTruck used by UPS), which would eliminate the long somewhat complicated chainline to the rear wheel that it presently has. Then ti will be 3WD, pedals on the front and geared hubmotors on the rear.

You don't have to use the cambered rear wheels, they can be vertical isntead, but the camber gives it a wider track for better stability in turns.

BTW, a larger front wheel will give you a much better ride quality.



The SB Cruiser trike is a better front end thru rider section design, and some day I'l change out the rear end deck frame for the style of the Raine Trike, with larger wheels, but this is a major change to the whole backend, so it won't happen soon (and probably not until I can make a middrive to drive the rear wheels, and pedals run the front wheel, rather than pedals to the left rear wheel only and heavy hubmotors in the rear wheels).
 
I don't know what these things weigh, but it sure looks like a different world than your typical bicycle design, where ounces count. I wonder if that might give you both the freedom and the incentive to add a geared transmission to the electric drive?

Like crank drive / so called "mid-drive", but without the awkward piggyback on the pedal drive, just a simple two or three speed transmission ... I don't have the vaguest idea myself, but I see endless-sphere is loaded with examples that people seem to have done.
 
Unless there is a fairly wide range of speeds needed, or situations where significnatly more torque is needed and other situations where more speed is needed, there's not much reason for a transmission, unless the motor is simply not capable of the torque (or speed) needed by the system.

That said, with a couple hundred watts or few, a simple 3-speed IGH used as a jackshaft, or in the driven wheel if only one is powered by the motor, can be used. There are a number of ways to use a hubmotor to drive a chain, rather than be driven by one.


Regarding weight...80-100lbs for a trike of this design, without any electrics on it, would not be uncommon.

Trikes like my SB Cruiser, with relatively huge batteries and twin heavy hubmotors, big wooden dog-carrier shell, etc., are triple that (or more).
 
Update!!!!

The motor came in the mail yesterday and I couldn’t wait to split it apart just to see the inside.

So I recorded a video taking it apart . I will upload it to YouTube when I get home and post a link.

I have a question about the hall sensors. I had to cut them off the circuit board but they were not labeled. Does anyone know positive , negative, and signal?

I will post a picture when I get home.
 
lithail said:
I have a question about the hall sensors. I had to cut them off the circuit board but they were not labeled. Does anyone know positive , negative, and signal?

This is for a BionX Hall sensor "502A 4117", which is a digital switching/latching sensor. If looking at it so as to read the label printed on the body, the leftmost lead is Vcc, the middle one is Ground/Common, the one on the right is the signal.

I'd post a picture, but what I've got is too high res for ES, so I have to downsample it. I'm just switching over to a Win10 box for this so am still building a toolbox for a basic image manipulation utility. (If anyone has recommendations, I'm listening!)
 
rowbiker said:
lithail said:
I have a question about the hall sensors. I had to cut them off the circuit board but they were not labeled. Does anyone know positive , negative, and signal?

This is for a BionX Hall sensor "502A 4117", which is a digital switching/latching sensor. If looking at it so as to read the label printed on the body, the leftmost lead is Vcc, the middle one is Ground/Common, the one on the right is the signal.

I'd post a picture, but what I've got is too high res for ES, so I have to downsample it. I'm just switching over to a Win10 box for this so am still building a toolbox for a basic image manipulation utility. (If anyone has recommendations, I'm listening!)

Ok thanks. The sensor inside this unit has no writing on it but I will look closer at it. Is there a resistance test to check it?

I use light image resize for resizing photos.
 
My experience with Hall sensors is that as long as you limit Vcc voltage to ~5VDC, you're not going to 'fry' the sensor. If the connections are wrong it just won't work. I couldn't find the data sheet for this particular sensor so I just identified the leads by trial and error. The test method I used was to connect plus and minus 5VDC to any two pins and put a voltmeter one one 'common' pin and the remaining pin. Then pass a magnet near the label side of the Hall sensor and watch the voltmeter. You know when you've identified the right leads by being able to 'latch' the signal line in either a high or low voltage state, alternating between the north and south poles of your magnet. I.e., just dink around until the thing works and then note the connections! :D
 
rowbiker said:
I'm just switching over to a Win10 box for this so am still building a toolbox for a basic image manipulation utility. (If anyone has recommendations, I'm listening!)
at32resizer
it's free, and can do batches, with different (or same) source and destination folders. I just set it to 600pix width / maintain aspect ratio from width, and leave the rest alone, and it makes stuff that fits.
 
rowbiker said:
My experience with Hall sensors is that as long as you limit Vcc voltage to ~5VDC, you're not going to 'fry' the sensor. If the connections are wrong it just won't work. I couldn't find the data sheet for this particular sensor so I just identified the leads by trial and error. The test method I used was to connect plus and minus 5VDC to any two pins and put a voltmeter one one 'common' pin and the remaining pin. Then pass a magnet near the label side of the Hall sensor and watch the voltmeter. You know when you've identified the right leads by being able to 'latch' the signal line in either a high or low voltage state, alternating between the north and south poles of your magnet. I.e., just dink around until the thing works and then note the connections! :D

I have done this method. I could not get it to work but maybe my magnet was not strong enough. I believe from poking at the circuit board with my multi meter from left to right with writing facing out. it goes. signal, ground ,positive.

Does that seem right?
 
I am going to wire it up and try it. the worst that can happen is I have to take it apart again.

What does the hall sensor do? Does it send signal back so the speed controller knows what signal to send to the motor?

I guess I should Have researched that more.

Thanks again for all the great help.
 
lithail said:
What does the hall sensor do? Does it send signal back so the speed controller knows what signal to send to the motor?

The Hall sensors assist your controller in knowing the relative positions of the stator (non moving part, or the 'axle' assembly in the case of the BionX motor) and the rotor (the motor 'shell' containing the magnets). It also helps the controller determine the RPM's of the motor, so it can give 'speed' data.

You can certainly use the motor without functioning Hall sensors by simply choosing to use a controller that can function in 'sensorless mode'. Controllers like Grin's Phaserunner or Grinfineon (see ebikes.ca) can drive this particular motor without Halls because they can figure out the relative positions of stator and rotor using other means. Many users on ES have argued that Hall sensors can make substantial improvements, especially for BLDC motors like this BionX, when starting up under significant loads.

Before closing up the motor, spend an hour looking at youtube vids covering Hall sensors, especially the 'digital' (vs analog) type, since these all work pretty much the same way. Over the years I've 'recycled' many hard disc drives and recovered their neodymium magnets, which I use for testing Hall sensors. Since these magnets are of the type used in the motor itself, you know they're strong enough to trigger a working Hall sensor.

If you do close up the motor again, let us know if you develop any 'tricks' for pressing the two halves together without having the wheel bearing on the non-drive (brake) side damage the wires exiting the motor!
 
Update!

I tried again using another magnet. I believe I had success although none of my readings were over a volt. Should I have got a higher reading? I would say the average reading was 0.3 of a volt . Maybe less

I could get a reading then when I flipped the magnet over the reading would go away. I did it several times. It seems to be working. .

I will solder it back together tonight . I just have to find some signal wire for the hall sensors. I will probably use a USB cable or a Ethernet cable depending on what I have laying around.

Thanks
 
lithail said:
I could get a reading then when I flipped the magnet over the reading would go away. I did it several times. It seems to be working.

It's the sensor's switching when magnetic poles transition that's the critical part. The absolute voltage levels may be due to other issues, such as using the wrong 'common' lead for the signal wire, or something like that. Absent a datasheet for those specific sensors, we can't be positive about the specs, but based on what you've discovered so far I'm inclined to agree with you that "they're working".
 
So I started soldering the wires back together but I couldnt solder the winding wires from the coils. There is to much varnish on the wires from the windings.

How do I clean the varnish off the wires to make a solid connection?
 
Lots of ways:
https://www.google.com/search?q=remvoe+magnet+wire+vanish+to+solder&ie=utf-8&oe=utf-8&client=firefox-b-1

On this type of hall sensors, they are just current-sinking devices, they don't output a voltage. They just ground or not ground whatever signal is on their output pin. Inside the controller are pullup resistors to 5v (or 12v, etc) that provide the actual signal voltage to the MCU. The sensors then ground this volage when they are "on", and let it float to the +voltage when "off". So when you have them not connected to the controller, theyr'e hard to test.

Almost all (maybe all) hall sensors of this tyep, the wriing is the same. If the printing is facing up, so you can see the two bevelled edges, the the lead to the left is +V, the middle is Ground, and the right is Signal out.
 
Update!!!

I managed to get it working. Here is the video on taking it apart.

https://www.youtube.com/watch?v=0lMm6X39-S4&feature=youtu.be
 
Back
Top