Hub Motor, Throttle Won't work?

[furcifer]

So, buddy gave me a fat tire bike with a 350W hub motor that wasn't working. So I order a controller and of course almost everything is backwards male where female should be or black where red should be.

But NBD, the display connector, the throttle and the phase sensor all lines up. The phase wires didn't but I got them figured out.



with it connected like this the throttle won't work, I get an error showing on the display. The walk mode works and the motor will spin 11 km/h. If I connect the two green wires together (I don't know what they are called, it's some sort of bypass/reset wire) It will run full speed. (green wires right above the 00 on the display)
But when I disconnect the green wires and hook up the throttle I'm back to the error and the motor will not run.


Does anyone know what this could be? I bought a new throttle and the voltages check out. I tried changing the phase sensor wires to match the phase wires (grn-bl, ylw-grn, bl-ylw, you can make out the harness behind the phase wires in the first pic.) and it just won't work. I don't have any of the brake sensors connected but I have tried them (swapped the pins in one) and nothing changes.

None of this makes sense to me. I have power, the motor runs and input from the throttle tests good 0.8-4.6V between sensor wire and ground. I don't have the 3 speed, horn or light hooked up because this controller doesn't seem to support it.

The only thing slightly out of place is the throttle connection. The throttle connector is in the first pic covering the controller. It's white on the controller side and black on the throttle side. I don't think this is a problem, I have black to black and red to red, with the controller signal wire being grey and the throttle signal being green. When the throttle is pressed it's a positive (+) voltage increase in the signal wire. Based on wiring diagram provided (which is sketchy at best) and my knowledge of electrical this is functioning properly. But it's not functioning properly. The only way to get the motor to spin is either using "walk mode" or connecting the green wires on the controller.
Any ideas would be much appreciated. This bike is a week away from being sacrificed in a Labor Day ritual and ghost ridden into the Detroit River.

 

[furcifer]

Would a bad hall sensor cause this error? I tested them before I got this new controller and they spec'd out but maybe there's a short somewhere.

 

[TommyCat]

As you say, walk mode and self learn mode both work well. So I would think it unlikely.
(The OP error…)

 

[E-HP]

Would a bad hall sensor cause this error? I tested them before I got this new controller and they spec'd out but maybe there's a short somewhere.

Can you describe the exact steps you used when running the self learning routine?

etc.

 

[docw009]

I am concluding that you're not connecting to the throttle connector. Is there a three pin you are avoiding because it's the wrong gender?

Find the power and ground on all the three pins. The third pin is the signal. The red signal on your throttle and connect it to one of the +5V power leads you found, Doesn't matter which one, as they are all the same. Don't worry about the throttle ground. Just rotate the throttle signal wire to each of the signal pins. One of them will spin the wheel at max, as this simulates a throttle fault,

This is what I do when I pull an old controller out of my box and don't remember which of the three pins is throttle. Hook up power/motor and display, Turn it on, Instead of using a throttle, I use a diode. Acts the same.

 

[furcifer]

Can you describe the exact steps you used when running the self learning routine?

1. 
   

etc.

1. Turn on battery power switch.
2. Connect the green wires.
3. Wait

I might be wrong but if I swap the phase wires and connect the greens it sounds like it runs backwards for a couple seconds and doesn't move then goes forwards full throttle.

 

[furcifer]

I am concluding that you're not connecting to the throttle connector. Is there a three pin you are avoiding because it's the wrong gender?

Find the power and ground on all the three pins. The third pin is the signal. The red signal on your throttle and connect it to one of the +5V power leads you found, Doesn't matter which one, as they are all the same. Don't worry about the throttle ground. Just rotate the throttle signal wire to each of the signal pins. One of them will spin the wheel at max, as this simulates a throttle fault,

This is what I do when I pull an old controller out of my box and don't remember which of the three pins is throttle. Hook up power/motor and display, Turn it on, Instead of using a throttle, I use a diode. Acts the same.

Ahh, OK. I've been looking for a way to bypass the throttle. If the signal wire is the red one I may have been screwing this up. But I can find the +5V no problem and do this.

Yes, the black, red and blue is the odd man out of the 3x 3pin connectors.

 

[furcifer]

As you say, walk mode and self learn mode both work well. So I would think it unlikely.
(The OP error…)

I'm getting mixed messages on the learning mode. Most say the learning mode doesn't require fully functioning hall sensors and others say you do.

There's less information on the Walk mode.

It seems like there is infinite variation in these controllers so there's infinite wiring diagrams.

 

[E-HP]

I'm getting mixed messages on the learning mode. Most say the learning mode doesn't require fully functioning hall sensors and others say you do.

There's less information on the Walk mode.

It seems like there is infinite variation in these controllers so there's infinite wiring diagrams.

Self learning starts as a sensorless operation as jerks around getting oriented with the stator position, but can’t complete without working hall sensors. Walk mode needs halls in order to run smoothly at low speed, where sensorless struggles.
Are you following the self learning procedure that came with tour controller? Most start with the battery disconnected, connect the self learning wires, plug in the battery, turn on the controller, and wait. If it spins the right direction, unplug the self learning wires and power down. If not, unplug and replug the self learning wires.

 

[furcifer]

Self learning starts as a sensorless operation as jerks around getting oriented with the stator position, but can’t complete without working hall sensors. Walk mode needs halls in order to run smoothly at low speed, where sensorless struggles.
Are you following the self learning procedure that came with tour controller? Most start with the battery disconnected, connect the self learning wires, plug in the battery, turn on the controller, and wait. If it spins the right direction, unplug the self learning wires and power down. If not, unplug and replug the self learning wires.

lol, if by "learning procedure" you mean a single green wire labelled "intelligent identification" and deciding to plug the other green into it and waiting to see what happens then yes, yes I did. (I think in the other wiring diagram I have it's called a "self programming")

I don't think geared hub motors spin backwards, my understanding is they have a clutch. That's why I say it sounds like it's running but the wheel doesn't move, then it starts up, full speed.

I think I have followed that procedure inadvertently. I have been turning the battery on before connecting the greens, but subsequent to that I've shut it off with the greens connected and turned it back on. I've done so much fiddling with the phase wires I'm not 100% sure so I'll follow that more exact procedure and see if it changes anything.

 

[TommyCat]

If the signal wire is the red one I may have been screwing this up

I am sure Doc did not mean it that way! It would be very unusual for the throttle signal wire to be RED.

But we both agree that you’re not connecting your throttle to the correct 3 wire controller group.

I've been looking for a way to bypass the throttle.

If you can identify each of the controllers 3 wire groupings separately into power, ground, and signal wires.
I recommend connecting a 3 volt button cell battery just to each groups signal and ground wire, one group at a time. Starting with my best guess of the RED,BLACK, and WHITE female connector bunch.
The positive side of the battery will go to the signal wire. And the negative side of the battery will go to the ground wire.
The 5vdc positive power supply wire goes unused.
This test has the advantage of not taking a chance at shorting your valuable controller’s 5vdc regulated supply out.
And provides a sure, safe, midrange voltage for testing without using a throttle.

Some controllers may hesitate because of the throttle no voltage on startup. But if you have it wired so that all you have to do is make the final touch to button cell positive. You can “bump“ the signal a couple times and get it to start after first powering up your system.

For more details on identifying the correct controller wires, as well as the throttle’s wiring. And using the button cell battery to verify controller operation, see this thread…
Guide to Hall Sensor Throttle operation, testing, and modification. - Electricbike.com Ebike Forum

The safety recommendations in there are always a worth while read!

Any concerns or questions, please ask before experimenting if not sure.

 

[furcifer]

Starting with my best guess of the RED,BLACK, and WHITE female connector bunch.

Good call. I found the wiring diagram finally and it lists the r,b,w as the throttle. The way they bunch things up and label things in super small print threw me off. Almost all of my testing has been wired to the PAS. When I go back to the shop I'll use that one to run all these tests on. Hopefully I didn't burn out the 5V supply. I think I'm good, the error comes on when I press the throttle so there must be some voltage still.

 

[furcifer]

Solved.

Thanks to everyone for all the help.

Once we identified the throttle for sure I hooked it up and it threw the ! error. The (+) and (-) wires were swapped in the connector. I swapped them "color to color" and it started right up.

I suspect most of my problem was trying to use standard voltmeter probes. They're way too big for these JST SM "micro connectors". False readings because I thought I was making contact when I wasn't threw me off course.