Regarding hall effect sensor throttles...

[RAI]

I have an old scooter that ran well with a hall effect sensor throttle, at least until my son crashed it into a hedge.

The throttle wires broke, so I took it apart to repair it. Whle doing so, I broke the wires off of the sensor. So, I ordered a couple replacement sensors (UGN3503U shipped from Hong Kong). They're smaller than the original, but same part number. I carefully soldered one in the manner that it apparently should've gone in; but I could not get it to work. Waving it around a magnet and moving it around in the throttle, I could barely get the output voltage to budge by 0.1 volt or so. I swapped pins 1 & 3 (+5 and output) but that didn't do much at all.Thinking that I'd perhaps burned it out by initally have 1&3 swapped, I tried my second part, with 1&3 swapped from the first go around, then when that did nothing, I swapped 1&3 yet again. So now, I probably have 1 broken sensor and 2 burned out sensors… Scooter is still not working… Laddie not happy...

SO - I decided to look for a whole new throttle. The existing throttle looks identical to several Chinese hall effect sensor throttles that are for sale on many websites (about $20USD). HOWEVER, while I was trying to get my repairs to work, I noticed that shorting the output to GND makes the motor run full throttle, while shorting the output to +5v (or leaving it disconnected) leaves the motor idle (stopped).

This is where I hope that some experienced chaps here can assist… (I'm assuming that pedelec throttles are very much like electric scooter throttles.)

Isn't this backwards? From the info that I can gleen from the web, the voltage for full throttle should normally be up at +4volts, while +1volt should be stopped. My controller is expecting the reverse of that. Low (i.e 0)volts is full throttle, while high (+5) volts is stopped. Which way is, in fact, "normal"? And is my throttle "normal"?

I believe that it was flecc that said, in another post about a broken throttle, that flipping a magnet (or two) in the throttle can change the "polarity" of the throttle. It seems that if I purchase a new throttle, I may have to do this flipping. If so, can someone tell me how to open up this type of throttle to flip the magnet(s)? I can't see how to open it up.

My throttle looks exactly like this one: Scooter Throttle, Electric Scooter Throttle | Monster Scooter Parts

I'm also curious if anyone (flecc?) has any insight into why my hall effect sensor replacement efforts were such a dismal failure. I was very careful, used a very small soldering iron and surgical forceps on the leads to keep the sensor from over-heating.

Thanks in advance
& Cheers,
Roger

 

[flecc]

It was d8veh who posted about the magnet reversal which reversed the Hall effect on the sensor Roger.

There's a very wide range of Hall efect sensors with several characteristics varying so that makes it difficult to guess at a suitable one. E-bikes mainly use a narrow band of these, but from other's postings it seems that scooters and mobility scooters often use widely differing ones.

On the e-bikes where I've known the throttle characteristics, it has been full voltage for full throttle, so your reverse situation seems odd and I have no ready solution. Perhaps someone else (d8veh?) will have an insight into a way to move forward on this problem.

 

[RAI]

I've managed to figure out how to open up my broken throttle. I used four bamboo (shishkebob) skewers to release four plastic spring tabs inside it. The magnet is very small (about 3x3x8mm) and it's mounted 'vertically' on a piece of plastic that looks like it was also the rotation stop. When my son ran into the hedge, he must have twisted the throttle forcefully forward, breaking off the plastic tab (and magnet). That doesn't fully explain why my sensor replacement efforts were so unsuccessful, but it's part of the puzzle.

I've been emailing around asking the same questions. I found a place in Idaho that convincingly has the correct throttle that I need, but they want $30 plus shipping.
2 or 3 other places agree that my throttle is backwards, but don't sell what I need
While 1 place basically says don't worry, buy ours and it'll work

I found one place that only wants $15 w/shipping for the same throttle that the others are asking for ~$22 w/shipping. SO, I'm going to give that one a go, and try to flip its magnet around.

I hate to say it, but if that doesn't work, I may just give up. With Christmas approaching, I have many other things I should be doing. I got the scooter for free from a neighbor. My only regret will be that I spent $75 for new batteries for it. (But I should be able to re-purpose them to one of the computer UPS's in my house.) Of course I'd like to get it working again, but my son tends to be hard on things, so 1) he may just go out and crash it again, and 2) it might be a lesson to him to lose the use of his scooter due to his crashing it. I've also had to buy a new charger for it (he lost the old one), & had to rewire the connector for it, repaired its power switch (which he broke in another crash), spent for the hall sensors that didn't work, and will have to get new tires soon as they're getting 'bald'. I'm getting weary of keeping this "free" scooter running.

My son and I thank you for your help, maybe d8veh can add something, too.
Roger

 

[mike killay]

Roger,
Your nickname isn't 'Lucky' by any chance?

 

[flecc]

I hadn't mentioned that these Hall effect throttles come in two magnet arrangements. I think yours may be a one magnet type from what you've posted.

Those I've seen have been mostly twin magnet, one at each end of the throttle rotation arc, but there are some with a single curved magnet as well.

I hope you are lucky with that one you are risking trying.

 

[Deleted member 4366]

I think Flecc's covered most of it. Firstly the broken throttle. Is it possible that it had two magnets and one fell out when it got broken? I've broken a couple of throttles while I was riding at full throttle and went over a bump. The plastic stops were not very strong. so they got ripped out along with a magnet. As these throttles broke in the same way as yours, it's possible that they're the same. There were two magnets - one on each side of the hall sensor. One magnet is oriented north and the other south so that the magnetic field gradually changes at the hall sensor when you open the throttle. To reverse the throttle direction, you only need to flip both magnets so that it goes from south to north. Not all throttles are the same, so until you get one you can't be sure that you can do it.
There's a couple of things you can do: if you can wait a bit, get one of these cheap throttles. These are the weak ones that I had that definitely have two magnets in. They make take a couple of weeks to arrive but are very cheap.
e-bike e-scooter Twistgrip Throttle(pair) [BHZB-008] - $3.00 : EV Assemble, All for Electric Vehicle!
Or, as a short term measure, you could make a potentiometer throttle, which would work more like an on/off switch but with adjustable power. You can get a 10K pot from Maplin for a couple of pounds plus a little box to mount it in for 99p and then you need a push-button or toggle switch that you can mount near the grip on the handlebars so that you can operate it. The potentiometer does the same as the throttle: You connect the red (5v) and the black (0v) to the outer two pins and the third (signal) wire to the middle one. As you turn the shaft, it'll work just like the throttle. If it works backwards, just flip the red and the black wires.. If you have a look at this photo, you can just see my one on the left handlebar. It's the little black box with a shaft sticking out (before I got a proper knob for it). My switch on the side of it switches between the normal throttle and the pot so that I can use the pot as a cruise control, but you could use the switch to switch it on and off instead. In fact you can even buy pots with built in switches like the volume control on your radio. Put the switch needs to be between the red wire and the pot.

..

 

[RAI]

My $15 hall throttle arrived yesterday and I gave it a go... so I'm back with news

Mike K. - nah, no one calls me lucky... in fact, my lack of luck in this matter continues

flecc - might be two magnets... didn't see any curved magnet. I'm thinking that the sensor basically works on seperation distance from a single magnet.

d8veh - Thanks for the link on the pot throttle. Unfortunately, although it's only $3, shipping to the states is $29! Perhaps I can find it closer to home.

As you may have guessed by now, the new throttle isn't working any better. I had to cut and splice the old connector onto the end of its cable. Anyway, it's acting just as dead as my other attempts.

I'm really getting befuddled - and I'm an electrical engineer!

With the new throttle connected, the 'output' voltage is stuck at about 4.2volts. The power is dead on at 5.0 volts. Twisting the throttle handle has no effect. I was expecting it to work backwards, but it does nothing.

I was starting to think that my controller died, but with the throttle cable unplugged, shorting the sense wire input to gnd still causes the motor to run full speed. I put my meter in current mode and the the current is 0.5mA, which makes sense if there's an internal 10kohm pull up resistor to 5v. I believe that the hall sensor has an open collector output (i.e. a transistor that pulls the sense voltage down towards 0v, & lets it get pulled back up by the 10kohm resistor). My point is that the controller input seems to be fine and it agrees with the sensor having an open collector output.

So WHY doesn't the output from the sensor actually pull the voltage down based upon the throttle handle position!!!! ARGH!

I may try digging out an old pot from my old parts boxes. Unfortunately, I think most of the ones that I had were either 10-turn or log taper. Maybe I'll get "lucky". I have a suspicion that maybe I'm misinterpreting the wires(?) so fiddling with a pot might make that clearer. It's got to be some stupid simple error (I hope).

If that's the case, and I can get this hall throttle to respond, then I may come back to flipping the magnets around. If not, then maybe I can find that $3 throttle nearby.

Thanks all - wish me luck!
Roger

 

[Deleted member 4366]

Just a thought: Is it possible that the signal wire and ground wire are reversed? I think we had one before where that happened. The Chinese are not very good at getting the Colours wired to the right place, so it's possible that the black is the signal and the white the ground from the controller. Would it give that effect? If you can find a pot, you should be able to test how this is working. You can use a log pot too - it doesn't make that much difference.

 

[RAI]

The throttle came with a sheet of paper detailing the wire colors and it has a red stamp: "CHECK PASSED" or something like that. So, I would hope that the wire colours aren't swapped. Of course they could just put pre-stamped papers with every unit without actually checking them...

I was also thinking that the signal and gnd might be swapped; but it doesn't quite fit with what I'm seeing. I might go thru all the possibilities on paper to see if any combination of wires agrees with my observations.

The controller wires are red, yellow and blue. I think the old throttle wires matched. The new throttle wires are red, black and green. Red seems to be consistently 5v. You'd think that they'd be able to consistently make gnd a black wire...

 

[z0mb13e]

The throttle came with a sheet of paper detailing the wire colors and it has a red stamp: "CHECK PASSED" or something like that. So, I would hope that the wire colours aren't swapped. Of course they could just put pre-stamped papers with every unit without actually checking them...

I was also thinking that the signal and gnd might be swapped; but it doesn't quite fit with what I'm seeing. I might go thru all the possibilities on paper to see if any combination of wires agrees with my observations.

The controller wires are red, yellow and blue. I think the old throttle wires matched. The new throttle wires are red, black and green. Red seems to be consistently 5v. You'd think that they'd be able to consistently make gnd a black wire...

Might be easier just to buy a hall effect transistor so you know the pinout and specs...

Hall Effect | Farnell United Kingdom

 

[Deleted member 4366]

Out of the yellow and blue, how do you know which is ground? Did you check them?

 

[RAI]

Had other stuff to attend to, but finally got back to this.

Zombie, I originally tried using a new UGN3503 replacement hall effect sensors in the old throttle. I followed the spec sheet, but it didn't do anything. Grasping at straws, I started swapping wires around, but still no luck. Turned out that the magnet wasn't where it should've been. With the new (suspected backwards polarity) throttle, still no luck - no response.

d8veh, for the yellow and blue wires, using my volt meter, there's 5v from red to yellow and 4.97 from blue to yellow. This is measuring on the controller end of the cable with the throttle disconnected.

I dug out an old 5kohm pot and put it across the controller end of the cable. With the pot across red and yellow, with the wiper on blue, the motor runs fine, and the voltage sense is still backwards. High voltage is stopped, low voltage runs fast. The odd thing is that there's only a 0.5 volt range from stop to full speed; and it's offset, too. Specifically, 3.0 volts is full speed, and 3.5 volts is full stop. Very narrow range. This is with the wheel off the ground though -- maybe it has a wider voltage swing under load.

Still, this doesn't explain why the output voltage on the new throttle that I bought, doesn't change. It confirms that the controller is probably still okay (caveat on the voltage range). And it confirms that I didn't have any wiring misunderstanding. Red is 5v, Yellow is Gnd, and Blue is control voltage.

I searched for several hours, trying to find a local source for that $3.00 resistive throttle from EV Assemble. I thought that place was in England... It's actually in China. Shipping that throttle to England is still 20GBP (plus 2GBP for the part itself). Other resistive twist grip throttles are available here, but for some reason, they're like $50 or more (all seem to be "Magura" replacements). Compared to those, the one you suggested is still a bargain, but I can't bring myself to spend $31 for shipping a $3 part.

I have one last thing to try before I throw in the towel, take the SLAs out and drag this old scooter out to pasture (so to speak). I'm going to take apart the new Hall Sensor throttle and make sure that I see nothing wrong with the magnet, sensor, etc. IOW, see if I can figure out why it's not responding. Perhaps the magnet simply fell out and isn't where it should be.

<sigh> Roger

 

[mike killay]

Roger,
You say all this started after a crash.
Have you checked that all the other parts are OK?

 

[Deleted member 4366]

The voltage range seems very small. Why not order a new controller as well as the throttle from BMsbattery. The postage will be the same, so you'll get the throttle and controller for about £30 (guess) and your problem will definitely be solved.

Or buy a whole kit as well as the controller and throttle. Sell the kit on Ebay or to a friend,and you'll probably make enough profit to get the new controller and throttle for nothing, or you can build yourself a nice Ebike - that BMP kit is tempting!

 

[RAI]

Mike, This is a stand up, electric scooter, so there's not much to break on it. Except of course, the throttle, which is (after inspection) quite fragile/vulnerable in a crash.

I managed to wrestle the new throttle apart. It does, indeed, have TWO magnets in it. They're mounted about 90 degrees apart, so one or the other of the magnets is near the hall effect sensor at the rotation limits. I'm convinced that they sent me a bad throttle. Its sensor just is not responding to the position of the magnets. It's all wired fine, with no breaks in the wires nor thru the connectors.

d8veh, I looked at controller alternatives. Seems that e-bike controllers are a fair bit different than scooter controllers. I'd been hesitant to go this route because I couldn't determine what the motor specs are. This scooter is about 8 years old and obsolete. There's hardly Any information on it Anywhere. I Could replace the controller (I'd guess a 250watt model would work) but I need a new throttle, so it'd cost me $30 ~ $40.

We have a thing here called Craig's List, it's like a bulletiin board service where you place classified ads. I checked and there are several scooters for sale by owners. Razor brand, going for $50 to $120. I'm more inclined to buy one of these and start off fresh. These are only a few years old and parts will be more standardized and common. I'll end up buying new batteries again, but that's pretty painless. So, that's probably what I'm going to to.

I'm sure that I could continue to work on this old scooter and get it working (either the new controller route, or try another throttle) but frankly, I'm tired of looking at this old thing.

Thanks everyone, for your comments and suggestions.
Cheers,
Roger

 

[mike killay]

Roger, I know how you feel, I began to strip my 20 year old pedal bike down with a view to repainting. I found that I could buy a brand new one in Tesco's for less than £100. Worse, these new bikes have full (cheap) suspension and alloy wheels.
Sometimes its just as well to let go!

 

[xyzzy]

Hi. Just been reading through this thread as a friend of mine has a very similar problem. He bust his twist grip and ordered a replacement. The replacement operates in reverse, i.e. at rest the motor is at full speed and when it is twisted fully round, the motor is at rest. The grip was returned but another replacement operates in the same way.

I wonder if RAI ever resolved his issue or anyone else has any ideas?

BTW does anyone know why the Chinese call the twist grip control, the steer?

 

[flecc]

Hi. Just been reading through this thread as a friend of mine has a very similar problem. He bust his twist grip and ordered a replacement. The replacement operates in reverse, i.e. at rest the motor is at full speed and when it is twisted fully round, the motor is at rest. The grip was returned but another replacement operates in the same way.

I wonder if RAI ever resolved his issue or anyone else has any ideas?

BTW does anyone know why the Chinese call the twist grip control, the steer?

You need to reverse the magnets within the throttle. Once you've got it's housing apart, often four claws internally to release, turn each magnet round individually to change their polarities.

 

[xyzzy]

Ok, thanks for that; we'll give it a try and report back.

 

[thomas_d.b]

Ok, thanks for that; we'll give it a try and report back.

what you might really need is a 10 kilo-ohms pulldown-resistor between the 0 volt ground and the signal wire of the hall sensor , cause your signal wire is floating at 5 volt and overrules the output of the sensor itselve, wich is lower. this fixed my goldenmotor kit to be able to support the pedal assist