Neumotor 8057/75 middrive project, belt drive, adding halls, 10kW+

[fechter]

If the hall sensor is located exactly at the middle of the magnet you could have no flux in the direction of the sensor axis (fail). If the sensor axis is off a little from the middle of the magnet, it should work fine.

If you powered a sensor and watched the output while trying different positions you could verify a good signal.

[coleasterling]

It is tested and switching well. The slots are slightly longer than the sensors to allow moving them across the thickness, but I didn't end up needing it. Controller testing will happen tomorrow or Tuesday and we'll see for sure. I'll probably throw the scope on it, too.

[coleasterling]

I realized it was somewhat difficult to tune the hall position due to some dumb dimensioning. I reworked the part today for more adjustment range (On gap? Middle of tooth? Somewhere in-between?) and better centering. Don't quite get the relationship between initial hall position and termination yet, other than being 30 degrees different.

Anyway, I made the ID of the ring smaller than the motor mount's hole (instead of larger) and turned a tight-tolerance insert that allows me to tune the hall position without worrying about concentricity until the mounting screws are tightened.The screw in the end of the insert lets me pop it off when tuning is done. I didn't get it fully wired up today since I made the design revisions, but will have it running tomorrow.

[coleasterling]

Made yet another revision, this time to allow building of the hall-sensor module off of the motor, and for much better wire routing.

This is after tumbling it to deburr in our 4-foot diameter vibratory tumbler with polyester tetrahedrons.

Edit: I can't change the orientation of the image for some reason.

[nowholeself]

Made yet another revision, this time to allow building of the hall-sensor module off of the motor, and for much better wire routing.

This is after tumbling it to deburr in our 4-foot diameter vibratory tumbler with polyester tetrahedrons.

Edit: I can't change the orientation of the image for some reason.

I want one!

[nowholeself]

My Neumotors 8025 has just arrived. It is so small! My original plan was to attempt to at least match, if not beat, the power of an ASI BAC800 controlled BBSHD. But can something so tiny really do that!?!? Maybe I should have gone with the 8038...

[fechter]

If you apply a steady current to a pair of phase wires, the rotor will align to a certain spot and this will be exactly where one of the halls should be right at the switching point. By jiggling the rotor a little, you should be able to dial in the position so the hall signal toggles, indicating the timing is neutral. This would be a good starting point for the timing.

[coleasterling]

Ah, makes sense. Manual version of setting a digital encoder or resolver.

This solution works quite well...when the halls are functioning. Twice now, I've had the bike running well, then come back and the halls were all shot. Not sure why. They definitely aren't touching anything and there's no physical damage to the halls, mount, or hall harness. It wasn't getting hot enough to kill them, either. Color me confused.

Anyway, nowholeself, if you did want one, I'm happy to send the last one I took photos of to you. I realized there wasn't enough clearance to the stator windings with it, so it couldn't move as much as I wanted. I think the easy solution if you did want it is to just bend the legs slightly out. I haven't been able to tell any difference in function with a bit of angle on the halls. I can't really envision the 3d flux path that makes it work in the first place.

[madin88]

Twice now, I've had the bike running well, then come back and the halls were all shot.

Thats weird. hope you find out what the problem is.

It wasn't getting hot enough to kill them, either.

The holder is made of aluminum so there will be eddy currents, but i hardly think that this small surface area behind the halls would heat up so much to toast them.

[coleasterling]

The holder is made of aluminum so there will be eddy currents, but i hardly think that this small surface area behind the halls would heat up so much to toast them.

Thanks, they definitely weren't significantly heating. I really don't know. Now, I'm not running resistors, but I never have on ss411a's and they've been just fine, even with this exact controller. Strange deal. I'm out of halls now, but might buy more and wire the resistors just to check.

[Okami]

Hi, very nice and detailed build log.

What caught my interest was -

U said it is is registered as moped. Did u have to go individual test or u had different approach how to make it road legal?

At least it doesnt seem to me like it was commercially made ebike/frame from start but I might be wrong, just seems like u did so many 'mods' that it was custom made from start.

[ARod1993]

I realized it was somewhat difficult to tune the hall position due to some dumb dimensioning. I reworked the part today for more adjustment range (On gap? Middle of tooth? Somewhere in-between?) and better centering. Don't quite get the relationship between initial hall position and termination yet, other than being 30 degrees different.

Anyway, I made the ID of the ring smaller than the motor mount's hole (instead of larger) and turned a tight-tolerance insert that allows me to tune the hall position without worrying about concentricity until the mounting screws are tightened.The screw in the end of the insert lets me pop it off when tuning is done. I didn't get it fully wired up today since I made the design revisions, but will have it running tomorrow.

Quick question; do you have a more complete CAD drawing of the motor than what's on the website, and if so would you be OK posting it somewhere? I'm playing with it in Solidworks using a dummy model based on the dimensions on their website, but if you have an official CAD model that would be awesome!

[madin88]

Hi, very nice and detailed build log.

What caught my interest was -

U said it is is registered as moped. Did u have to go individual test or u had different approach how to make it road legal?

At least it doesnt seem to me like it was commercially made ebike/frame from start but I might be wrong, just seems like u did so many 'mods' that it was custom made from start.

Yes that is Votec VF 195 DH frame. It was not an ebike.
To get vehicle papers for registration i had to make everything street legal (lights, reflectors, automatic stand or one with a switch which disables motor, mirror, horn, turn signals etc) and then they asked for approval of electric (EMI, EMV tests in lab) and test of braking system for 45kmh.

That was already a few years ago and now i was told it is more complicated, maybe they want tests of the frame idk..

[nowholeself]

Ah, makes sense. Manual version of setting a digital encoder or resolver.

This solution works quite well...when the halls are functioning. Twice now, I've had the bike running well, then come back and the halls were all shot. Not sure why. They definitely aren't touching anything and there's no physical damage to the halls, mount, or hall harness. It wasn't getting hot enough to kill them, either. Color me confused.

Anyway, nowholeself, if you did want one, I'm happy to send the last one I took photos of to you. I realized there wasn't enough clearance to the stator windings with it, so it couldn't move as much as I wanted. I think the easy solution if you did want it is to just bend the legs slightly out. I haven't been able to tell any difference in function with a bit of angle on the halls. I can't really envision the 3d flux path that makes it work in the first place.

Thank you for the offer. Talia from Neumotors mentioned to me a while back that they are now installing hall sensors on their 80-series motors. A little over a month ago, she offered to have the company install some on my 8025. I took her up on the offer, and immediately sent it in. But there hasn't been any response since. Their products seem great but the communication there is sluggish to say the least. Still, factory-installed halls would be so much easier.

[nowholeself]

Neumotors 8025 with factory installed hall sensors.

WIN_20200719_13_58_28_Pro.jpg (268.89 KiB) Viewed 644 times

[madin88]

Neumotors 8025 with factory installed hall sensors.WIN_20200719_13_58_28_Pro.jpg

Thats nice to see. Did you spin it up already and does it run smooth?
Would you mind to share more photos of the hall board (with the rotor removed)?

[larsb]

Those hall wires look fragile and no strain relief. That’s only halfway decent. Can it be adjusted?

[nowholeself]

Neumotors 8025 with factory installed hall sensors.WIN_20200719_13_58_28_Pro.jpg

Thats nice to see. Did you spin it up already and does it run smooth?
Would you mind to share more photos of the hall board (with the rotor removed)?

I'm still fussing with the ASI controller. So far so good. Here are a few photos with the rotor removed.

IMG_20200722_204308586.jpg (330.83 KiB) Viewed 584 times

IMG_20200722_204331589.jpg (344.89 KiB) Viewed 584 times

IMG_20200722_204339469.jpg (355.32 KiB) Viewed 584 times

[fechter]

The hall sensor board looks great. The whole thing should be potted if you want to operate in wet weather. Even conformal coating would be better than nothing.

Can you estimate the clearance from the top of the hall sensors to the rotor? I'm assuming it's pretty tight.

[coleasterling]

Little bit more elegant than mine. I wish I had thought to cut the housing, would have made it much easier.

[larsb]

that's a lot better looking up close, still easy to crack with only a slight bump. did they charge extra for the hall option?

[nowholeself]

that's a lot better looking up close, still easy to crack with only a slight bump. did they charge extra for the hall option?

They charged $70 to install the halls on an engine I already had. I suspect they'd charge less to sell an engine made that way from the beginning, though I don't know for sure. I agree that the hall sensor board is a bit too brittle and exposed to be usable in most circumstances. Fortunately, I have it hidden behind the mount on my setup so I'm probably alright. Fingers crossed.

[larsb]

one could 3d-print a cover for it, would look nice if done well

[nowholeself]

I was hoping to give an update on how well the halls work. But so far I've had little luck getting this engine to work with the ASI BAC800. I can get BacDoor to spin the motor up to around 3500rpms with a 7759rpm setting using 'motor discovery' in both unsensored and sensored mode. But trying to spin the engine with anything greater than a 7759 rpm setting results in a glitch. So does the use of the throttle. The engine starts to spin up but then stops abruptly. Sometimes it'll spin fast in the wrong direction for a split second before stopping. But it never runs at the appropriate RPM setting or with the throttle. This is the case even using a 48volt as opposed to 56 volt pack. I've also checked and mixed-and-matched the phase and hall wiring. Anyone have any ideas?