My electric bikes & projects

[Stubby79]

Blue bike found a new home a day or two after being fixed. Red is waiting on parts.

Scooter/moped upgrade time! Brought this "fixer-upper" home this morning:



Doesn't appear to need much. New batteries, a few minor bits and bobs. Better shape than the one it's replacing started out.

Much more modern, and better quality. Hydraulic disc brake on the front, drum on the rear, more solid frame. Still Chinese, but made by a company that mostly builds gas scooters rather than cheap e-bikes. Hopefully it lives up to expectation!

[Stubby79]

It does!

It's not quite as strong as I'd like, but that's probably the half-done lead acid batteries I was using. I'll find out for sure once I reconfigure my lithium pack to fit!

But, making up for it, it turns out that it has regenerative braking! I'm thrilled!

That might not seem all that note worthy, especially with the sturdy brakes on it, but it was one big let-down on my previous one. And good luck buying an after-market controller affordably that will do regen.

Despite the lead, I got it up to 40km/hr on level ground, in "third gear". It has 3 virtual gears, and a switch to choose them. And a mechanical speedo, so it's not completely out to lunch with the reported speed. Bench tested/no-load, with my lithium pack (~56v), it was spinning at ~50km/hr. Not bad for a 500-watt motor!

[Stubby79]

Reconfigured my lithium pack to make it shorter, and swapped out two of the individual packs for healthier one, so I can get the full 21+ amp-hours out of it.

Rebuilt the battery tray to be 3" longer and 2" wider, out of old 1/4" ABS(?). Took longer than expected.



Now it will hold either chemistry of battery belowthe under-seat storage.



I will make a false bottom for the under-seat storage out of a new piece of 1/4" ABS I have laying around. And I will re-route the breaker, if it will fit underneath.

That's the big stuff out of the way.

[Stubby79]

Rusty chain...


CLR...


After...


Excessively long wiring between the battery and the controller...


A good 3 feet removed...


Breaker relocated...


Access hole...


False bottom made for under-seat storage...


And, just for the heck of it, smoked out front signal covers...


That makes up what I got up to last week.

[Stubby79]

Today...since I can't leave well enough alone...

12 mosfets in controller. Good stuff. 63V rather small capacitors, not so grand, but not really an issue either. Thankfully, right there when you open it up, the shunt...


Add a couple of wires...


Run them out of the case, and add an insulated paperclip between 'em...


And you have a "shunt-mod" that you don't have to open the case and re-solder to adjust the resistance!

[Stubby79]

Picked up a 1500w "sensor-less" compatible electric bike controller off fleabay for ~$50, so that I could try running an alternator as a motor...

The crude setup, using a common/spare alternator I had laying around, with the voltage regulator removed:



It works!

http://www.youtube.com/watch?v=64lt7WMgn8g

I was varying the speed via an ebike throttle in the video, just to show it was variable.

Max RPM is determined by the battery voltage, of course, and, more modestly, by the voltage being fed to the rotor/field. For example, feeding it 80v with 10v at the field, it ran at 4700rpm. With 80v and 5v on the field, it ran at 5000rpm. You can get ~3600 rpm off of 56v(48v lithium) with 5v on the field, a typical rpm for a lot of motors.

No-load efficiency falls off based on two things...the RPM of the motor, and the field voltage; the higher in either, the more power it wastes.

For example, at 80v, 10v on the field, max RPM of 4700:



13a (1040 watts) just to spin the thing!

However with 4v on the field, it's a lot more reasonable:



Of course, it wasn't 100% stable at 4v...it was fine at 5v though. One of two things comes to mind there...this well used alternator's brushes didn't want to conduct throughout the rotation at a lower voltage, or else the magnetic force from the 3 phase wingdings(the 80v side) was enough to interfere with the weaker field. I'm leaning towards the latter, since it would crap out on hard acceleration(more 3-phase current, more likely to interfere). The other option is that my adjustable power supply was the culprit and couldn't supply more current when strained suddenly.

Low RPMs barely take any current to drive the motor. As in 1-2 amps at 1200 rpm.

Efficiency? Well, if the wasted current stays the same under load (no, it won't, but I don't have much else to go on), it would be ~78% efficient at full output (max rpm and controller current). That's better than I was expecting. Without rigging up a dyno, I won't be able to be sure though.

Torque? Yessir. Not very scientific, but I tried to hold the pulley from moving by hand with a piece of rubber for friction/protection, and from a dead standstill it easily overpowered said friction, however hard I tried to stop it. And with very little current; the most I saw it draw was 4 amps at a few hundred RPM when I held on as hard as I could. Plenty of starting torque...

Anyway, I'm pretty happy with the results. Yes, I'd like it if it were more efficient at higher RPMs, and/or if it would spin considerably faster, but that's about it.

Why bother? Well...motor cost is the first reason. A 1500w brushless motor would cost north of $200, and still need the controller to work, whereas you can pick up an alternator for next to nothing. The second reason is RPM...I wanted a motor that, even if I couldn't drive it particularly fast, would survive spinning at 10,000 rpm or more. Third is size...this little guy is just under 6" across and 6" wide...a much more reasonable size to work with than the Etek motor I have. I'm willing to trade off some efficiency for those reasons.

What could it be used for? Well...for a cheap ebike with plenty of power, or for a very mild hybrid of either a car or a motorcycle comes to mind. It's small enough and light enough for those purposes, and can handle the rpms if connected to an engine. You could run whatever else you want off it instead...1500w(2 horsepower) at ~3600 rpm will run a lot of things...a lawnmower, for example...

Bla, bla, bla...if you're interested in trying to do the same, just make sure the controller you get can run "sensorless".

Peace.

[Grant-53]

There are many technical books on AC motors and controls for those interested.
Remember torque is a function of current supplied. A capacitor can supply current for a faster start.

[Stubby79]

Pulled apart another one to try it out. This one is wired in wye rather than delta, and I wanted to see what kind of difference it made to the numbers...



A lot more work to get connections on this one; the regulator et al is internal on it, whereas the first one is just sitting under a cover on the back. Had to run wires to everything to get external connections.



Final setup, about the same as the other one.



Results?

Less current and less RPM at full 80v/10v - 3750 rpm rather than 4700, pulling 10a max vs 13. Got up to almost as fast with field weakening, 80v/4v - 4900 rpm vs 5200, dropping the current down to 7a. Similarly unstable below 5v on the field.

Now, these aren't identical alternators, so they're not going to play the same, but it wasn't so different as I thought. The speed difference could be (partially) due to the field winding pulling 30-50% more current than the previous one.

It makes sense that a wye wound would run slower than a delta...as an alternator, a wye puts out more voltage vs rpm than a delta...so it's going to take more voltage going in to spin it at the same speed.

Wye is supposed to have more low end torque...I'd have to agree with that...the first alternator rocked a bit on a gentle start up, this one would try to jump, however gently you gave it throttle from a dead stop. Good to be able to choose to either have high starting torque or high full speed torque, depending on how it's wired up.

Bit surprised not to see either rev up stupidly fast...I hate to guess what kind of voltage can come out of an alternator with no load on it and the regulator bypassed...sounds like my next project.

I'm rather enjoying this silly project.

[Stubby79]

Third experiment...



Wanted to know if any of the "self-learning" controllers are capable of running without the hall sensors...I went out of the way to find a 1500w 72v controlelr that specifically stated it could run "sensorless"; it was hard to find and at least 25% more expensive than the cheapest one that didn't state it either way.

Well, they do! No need to pay extra for one that states it...if it has the two little white wires for "Self learning", it should be good to go.

Ok, rather a blanket statement, maybe I just lucked out...I have at least one more controller that has self-learning that I can test it again with...but I'm pretty confident. Those little "ESC" RC controlelrs have been sensorless for years, no reason these shouldn't be too by now.

I think I only paid $10 or $15 for this little 350w controller, and less than $25 for the 500w and 800w ones. Canadian. So about 50 cents US.

There ya go. Happy hunting, DIY'ers!

Oh, and the motor ran modestly warm at 40v, compared to really warm at 52 and hot at 80v...so I'd suggest sticking with a 48v/1000w controller or less, unless you get your hands on a high-amp (130-200a) alternator. (and even then, I don't promise it won't get hot at "72v")

[Stubby79]

Someone was nice enough to document all the important stuff(power draw/efficiency) in some of his videos of his alternator-powered mid-drive electric bike build:

http://www.youtube.com/watch?v=K8MdOs0HtgM

If you look closely, it's pulling over 2000 watts just spinning the tire at full speed with no real load on it!

Another video he shows the numbers comparing the efficiency of it vs a hub motor; the alternator bike uses about 50% more watt-hours than the hub motor for the same speed.

Yeah, sounds about right.

Then again, maybe not...looking at his chart of amp draw vs speed with no load, he's pulling about twice the current I was with the same test; his alternator must be even less efficient than the two of mine.

Even so...if range and maximizing your pack size/weight are important, this probably isn't the way to go. There are better options, though not nearly as cheap, or likely in as small of a package.

It would do well when you want a lot of power in short burst, or if you kept the power demands reasonable most of the time by keeping the RPM down. You do get some pretty impressive torque, so you could get away with keeping the speed (voltage) down, run it at low RPM and not waste as much power just making the motor idle. You could also run it through the gears and optimize the output you need by down shifting for hills and up-shifting for speed.

Not a complete disappointment. Glad someone documented some of what I was looking to find out, before I put in all the same work. It still applies for what I intended it for, though cutting my battery range down by 1/3rd might limit it a bit much. Decisions, decisions...