If you have ever seen the inside of a direct drive (DD) hubmotor, the thought might have occurred to you that there is a surprising amount of unused space in the middle of it, and that it is also just about the right size to mount a controller there.
Well, you are not the first person to realize that. So, lets take a look at who has done this in the past, and who is doing this now.
_______________________________________________
BionX, the “integrated controller” Original Gangster
Two years ago, we wrote about how Canada-based BionX had fallen onto hard times (to see that article, click here). I am a fan of history, and also the evolution of electrical and mechanical things (to see our article on the history of ebikes, click here).
History is filled with examples of clever inventions that never “made it” due to the business side of things being managed poorly. Sometimes the product was “ahead of it’s time”, or it was simply not marketed properly. Here are some of the details about the BionX company and hub design.
Ebike controllers are most often designed to work with only one voltage. If you want to “hot rod” an ebike, garage builders often raise the volts and the amps. However, raising the amps is one method that is noteworthy because it also raises the amount of HEAT that the controller and motor suffer from.
If you want to avoid too much motor-heat, that leaves the volts to raise, since controllers and motors can be sensitive to heat-damage. Back when high-volt battery packs were simply unavailable, builders had to make their own high-voltage packs from cordless tool cells (to see our article on building a battery pack at home, click here). But even if you had the tools and skills to build a custom high-volt pack, the BionX controller has a very custom size and shape. So, it was quite an involved process to try and use an external controller to boost a BionX hub.
There are many other configuration clues that suggest that BionX was focused on making a design that was street-legal, and also very difficult to “hack” (just my opinion). Ebike laws were in their infancy in 1998, and it was feared that hot-rodders could injure pedestrians, leading to the image of ebikes being tarnished, rather than embraced.
By having the controller inside the hubmotor, Some of the wiring clutter was eliminated, and it was MUCH harder for an owner to “mess around” with the guts. Regardless of the details, I had to mention BionX first because they are the earliest reference I could find to a controller being located inside the hubmotor.
_______________________________________________
All-in-One hubmotors?
We wrote an article in January of 2019 about the so-called “All-in-One” hubmotors. As stated in the article, this is the most uncluttered type of drive system (with the motor, controller, and battery all in one housing), but…I don’t like them because the battery pack is forced to be small, and that also means that the battery pack is typically limited on how many volts and amps it can provide, not to mention it is also restricted to short range. (To read that article on All-in-Ones, click here)
That being said, I want to be clear that I actually like the idea of mounting the controller inside the hub…for “SOME” designs. A horribly designed ebike does not mean that all ebike’s are junk. The same goes for integrated controllers as an engineering principle.
_______________________________________________
Tidalforce / Wavecrest motor
The 2004 Wavecrest motor (used by the Tidalforce ebike) is no longer made, but it is noteworthy for several reasons. It had a proprietary 7-phase sine-wave controller that was “slightly” more efficient than a common 3-phase controller. Of course, I am mentioning it because it’s controller is integrated into he hub.
In the pic, you can see how the shank of the axle has a much bigger diameter than a common ebike axle, so the motor phase wires do not have to pass through a small-diameter hollow axle (which would limit them to being thin and weak). It also had optical sensors, which are impervious to heat, unlike the common and cheap Hall-sensors (to see our article on Hall sensors, click here).
The last thing that they were noteworthy for was that they built the battery pack into the front hubmotor. It “looked like” a 2WD ebike, but the rear hub was the motor/controller, and the front hub was the battery pack.
If you want to read our article on the Tidalforce electric bike, click here.
_______________________________________________
The Ferro Fluid question
Back in April of 2017, we wrote about successful experiments that involved adding a small amount of ferro-fluid to a direct drive hubmotor. However, it doesn’t help mid-drives or geared hubs, due to construction details that you can read-up about (to see our article of ferro fluid, click here).
You don’t necessarily need to take a hubbie apart to add ferro fluid, but the pic above was provided to show readers what it would look like inside. The FF is the dark goo in-between the rotor magnets. FF is definitely a high-performance trick, and if you like large hubmotors because of their ability to absorb huge amps, then FF can really boost your performance for just a few dollars (to read our article on big hubmotors, click here).
There is normally a slim air-gap between the stator-coils (the electro-magnets that cycle on and off), and the magnets in the rotor. The side-plates of the hubmotor are typically aluminum, which ‘could’ absorb and shed heat well. However, the air-gap is a significant barrier to transferring the heat from the stator to the outside air.
Just a tiny amount of FF will fill the air-gap between the stator and rotor. It is a liquid, so it adds very little drag. Since it is magnetic, you don’t need to worry about it leaking out or settling into a pool. It sticks to the rotor magnets, and creates a fluid bridge that the heat can travel through. This means that your temporary peak amps (used for a few seconds when accelerating) can be as much as doubled in some cases without any harm to the motor.
Since FF provides a surprising increase in DD hubmotor performance, certain models of hubmotor that had previously been “a little boring” can now be considered a viable option for decent performance. The hubs listed below must all be considered in reference to their stock abilities, and then also their potential if FF is added.
_______________________________________________
Stromer ST1
I wouldn’t call Stromer a hot rod, but…for reasonably flat land (like a beach community), using a DD hubmotor that is restricted to 500W isn’t horrible. Mid drives often need to be shifted often to take full advantage of the gears (which allows it to climb well). However, some people like the simplicity of an E-drive with no shifting (To see our 2013 review of the Stromer ST1, click here).
My initial reaction to seeing the construction details of the Stromer motor is that…adding FF would help it a lot…IF…you could hack the controller to add more amps (which you can’t). If you find a used Stromer at a very good price, be aware you can toss the stock “hockey puck” controller and add an external controller that unlocks the potential of this well-made product.
The Stromer is a Swiss-designed and German-made “Mercedes” type of ebike. The torque-sensor and control programming make this a low-powered exotic, like an expensive 4-cylinder car. It rides absolutely fantastic, but…there is no serious power available.
Like the Voilamart hub below, the Stromer controller is fully “potted” for waterproofing. To see our article on waterproofing your ebike, click here.
_______________________________________________
Golden Motor “Magic Pie”
If you absolutely must have a DD hubmotor right now, and you need an integrated controller, then a Golden Motor kit is the “go to” system. They have been doing this for a while, so they are not new at it, and this is a very real product.
These hubmotors are advertised as having a programmable integrated controller that will function with voltages at 24V, 36V, and 48V. If you want to convert a BMX that has a 20-inch wheel for a child, this is probably the easiest way to do it. Their website can be found by clicking here.
I don’t know of anyone who has added FF to a Magic Pie, but that would be awesome, and I will keep an eye out for any data on that.
_______________________________________________
Voilamart 1500W
I don’t know anything about this company, and as of March 2020, they are new to me. However, this design is very intriguing. If you use a cast aluminum spider in the center of a DD hub (instead of the common stamped-steel disc) the hub will be heavier and more expensive, but…the aluminum will act as a thermal-sponge to absorb the heat from any temporary amp-spikes you experience during acceleration. Then, that captured heat can be dissipated over time when you are just cruising along.
I mentioned the cast-aluminum stator-core as a key feature found on the so-called “1500W” hubmotors (which have 35mm wide stators), and these are one of my favorite recommendations, since they have typically been able to run reliably on 2600W (52V X 50A). If you can add ferro fluid, you can even use many more amps for at least a few seconds, which is often all it takes to get to it’s top-speed. To read our article on a popular 1500W DD hubmotor, click here.
This motor is advertised as a 48V system, and that means that it likely has the common 63V max capacitors. It “might be possible” to use a 14S 52V battery pack, but it has not been verified yet. If you use the recommended 48V, then the 1500W power rating results in a max amps of 31A. Not bad, but it will certainly not feel like 50A (which is quite nice, if you can get that).
I suspect that this motor would work quite well with ferro fluid, so the stator coils at an occasional 2600W (52V X 50A) would not be in any danger of damage from overheating. However, the stock controller does not seem to have a decent heat-shedding pathway. The wide and flat spokes of the stator spider (that the controller sits on) would absorb some heat, so that is good news.
If I had one, I would add a temp sensor to the controller, and then see how much she could take. If you added FF, then I suspect the sideplates could absorb quite a bit of heat from the stator-spider, and as a result, the controller would significantly benefit from adding FF. I would also add some thermal compound between the controller base and the stator-spider contact-points, to ensure that any controller-heat that was building would have a decent escape pathway.
The 34-MPH (55km/h) top-speed on flat land in a 26-inch diameter tire (450-RPM’s) suggests that the default model has a kV of 12.8/RPM’s per volt. (To see our article on kV and motor design details, click here). The listing implies it has a 7-speed cassette instead of a 7-speed freewheel, but don’t count on it.
If you are comfortable with risky behavior, and are curious about this new model, click here for the link as of March 2020.
Ok, that is three DD hubmotors with integrated controllers that actually exist, and two from the history of ebikes. This does not include the All-in-Ones, but…this is an article that needed to be written, and I’m just the guy to do it….
_______________________________________________
Written by Ron/spinningmagnets, March 2020
