Friday, November 16, 2012

SpnKix Technical Teardown

-->

I backed this company on kick-starter by 'buying' a pair of electric roller skates called Spnkix.
The design looked cool and fun, but I am disappointed in the result. I thought I might as well take a few minutes to publish this as a review here so that anybody else who is considering buying them knows what they are getting.

In this post, I'm also posting some results of me engaging in my other hobby: taking things apart. I actually enjoyed this part, at least. Since this is the first time I have been the first to take apart a new category of device, I'll be through in posting photos for others who enjoy this stuff. My e-vehicle engineering experience is relevant enough that I hope you find my comments interesting too.

Teardown and Comments:

One skate shoe weighs 4.35 kg / 9.6 lb - it feels every bit of it.

Your shoes are secured in with two large buckles, each is in the style of a snowboard binding.

Steel support between the wheels underneath the foot.

The handheld remote control has a good range. It worked 30 feet away, which implies it should be reliable when used from just a few feet away as intended.

The wireless modules are prebuilt Nordic Semi RF modules. These are a digital radio and so should be better than the old electric skateboard remote controls that were prone to interference. It seems a bit wasteful to pick a two-way protocol for one way communications, though.



The nylon frame feel very solid and has a nicely done honeycomb reenforcement pattern in the thicker areas.





It seems a bit odd that there is motor wiring into each wheel, but only a motor on one side. The only plausible explanation is that there was so little space in the main body, that the only place to put the motor driver electronics was in the other wheel. This is an unusual approach, as a 75W motor driver can be almost small enough to sit in the space below the battery, but putting it in the other wheel requires significantly more parts and wiring in a wheel which otherwise could have been simple.



It has a substantial battery pack. At 125 Wh per pack, this is bigger than any laptop batteries and probably not legal to take onto an airplane.
Does not have enough connections to have a full charge balancer. The 7S2P LiO cell configuration only has 4 charge balancer channels. This will somewhat shorten battery life because the cells have different capacities and over time the stronger cell in each pair will start to get more and more overcharged and the weakest more and more undercharged. But 4 is better than none.




The motor has three thicker phase wires, and a red and a black wire in one connector and three wires in another connector. This means it is a brushless motor with a hall sensor, which is expected, since this is really the only way to get this much torque and power into a hub without using a gearbox. Kudos on not going cheap on the motor technology.

Hub motor - not disassembled
I didn't disassemble the hub motor. This is kind of tricky and dangerous to do and my curiosity was not that strong. The strong magnets in hub motors will snap the pieces around viscously. I am fairly confident it would have been like a smaller version of the normal bike hub motors. (Search youtube for “hub motor disassembly” if you're curious).
Motor connectors

I did notice that the manual suggests you replace the tires by taking off the hubs of the wheels. Lies!This is effectively impossible to do. Even after struggling to undo the screw that hold on the hub plates (the Phillips socket strips very easily), it is still impossible to simply pull off the side plate like the manual says. I can't even prise it off when the motor is in a vise. I assume this is either because it is too tight, or because the magnets of the hub motor are holding it on. You often need a special press to take hub motors apart. If the tires wear out or get damaged, you're in for a lot of trouble.

There isn't any sealing or waterproofing.

A quick high school physics calculation on the specified maximum weight, slope and speed gives a required power of 161W required per foot, which is considerably more than the 85W motor rating.
( 0.5*(81.6 kg+2*4.35 kg)*9.81*sin(5 deg)*4.17m/s )
So this is misleading, but on the other hand, it would require a fantastic amount of skill to remain balanced at top speed and top power, so this is not likely to be an issue in reality.

Given the commendable lack of second rate materials, and the un-commendable poor DFM (design for manufacturing), I seriously doubt that a first production run was profitable at $375 per pair, even ignoring the tooling costs. The tooling costs would have been substantial, even considering how slowly it was done.

Overall, this is a good effort for the first version of a product. But given the flawed concept, I doubt there will be a second.



SpnKix Review

Image from spnkix.com
I backed this company on kick-starter by 'buying' a pair of electric roller skates called Spnkix.

The design looked cool and fun, but I am very disappointed in the result. I thought I might as well take a few minutes to publish this as a review here so that anybody else who is considering buying them knows what they are getting.

Review of the Product

TL;DR: Seems like a good idea, but isn't.

The experience is not at all like roller skates, so it takes a bit of getting used to. The general feel is different but not necessarily too bad. The braking effect is cool – it applies a nice constant stopping force and is much better than any mechanical brakes.



It is kind of a shame that that only the wheel on one side is motorized. Because it kind of twists your body as it applies power, so you are always having to work against it.

Also, for something that is supposed to be easier than walking, it is certainly really tiring – maintaining the posture and balance is a lot of work. After the first 15 minutes of using it I was sweating profusely like I'd just had a workout.

Also, I don't know if they are all like this, but my pair always pulls a bit to the right. I think this is because each skate applies a slightly different amount of torque in response to the controller.

Swapping the left and right ones over from what they recommend feels much better. Having each skate on the foot that is indicated L/R means that the motors are on the inside and they kind of push outwards, this requires always pulling your legs together like you need to go to the bathroom. I didn't like this.
Switching them over onto other feet, so that the motors were on the outside felt much better. Now they require pushing outwards to counteract them steering inwards, which feels better for me.

For people with smaller feet they have a lot of difficulty standing still on their toes. The strap seems oversized for regular shoes. The width is non-adjustable causing my feet to shift left and right at times (the large straps made this worse). Also note that I'm 5'10", and shoe size 10, so poor fitting must be a very common problem.

Overall, the effort it takes you to compensate for the strange torque they
cause on your feet and body seems to not be worth the fact that motor is
powering you forward. I can't imagine any situation where I would prefer to use these instead of regular roller skates.