3 ms·

1) But you can't just handwave away some inconsequential components in terms of cost (brakes and a chain drive), describe a bunch of extra complexities by not h

1) But you can't just handwave away some inconsequential components in terms of cost (brakes and a chain drive), describe a bunch of extra complexities by not having those components and then say it will be cheaper. Everything described in the article looks like an extra complexity and cost to me, compared to the way most e-bikes are built now. A few red flags stand out to me: A) Brakes. Mechanical bicycle brakes can provide kW of power dissipation for short time periods. e-bike motors are rated for ~250W and even if one were rated for 1000W, they are not going to be able to dump that power into a battery. So if you want regenerative braking, it's either going to need mechanical or resistive braking as well. There's a reason virtually every e-bike design from low-end to high-end does not even have regenerative braking and only has mechanical brakes. Because physics dictates you need mechanical brakes, and regenerative braking barely helps you on a bike anyway. If you're going to go against this trade of of every e-bike that came before, you had better show some numbers or reasoning. This is a physics problem so it's really not something where you can say "in the future when technology is better" especially in the context of claiming that cost will be lower. I'm not aware of any vehicles (bikes or other) that use only dynamic braking. Even streetcars, which historically use resistive braking (with very large components to do it by the way), always have emergency mechanical brakes. B) Mass. If you're going to claim removing the chain and brakes saves significant mass, you better show some numbers. A hub motor is like 2-3kg and I guess a matching generator should be a similar mass. C) FreeDrive system linked by the author. The claim is 5% less efficient than a chain, but this is really hard to believe given that power must be converted from mechanical to electrical and back again. The image of FreeDrive looks like a premium product, not cheap and is controlled by a CAN interface and a human-machine-interface, according to its press release. This is not a component that's going to make your bike cheaper. And if it wasn't heavy, the mass would be in the press release. It's heavy, I'd guess also about 3kg from images. OK, you can think of something lighter but nothing will be able to approach a chain drive. After all, you have to support the same input and output loads in addition to extra components. D) Software handwaves. Why would a bluetooth connection to your phone be preferable or cheaper than a wired display? Sure, you've eliminated 2 or 3 wires, but now you need an adjustable mount and ongoing software maintenance for Android and iOS? What does "<motors that> will hand more control off to software, allowing physical simplification" mean? BLDC motors are dead simple mechanically but basically cannot function without software. Fortunately such software is heavily optimized and available for practically nothing inside commodity motor controllers. Now, all of the above is fine as speculation anyway, except for the fact that the premise of the article is making a cheap Corolla-like e-bike. It gets so close by saying "maybe one will come out of Asia." I would counter this whole article by saying "why aren't the base models of Panasonic and Bridgestone e-bikes sold worldwide?" They start at just over $1000, but are definitely the Corollas of the e-bike world. If they sold 10x as many they probably could be $500. I'm not sure why North America seems to have cheap junky bikes and high-end bikes in stores, with barely any decent commuter bikes (e or non-e) on the market.

rjsw 5y ago
CAN isn't expensive to implement and is built in to any SoC that you are likely to use for a display.
mft_ 5y ago
You make some great points. FWIW, they don't really relate to the original post I was disagreeing with, but great points nonetheless :) A few thoughts back: A) There are plenty of (heavy) bikes out there with a front caliper and a rear brake activated by reverse pedalling. Suspect a similar compromise would work, and a decent front caliper/brake level pair isn't heavy. B) In theory, simplifying the drive-train doesn't just subtract the mass of the chain - there's also a freewheel mechanism, rear cassette, chain wheels (usually 2-3), front and rear derailleur, multiple cables, handlebar shifters... and of course, the cost/complexity of purchase and fitting for these items. This would be quite a considerable weight, especially when we're talking about bottom-end components. C) Agree totally, but (for the sake of this discussion) the comparison isn't just the drive-train of a standard bike - it's a whole standard drive-train, plus a hub or bottom-bracket motor - as deployed in current e-bikes. Current Bosch BB motor assemblies seem to be ~6.3 - 7.1 lbs each. [1] D) Agree - ideally, there's no need for a display at all - a simple e-bike could just work. [1] https://www.bosch-ebike.com/us/products/drive-units https://www.bosch-ebike.com/us/products/drive-units
- johnwalkr 5y ago
  A) I agree regenerative braking plus a front brake could be a reasonable compromise. This is actually a pretty good idea. Could be simpler than trying to have some algorithm decide when to use dynamic vs mechanical braking. Front and rear brakes feel different already so it might be pretty intuitive to operate. B) Potentially those other things are subtracted, but a lot of e-bikes get rid of a lot of those components too and keep either a single speed with motor assist, or use a 3 speed internal gear hub. Really, if you're making the argument for a cheap, decent bike, you have to compare against the single speed option. The article didn't go into this detail but as long as the battery has juice, single speed e-bikes with motor assist don't have a big disadvantage compared to multiple speeds. C) And in this article a bottom bracket generator is required (about 2-3kg, more likely 3 for a "Corolla") plus a hub motor (also 2-3kg and more likely 3 for a "Corolla"). So it's "drivetrain plus motor" vs "generator plus motor" with some room for discussion about how many drivetrain components are involved. Until I read your reply I didn't even think about the fact that this proposed e-bike needs a hub motor instead of a BB motor, so it loses some more efficiency there automatically. D) I think front/rear lights are probably required for many jurisdictions and ought to be permanently fixed to the bike in any decent design to take advantage of the battery and be harder to steal, and battery level indication is of course required. A speedometer and programmable speed limiter would be good things to include if you want to make an e-bike that can be sold in as many places as possible without changes.

rjsw 5y ago

CAN isn't expensive to implement and is built in to any SoC that you are likely to use for a display.

mft_ 5y ago

You make some great points. FWIW, they don't really relate to the original post I was disagreeing with, but great points nonetheless :) A few thoughts back: A) There are plenty of (heavy) bikes out there with a front caliper and a rear brake activated by reverse pedalling. Suspect a similar compromise would work, and a decent front caliper/brake level pair isn't heavy. B) In theory, simplifying the drive-train doesn't just subtract the mass of the chain - there's also a freewheel mechanism, rear cassette, chain wheels (usually 2-3), front and rear derailleur, multiple cables, handlebar shifters... and of course, the cost/complexity of purchase and fitting for these items. This would be quite a considerable weight, especially when we're talking about bottom-end components. C) Agree totally, but (for the sake of this discussion) the comparison isn't just the drive-train of a standard bike - it's a whole standard drive-train, plus a hub or bottom-bracket motor - as deployed in current e-bikes. Current Bosch BB motor assemblies seem to be ~6.3 - 7.1 lbs each. [1] D) Agree - ideally, there's no need for a display at all - a simple e-bike could just work. [1] https://www.bosch-ebike.com/us/products/drive-units https://www.bosch-ebike.com/us/products/drive-units

johnwalkr 5y ago
A) I agree regenerative braking plus a front brake could be a reasonable compromise. This is actually a pretty good idea. Could be simpler than trying to have some algorithm decide when to use dynamic vs mechanical braking. Front and rear brakes feel different already so it might be pretty intuitive to operate. B) Potentially those other things are subtracted, but a lot of e-bikes get rid of a lot of those components too and keep either a single speed with motor assist, or use a 3 speed internal gear hub. Really, if you're making the argument for a cheap, decent bike, you have to compare against the single speed option. The article didn't go into this detail but as long as the battery has juice, single speed e-bikes with motor assist don't have a big disadvantage compared to multiple speeds. C) And in this article a bottom bracket generator is required (about 2-3kg, more likely 3 for a "Corolla") plus a hub motor (also 2-3kg and more likely 3 for a "Corolla"). So it's "drivetrain plus motor" vs "generator plus motor" with some room for discussion about how many drivetrain components are involved. Until I read your reply I didn't even think about the fact that this proposed e-bike needs a hub motor instead of a BB motor, so it loses some more efficiency there automatically. D) I think front/rear lights are probably required for many jurisdictions and ought to be permanently fixed to the bike in any decent design to take advantage of the battery and be harder to steal, and battery level indication is of course required. A speedometer and programmable speed limiter would be good things to include if you want to make an e-bike that can be sold in as many places as possible without changes.