Regenerative braking for braking assist

Question

I want to add regenerative braking to my bike, not for energy recovery, but as a braking supplement/assist on long downhills. Is anyone doing this and how effective is it?

A 3 phase hub motor IS a generator when freewheeling. All that's needed to create braking resistance is a load that can be switched in on demand. In fact, a short circuit might be effective if heat buildup in the windings can be managed. As an electrical engineer, I understand this much.

My question is, has anyone done it and how effective is it as an assist, not the primary brake?

Joe

Answer 1

Dynamic brakes are used on electric trains (sometime true regenerative brakes are not used because dynamic brakes are far simpler). I think they may also be used on trucks to regulate speed on long downhill roads.

You’d need a generator built into a wheel, like a dynamo hub, but somewhat larger to provide greater braking force. Power is dissipated in large resistors with cooling fins much like the cooler on an PC CPU. Depending on how sophisticated you want to be you could include a controller circuit to vary the braking force.

The issue is that all this adds significant mass to the bicycle, and the generator will create a non-zero resistance even when off (unless you provide a clutch to disengage it). Given that by definition you need to ride this bike up a hill in order to descend it, is the trade off worth it?

Answer 2

If you've already got a hub motor, and you've got electronics skills, a power resistor (rated to perhaps 100W) per phase on a heatsink (perhaps under the downtube) should do. You want the heat dissipated in moving air, rather in the windings, which will determine the resistance you can use. The difficult bit is likely to be getting motor specs in enough detail to avoid too much trial and error.

To switch it in, at least initially, I'd use relays (or one relay if you can find one with enough contacts. That would make it easy to disconnect the existing drive circuit at the same time, using changeover contacts. This would act as a drag brake with a fixed relationship to speed, so I'd start it at the top before building up much speed, otherwise the turn-on would be rather sudden.

I'd probably also add a little digital thermometer to the resistors - you can get cheap battery ones designed for panel mounting, with the sensor already fitted. This could be removed after testing, but I'd want to know if I was likely to melt the solder and lose contact, easily done with big load resistors unless they've got excellent heatsinking.

With well-chosen resistors, it should be possible to make this an effective drag brake to keep your speed down, but almost useless for stopping you.

There are also active load braking circuits, that dump the heat into MOSFET motor drivers (again with big heatsinks). I haven't used them myself in this application. They'd need more of a control circuit, but could be used to slow you down gently once you were going fast. However packaged versions may not be designed to provide long-term braking. Looking at the datasheets on these may give you a hint as to resistances to use.