Evening all,

 

I'm from London and am working on an ebike project I would like to share with this great place..

 

It's a custom lightweight framed city commuter aimed at UK/European cities.

 

I believe I have settled on using a 250W geared hub motor for rear drive. The wheel size I am working with is 700C with a 28mm tyre. I am intending to use single speed belt drive and disc brakes front and rear.

 

The aim of this concept is of city commuter use, fairly flat-tarmac terrain and a range of around 25 miles or so. (the more the better)

 

I would like to have a PAS system and no hand throttle whereby the more you pedal you more power ids delivered.

 

I am working on a custom aluminium frame to be a light as possible while being strong enough. The idea is to make the bicycle as subtle as possible and the battery pack to be integrated in to the frame and removable for recharging also.

 

The controller is to be integrated in to he frame.

 

I will hopefully be ordering parts over the next week or so and will document everything on here.. the frame design is almost settled pending stress analysis results.

 

If anyone has any input, advice or would like to ask anything please be my guest,

 

Best,

 

Jamie

Drive systems that give power or torque proportional to how hard you pedal sound sort of logical. but in practice, they're not very pleasant to use. There are also systems that give power proportional to cadence. They're even worse. How nice the bike is to ride depends a lot on the power algorithms you choose. You need to try a lot of ebikes, and then try and figure out what exactly is happening. The best one I've ever tried was the Woosh Sirocco with Torque sensor.

 

Are you going to build your own controller or use a proprietary one?

Hi d8veh,

 

I tried a Mercedes Smart eBike a while ago which i believe uses a torque sensor also. It was very intuitive to ride and took less than a minute to get used to. I really liked there was not a hand throttle and you kind of became mechanically part of the bike to control the power output.

 

At the moment with regard to the controller I'm not totally sure..initially i wanted to build my own however in the interest of time I may have to buy one.

 

Thanks for that advice about the drive systems you mentioned.

The algorithms in the controller on that Mercedes will be a lot more complicated than just giving proportional torque.

 

If you buy a ready-made controller, the chances are that it'll use speed-control rather than torque control. Even if it uses torque control, it'll probably have fixed levels of torque that are activated when you pedal. You'll need to find a controller supplier that makes controllers specifically for European systems.

Yes it was super smooth and must have taken quite some development to get it like that.

 

I would really like to get mine as smooth as possible so the speed control system just won't cut it.

 

I am thinking of using Arduino and a torque sensor bb to give a throttle signal..

The problem is that most controllers use the throttle for speed control, so your torque sensor/arduino can only send a speed request to the controller. The throttle signal is a voltage between about 1v and 4v, which is divided into a fixed number of steps (probably 16). Each step corresponds to a speed. Then there's an algorithm in the controller's CPU that determines the amount of power to give , which is based on the difference between the motor's actual speed and the target speed, so, as the bike slows down (say on a hill), the controller will increase the power in an attempt to maintain the target speed, which is very convenient and makes bikes with speed control controllers very comfortable to ride - much better than torque control. The PAS levels correspond with certain values too, as if the throttle were applied at a fixed level.

 

To do what you want requites a current control controller, which are very rare in the ebike components market. You need to do some research on "current control". Do a search on Endless-sphere, where a few guys have tried to do something.

the way the XOFO rear hub motor with in motor torque sensor works is quite simple: the cassette is spring loaded, 12 magnets arranged on a ring activate 4 Hall sensors which translate the angular deplacement (which is proportional to pedalling force) directly to a throttle voltage, wired into the controller throttle input. See my other post http://www.pedelecs.co.uk/forum/threads/uk-ebike-parts-suppliers.16885/#post-207905

the way the XOFO rear hub motor with in motor torque sensor works is quite simple: the cassette is spring loaded, 12 magnets arranged on a ring activate 4 Hall sensors which translate the angular deplacement (which is proportional to pedalling force) directly to a throttle voltage, wired into the controller throttle input. See my other post http://www.pedelecs.co.uk/forum/threads/uk-ebike-parts-suppliers.16885/#post-207905

 

It's absolute brilliant - the best I've ever tried - until you get to a steep hill. Then the spring effect in the motor shows itself as you rattle against the stop with each pedal stroke. It feels like you've got loose cranks. If you owned the bike, I guess that you'd soon get used to it, so not a deal-breaker, but just a bit disconcerting.

if you get to a hill and push hard, I agree that the motor simply maxes out.

You can still adjust the throttle voltage with the LCD +/- or change gears.

I think that you've misunderstood me. The problem is that the sprocket is not rigidly fixed to the motor. When you pedal hard, you can feel free-play because of the spring movement. If you loosen off the bolts on a square taper crank-set. It'll feel the same, except that you won't get far.

Thanks for the advice guys..certainly given me a lot to research.

So far..

That's a lot of custom welding & fabrication right there. To get the rigidity from an alloy U section it'll be thicker than a bull's whatsit & heavy, uncomfortable to ride. Don't forget it rains here too.

 

Obvious question is obvious; how's that weight saving over say a carbon, Easton/Columbus steel, or an alloy frame with a battery slung in the triangle in a bag? Or even a battery slotted behind the seat post..

 

Not sure what market research you've done before spending your (precious) time designing the frame but have you checked out the style of bikes ridden in must European cities? Step through with racks & baskets galore. They don't try as hard to look "hardcore" as the UK market.

Edited February 26, 201412 yr by Tangsoo

Hi Tangsoo,

 

Yes a fair bit of fabrication involved at this stage, this is due to it being a prototype.

The top tube and down tube sections also have a box section running through inside that is out of view here, adding integrity.

 

That design is from 2mm aluminium sheet, and after some initial finite element analysis shows to be surprisingly strong and light.

 

The frame shape is designed to be something between a step through and male frame, the image is a bit deceiving as the top of the seat post is only represented with a centre line. It it supposed to represent a type of uni sex frame geometry for a 95th percentile European adult.

 

The target market is UK city commuters. After many surveys, interviews, questionnaires etc I concluded UK users do not so much like the 'European' step through/racks/baskets galore style electric bicycles and prefer subtlety. Hence the design is to have some kind of hidden integrated battery that is also removable for practical recharging.

A frame shape that is not too 'hardcore' so to speak but looks close to a regular bicycle.

 

The design is changing all the time with finite element analysis checking and material testing in the lab.

 

I have some basic renders I will upload shortly..

Edited March 2, 201412 yr by Jamie

I've got a carbon MTB commuter with 250w front hub and pedal sensor and twist throttle on the right bar

 

 

http://endless-sphere.com/forums/viewtopic.php?f=4&t=56711

 

I love the twist throttle because I can leave it in top or second gear and still get away from the lights with instant max power whereas the PAS takes a half second and a few crank revs to register and provide thrust.