Setting a speed limit on bafang mid drive

[matthewslack]

It's not compliant if it overheats in the test. The paradigm is that low power is OK and high power isn't, but that's not what the law says. The law says it mustn't overheat and the motor can't be used if it's rated at more than 250w. The rule is perfectly functional and understandable. What makes you think that you know that something different to what was written was intended?

Leaving aside the unnecessary personal stuff, common sense.

What other intent can there be from the inclusion of a number in a regulation than it has some functional meaning? And given the low power of EAPCs is to allow them to be treated as bicycles and safely coexist with more vulnerable users of the same space, it seems reasonable to infer that the intent was to recognise that a power level of 250W gives performance and therefore risk levels on a par with unpowered cyclists.

The defect in the regulation as you know is that the power test only establishes minimum continuous power capability, not maximum, and the test protocol does not require the maximum to be determined.

The rating as currently interpreted means nothing more than a number inscribed on the motor. That is not what the public understand and expect it to mean.

A motor that overheats at 250W but not at 200W can be rated at 200W, which is compliant with EN15194 because at least until now, 200 is less than 250.

A rule which allows any and all motors to be treated as compliant provided they didn't overheat in a 250W test case and have 250W tattooed on the back of their neck is the very definition of disfunction.

 

[saneagle]

Leaving aside the unnecessary personal stuff, common sense.

What other intent can there be from the inclusion of a number in a regulation than it has some functional meaning? And given the low power of EAPCs is to allow them to be treated as bicycles and safely coexist with more vulnerable users of the same space, it seems reasonable to infer that the intent was to recognise that a power level of 250W gives performance and therefore risk levels on a par with unpowered cyclists.

The defect in the regulation as you know is that the power test only establishes minimum continuous power capability, not maximum, and the test protocol does not require the maximum to be determined.

The rating as currently interpreted means nothing more than a number inscribed on the motor. That is not what the public understand and expect it to mean.

A motor that overheats at 250W but not at 200W can be rated at 200W, which is compliant with EN15194 because at least until now, 200 is less than 250.

A rule which allows any and all motors to be treated as compliant provided they didn't overheat in a 250W test case and have 250W tattooed on the back of their neck is the very definition of disfunction.

There's no defect in the test. The test is there to make sure that manufacturers don't over-rate their motors, which could cause fires. It's your understanding that's deficient, nothing to do with the public. The rating is there to protect the public from a wrongly described product.

Think about lifting gear. You wouldn't want a 10T hoist to snap when lifting 5T. As well as the tension in the line, the hoist has to deal with rough treatment that causes unspecified impact forces while under load, so it has a large safety factor built in and can probably lift 30T if you're careful. There's no misdirection or cheating. Is anybody going to complain, " You sold me a 10T hoist, but I found out that it can lift 20T with no issues. I want my money back and I'm going to report you to Trading Standards"?

 

[matthewslack]

There's no defect in the test. The test is there to make sure that manufacturers don't over-rate their motors, which could cause fires. It's your understanding that's deficient, nothing to do with the public. The rating is there to protect the public from a wrongly described product.

Think about lifting gear. You wouldn't want a 10T hoist to snap when lifting 5T. As well as the tension in the line, the hoist has to deal with rough treatment that causes unspecified impact forces while under load, so it has a large safety factor built in and can probably lift 30T if you're careful. There's no misdirection or cheating. Is anybody going to complain, " You sold me a 10T hoist, but I found out that it can lift 20T with no issues. I want my money back and I'm going to report you to Trading Standards"?

In it's original application, which is the general case of motors rather than ebike motors, that is so.

The defect in the test procedure when applied in EN15194 is when a motor that would pass the test at 500, 750, 1000W with flying colours and should therefore be rated as such is deliberately underrated as 250W.

It is not difficult to devise a test procedure for EN15194 based on the same motor test: it would simply demand two tests be performed, one at the claimed rated power and one at some mark-up of the rated power, say 150 or 200%. If the motor fails the second test then it is OK to be rated as 250W. But if it passes, it is clearly significantly more powerful than 250W and so not compliant.

The issues about variable operating conditions in the real world are also fairly easy to solve.

For basic bikes they have already been solved: the typical generic hub motor 7A continuous, 15A peak controller at 36V is ideal. In the motor's efficient speed range, 7A x 36V x 80% is near enough 250W, and 15A for short periods gets you started and copes with operating conditions down to about 50% efficiency, i.e. non-extreme hills.

There's a good argument for considering the controller and motor as a 'unit', which makes control of performance very easy, by placing requirements on the controller, and removes the need to worry about the motor.

Placing requirements on the control algorithms is an easy way to get sensible results. For example, having a maximum multiplication factor on torque sensor bikes, which makes the rider the limiting factor rather than the motor.

At the end of the day, the intent of EAPCs is to be as safe as bicycles in places where bicycles are allowed to go, not a backdoor to a cheap, powerful but slow electric motorbike.

 

[guerney]

@matthewslack - what have you got against fatsos, the old, the infirm and cargo bikes? They'd all be fskced with what you're suggesting.

 

[saneagle]

In it's original application, which is the general case of motors rather than ebike motors, that is so.

The defect in the test procedure when applied in EN15194 is when a motor that would pass the test at 500, 750, 1000W with flying colours and should therefore be rated as such is deliberately underrated as 250W.

It is not difficult to devise a test procedure for EN15194 based on the same motor test: it would simply demand two tests be performed, one at the claimed rated power and one at some mark-up of the rated power, say 150 or 200%. If the motor fails the second test then it is OK to be rated as 250W. But if it passes, it is clearly significantly more powerful than 250W and so not compliant.

The issues about variable operating conditions in the real world are also fairly easy to solve.

For basic bikes they have already been solved: the typical generic hub motor 7A continuous, 15A peak controller at 36V is ideal. In the motor's efficient speed range, 7A x 36V x 80% is near enough 250W, and 15A for short periods gets you started and copes with operating conditions down to about 50% efficiency, i.e. non-extreme hills.

There's a good argument for considering the controller and motor as a 'unit', which makes control of performance very easy, by placing requirements on the controller, and removes the need to worry about the motor.

Placing requirements on the control algorithms is an easy way to get sensible results. For example, having a maximum multiplication factor on torque sensor bikes, which makes the rider the limiting factor rather than the motor.

At the end of the day, the intent of EAPCs is to be as safe as bicycles in places where bicycles are allowed to go, not a backdoor to a cheap, powerful but slow electric motorbike.

Whoever made the rules decided that the 25km/hr power limit would be enough to make ebikes as safe as bicycles. The amount of power has nothing to do with it because it's speed which increases any impact damage. Any bicycle can do 25km/h on the flat and a lot more downhill. The damage done is proportional to mass x velocity squared. A 75kg guy coming at you down a steep hill on a road bike at 30 mph is going to do 3 times more damage than a fatso on an ebike with a 5kw motor travelling at 15 mph.

 

[sjpt]

The amount of power has nothing to do with it because it's speed which increases any impact damage.

I think the amount of power has a significant effect because it can change the way the rider behaves. They (the riders of the higher powered machines) are more likely to make sudden unexpected changes, and do them at speed, and weave through pedestrians in shared spaces.

They are also more likely to be riding at 15.5 mph on bridleways, and accelerate faster and take corners faster, so doing more damage to the surfaces.

Of course thoughtless riding isn't limited to those on more powerful machines; but the temptation is there. Same with drivers. Many drivers of high powered cars use that power to ensure they are driving safely, but a significant few use it very dangerously.

 

[saneagle]

I think the amount of power has a significant effect because it can change the way the rider behaves. They (the riders of the higher powered machines) are more likely to make sudden unexpected changes, and do them at speed, and weave through pedestrians in shared spaces.

They are also more likely to be riding at 15.5 mph on bridleways, and accelerate faster and take corners faster, so doing more damage to the surfaces.

Of course thoughtless riding isn't limited to those on more powerful machines; but the temptation is there. Same with drivers. Many drivers of high powered cars use that power to ensure they are driving safely, but a significant few use it very dangerously.

The difference in reaching 15 mph between a typical dinner plate 1kw motor and a 250 one is minimal. In fact most 250w ones can probably get there quicker, since their power characteristics are more tuned to speeds of 12 to 15 mph. If you've ever ridden a bike with one of those motors, you'd know what I mean.

 

[matthewslack]

Whoever made the rules decided that the 25km/hr power limit would be enough to make ebikes as safe as bicycles. The amount of power has nothing to do with it because it's speed which increases any impact damage. Any bicycle can do 25km/h on the flat and a lot more downhill. The damage done is proportional to mass x velocity squared. A 75kg guy coming at you down a steep hill on a road bike at 30 mph is going to do 3 times more damage than a fatso on an ebike with a 5kw motor travelling at 15 mph.

Going back to controller and algorithm, which ties in with your response, this is how I would approach it:

To accommodate a wide range of operating conditions, flexibility is allowed. But I would impose a hard limit of 250W output power at 25km/h. More output power is allowed but only if speed is reduced in proportion, so that 500W is fine but only with a 12.5km/h ssistance speed cap, and 1000W is OK up to 6.25km/h.

That way heavy riders can get up hills, normal riders can get up steep hills, EAPC performance is in line with pedestrians' expectations i.e. they are not blasting up steep hills at unbicyclelike speed and scaring the children.

Simple to implement by treating the controller as part of the motor, and giving the controller knowledge of the motor's performance curves in the same way that an IC engine is constrained by the programming embedded in its ECU.

This approach also allows the use of any motor as needed to achieve the necessary performance, because any excess capability is constrained by the controller algorithm. For example, bikes aimed at 150kg riders can use a chunkier motor, or a slower motor wind to manage the extra load in a way that respects the intent of the regulations.

 

[Peter.Bridge]

Going back to controller and algorithm, which ties in with your response, this is how I would approach it:

To accommodate a wide range of operating conditions, flexibility is allowed. But I would impose a hard limit of 250W output power at 25km/h. More output power is allowed but only if speed is reduced in proportion, so that 500W is fine but only with a 12.5km/h ssistance speed cap, and 1000W is OK up to 6.25km/h.

That way heavy riders can get up hills, normal riders can get up steep hills, EAPC performance is in line with pedestrians' expectations i.e. they are not blasting up steep hills at unbicyclelike speed and scaring the children.

Simple to implement by treating the controller as part of the motor, and giving the controller knowledge of the motor's performance curves in the same way that an IC engine is constrained by the programming embedded in its ECU.

This approach also allows the use of any motor as needed to achieve the necessary performance, because any excess capability is constrained by the controller algorithm. For example, bikes aimed at 150kg riders can use a chunkier motor, or a slower motor wind to manage the extra load in a way that respects the intent of the regulations.

That wouldn't work very well with cheap hub motored bikes where the motor efficiency reduces dramatically with speed, so at less than 7 or 8 mph on high power levels the controller would quickly overheat

 

[saneagle]

Going back to controller and algorithm, which ties in with your response, this is how I would approach it:

To accommodate a wide range of operating conditions, flexibility is allowed. But I would impose a hard limit of 250W output power at 25km/h. More output power is allowed but only if speed is reduced in proportion, so that 500W is fine but only with a 12.5km/h ssistance speed cap, and 1000W is OK up to 6.25km/h.

That way heavy riders can get up hills, normal riders can get up steep hills, EAPC performance is in line with pedestrians' expectations i.e. they are not blasting up steep hills at unbicyclelike speed and scaring the children.

Simple to implement by treating the controller as part of the motor, and giving the controller knowledge of the motor's performance curves in the same way that an IC engine is constrained by the programming embedded in its ECU.

This approach also allows the use of any motor as needed to achieve the necessary performance, because any excess capability is constrained by the controller algorithm. For example, bikes aimed at 150kg riders can use a chunkier motor, or a slower motor wind to manage the extra load in a way that respects the intent of the regulations.

Do it on your bike if it's so simple, show us how well it works, then we'll all get together and recommend it to the Government.

 

[guerney]

It isn't illegal for any bicycle to ascend hilly roads at 25kph. What enables me to do so, also enables me to drag heavy objects uphill on my bike trailer, removing my need for a (more) polluting car. I avoid cycling on pavements and trails. I only corner fast when going downhill, and only when in a hurry - I did the same with my acoustic bikes, so there's no change in my behaviour there. Except uphill, the lycra lot blast about faster than I do. "Blasting" along at any speed beyond 8mph exeeds motorist expectations, and that's one reason why I installed an Oxford lollipop on my handlebar's right side - I got pissed off at cars misjudging my speed and attempting to overtake ahead of traffic islands or bends etc. and screeching to a near stop alongside me when confronted directly ahead with the obstacle or oncoming traffic (difficult to capture on camera, it's front facing lens isn't 240° wide, and the lollipop has stopped it happening). In my 25kmh hill climb video, there were cars behind my bike which normally would have overtaken exactly like that ahead of those islands, but the were forced to loiter by the repulsive power of my lollipop. Here's video of a driver fleeing using the wrong lane ahead of the traffic island - another one effectively repulsed by my lollipop. My speed was 25kph. This fool didn't want to or couldn't slow down and wait.

And it's so much faster to stop uphill because of the non-removable gravity assisted braking system installed by default on every bicycle by Loki, very early during the coalescing of this universe, after the gigantuan fat alien lady sneezed in an adjacent universe constrained by very tenuous physical laws of no particular intent.

This approach also allows the use of any motor as needed to achieve the necessary performance, because any excess capability is constrained by the controller algorithm. For example, bikes aimed at 150kg riders can use a chunkier motor, or a slower motor wind to manage the extra load in a way that respects the intent of the regulations.

There's that "Intent" divining rod/crystallball/chicken entrails/tea leaves stuff again. It's why judges don't use them anymore, they use gavels. Do you really think if such a bright bunch of people had intended different, they wouldn't have defined the law accordinly? Advocating for a number of different motors for different weights of rider and terrains is bonkers, will be bonkers expensive for manufacturers and consumers. Making ebikes more inclusive is good for the planet. Keep 'em cheap. Mine cost £809 in total. Your 40Nm £3,000-ish(?) torque-sensored Shimano sounds crap, I doubt it'd be capable of anything I need it to do, certainly not with me riding it. Yours is an expensive bike for fit people, which doesn't suit the majority of average UK buyers. What's more, it has proprietary comms, hard or impossible for the owner to to repair at home at low cost. I detest all such bikes - they're e-waste waiting to happen long before necessary. I smell yet another failed Kickstarter or Dragon's Den pitch. Duncan Duncan Bannatyne would sound like he's choking on a rat sandwich swearing in Scottish.