Just happened across this but can't comment on its accuracy

[Marctwo]

It doesn't change the fact that it's the rated output that's referenced, not the actual output. I'm sure you don't seriously think the legislators would expect the ebike manufacturers to test each individual motor to obtain the actual output for that specific motor.

If the rated output isn't on the label then you can look at the technical specifications for the motor.

 

[Jeremy]

It doesn't change the fact that it's the rated output that's referenced, not the actual output. I'm sure you don't seriously think the legislators would expect the ebike manufacturers to test each individual motor to obtain the actual output for that specific motor.

If the rated output isn't on the label then you can look at the technical specifications for the motor.

I think you're missing the definition of "rated output" in this specific context.

"Rated output" is defined in the text of the British Standard, and is a figure resulting from a measurement conducted using a specific method with specific test equipment at a specific temperature and with a specific battery voltage.

It isn't the fairly meaningless figure that manufacturers sometimes mark on the motor case, by a long way, in fact it's not even close.

The reason isn't that straightforward, but has to do with the physics of permanent magnet electric motors (either brushed or brushless). Any PM motor will be able to deliver a wide range of power output, often far greater than some of the fairly meaningless figures printed on them might indicate.

In practice, an ebike motor power output is set by the controller, not the motor. It is for this reason that the British Standard uses motor voltage when determining rated power. You can have a wide range of maximum motor voltage and motor current when using the same battery, because of the transformation between current and voltage that occurs in any ebike motor controller (the controller is an efficient buck converter, in effect).

On a more practical note, here is the motor manufacturers test data for the Tongxin 180W motor:



The test method used was not to BS1727:1971, in fact it was a very different test where the dynomometer load was increased with a fixed motor voltage applied and with the controller throttle set to 100%, with no temperature control or attempt to reach thermal equilibrium.

Look at the red line at the bottom. This shows the power output with varying torque. As you can see, this "180W" motor really delivers over 350W in practice.

To discover the continuous rated output of this "180W motor, Justin at Grin Cyclery has tested it on his dynamometer, using a methodology similar to that in BS1727:1971, but without the accreditation or equipment calibration certificates that the British Standard requires. His testing (the results of which can be seen by running the simulator on his website) reaches safe thermal equilibrium at about 350W continuous power output.

So, here we have a motor that has a manufacturers rated power of 180W ( I have one, the label clearly states 180W), yet would have a BS1727:1971 rated power that is around double.

It is for this reason that the law requires the motor to be rated using the British Standard. Does this make sense now?

 

[amigafan2003]

So, here we have a motor that has a manufacturers rated power of 180W ( I have one, the label clearly states 180W), yet would have a BS1727:1971 rated power that is around double.

So, the puniest motor commonly available has a motor rated at twice the legal max if using the BS standard.

I can only imagine it would be quicker to walk than to use a motor than actually confirms to the BS standard?

This alone confirms it's all a load of bull. It also raises the point that it's easier to ignore the law than it is to follow it - it's no wonder many do exactly that!

"If you have ten thousand regulations you destroy all respect for the law. "
Winston Churchill

 

[Jeremy]

So, the puniest motor commonly available has a motor rated at twice the legal max if using the BS standard.

I can only imagine it would be quicker to walk than to use a motor than actually confirms to the BS standard?

This alone confirms it's all a load of bull. It also raises the point that it's easier to ignore the law than it is to follow it - it's no wonder many do exactly that!

You can easily make the small Tongxin conform to the rating using the British Standard test method though, just by adjusting the controller current limit. All it illustrates is that the motor ratings used by manufacturers on their labels almost always bear little or no resemblance to reality.

In the specific case of the small Tongxin, then if you run it from a controller that has a low current limit you can keep the continuous power down to 200W output quite easily. It performs very well like this, if you also allow the controller to deliver a higher current for the first ten seconds of acceleration. This is legal under the BS test, as that test is a continuous power one, not a changing load one like the manufacturer has used.

The result is that by doing this you can take advantage of the higher power capacity of the motor for starting off, accelerating to speed or performing a hill start. With the right gearing the 200W continuous power is usually enough to give reasonable performance, if not great on hills (which is what you'd expect from a small hub motor anyway).

One of the popular controllers, the Xiechang range (a.k.a. "Infineon", "Lyen", "Keywin/e-crazyman" ect) has a setting to do just this, block time. It is specifically included to allow a high power boost under high torque load when accelerating, for a programmable time.

Edited to add:

Thinking about it, I strongly suspect that the early Panasonic ebikes would have met the UK 200W limit, as they were designed to meet the more restrictive Japanese regulations. IIRC they had a maximum continuous power rating of 200W, perhaps a little less.

They performed quite well, too. My neighbour has had one for a couple of years and it effortlessly climbs the steep hill at the back of our village, a hill that my old 1kw hub motor bike struggles a bit on.

 

[Marctwo]

It may help if you can post the actual definition of "rated output" that you are refering to.

 

[Marctwo]

You can easily make the small Tongxin conform to the rating using the British Standard test method though, just by adjusting the controller current limit. All it illustrates is that the motor ratings used by manufacturers on their labels almost always bear little or no resemblance to reality.

If it's so easy to control the output, surely it's easy enough to determine the output without testing.

 

[Jeremy]

I don't have the BS in electronic form, just a very old paper copy, but the gist of it is that the rated output is measured using a dynomometer brake that is calibrated to a British Standard, the motor temperature is measured and the motor is loaded to a set rpm (the loaded rpm) that is equivalent to the maximum power point. In practice this load will usually be around the 50% rpm point, i.e. halfway between the maximum no load rpm and zero, but has to be determined under test. The motor is run at this load until the temperature stabilises. If the temperature increases beyond the manufacturers safe maximum (usually it will be the maximum safe temperature of either the magnets or the wire, but the BS doesn't stipulate this) then the test is stopped, the motor allowed to cool down and the test is started again with a smaller load. It has to be run like this for a period long enough to determine that the temperature is stable. The test is undertaken at a fixed and controlled ambient temperature.

Once the motor temperature under load is stable, then the motor rpm together with the torque reading from the dynamometer are used to determine the motor continuous output power. This is then the rated output that would go on the data plate.

The test procedure includes the motor controller, but this is set to maximum speed for the test. Clearly the motor manufacturer can supply the motor for testing with a controller that limited the motor power, to ensure that it passes. The BS considers the motor and controller to be an integral unit for testing, so if the controller was changed or altered to increase the motor continuous power then this would make the data on the plate false.

 

[Jeremy]

If it's so easy to control the output, surely it's easy enough to determine the output without testing.

How? Unless you have a good motor dynomometer it's pretty near impossible. I've built a small electric motor dyno that uses an eddy current brake and strain gauge torque reaction sensor, but it has been very difficult to calibrate it accurately. I'd suggest that there aren't many dynos around that will measure down at these low power levels with any degree of accuracy.

Edited to add:

For interest, here's a link to pictures of my part finished motor dyno: http://endless-sphere.com/forums/viewtopic.php?f=2&t=24024&p=350082#p350002

 

[flecc]

I can only imagine it would be quicker to walk than to use a motor than actually confirms to the BS standard?

That was actually the case with the early British e-bikes in the 1980s, the power available was pathetically useless and it was that which impeded progress. It wasn't until the late 1990s that manufacturers started to stretch the rules and make e-bikes suddenly viable.

 

[flecc]

Thinking about it, I strongly suspect that the early Panasonic ebikes would have met the UK 200W limit, as they were designed to meet the more restrictive Japanese regulations. IIRC they had a maximum continuous power rating of 200W, perhaps a little less.

I'm quite sure you are right Jeremy, the graph that I posted earlier shows a mean output of just under 250 watts and it was from the second generation unit. That is markedly more powerful than the first generation one that was fitted to e-bikes made by Panasonic, Giant and BikeTec before mid 2006.

 

[NRG]

Jeremy you have the patience of a saint, really.

 

[Marctwo]

How? Unless you have a good motor dynomometer it's pretty near impossible. I've built a small electric motor dyno that uses an eddy current brake and strain gauge torque reaction sensor, but it has been very difficult to calibrate it accurately. I'd suggest that there aren't many dynos around that will measure down at these low power levels with any degree of accuracy.

Edited to add:

For interest, here's a link to pictures of my part finished motor dyno: Endless-sphere.com • View topic - DIY Dyno

Your earlier reply to amigafan2003 explains how you can easily set the output to accommodate the law. So if you know the output, why would you need to test it? The law doesn't require the test, just the output rating.

 

[Jeremy]

Your earlier reply to amigafan2003 explains how you can easily set the output to accommodate the law. So if you know the output, why would you need to test it? The law doesn't require the test, just the output rating.

How are you going to know what the output power is though? Being able to adjust the power output by adjusting the controller is fine, but without a defined method for actually measuring the power at the motor you aren't going to have a clue as to what to set the controller current limit at, are you?

There are several ways of measuring power and depending on the definition of the method and conditions you will get a different result, as illustrated by the Tongxin example I gave earlier. For example, I could decide that, for me, continuous output power means the power delivered when the motor is loaded for a period of one hour and the motor temperature stabilises at a temperature that is 20K above ambient. That's a perfectly valid definition, but isn't the one in the British Standard so would give a different result.

It is because the specific measurement method and conditions play such a big part in determining the continuous output power figure that the law, quite reasonably, requires it to be defined using a specified test method. The test method defined is repeatable and represents the legally defined continuous power, irrespective of any notional rating that a manufacturer might give the motor.

 

[amigafan2003]

What evidence do you have to provide that a motor complies with BS? Is a court going to send the bike - controller, battery and motor - off for testing?

What if any of those components were damaged in the accident?

Assuming I'm corpus mentus at the time of the accident it'd be quite easy for me to short the battery and destroy any evidence of it being 63v and later state it was a 36v battery fitted.

Actually, thinking about it, a fireball at the side of the road probably wouldn't help!

 

[Deleted member 4366]

Back to the Speedict then. The new version has reurrected the panic button, which instantly limits the current and speed to whatever limits you've set. Nobody knows you've done it because it's on your phone, not the bike. The Speedict would be part of the controller.

 

[Jeremy]

What evidence do you have to provide that a motor complies with BS? Is a court going to send the bike - controller, battery and motor - off for testing?

What if any of those components were damaged in the accident?

Assuming I'm corpus mentus at the time of the accident it'd be quite easy for me to short the battery and destroy any evidence of it being 63v and later state it was a 36v battery fitted.

Actually, thinking about it, a fireball at the side of the road probably wouldn't help!

The British Standards Institute accredit test houses to perform testing to their own standards, as do a number of other accreditation bodies, so if the test was performed by a test house with the appropriate accreditation then it would be OK, I'm sure.

There's no need for the whole bike to be tested, just the motor and controller (which are considered to be a unit for the purposes of this particular test). The nominal battery voltage has to be established using the method in the standard, but because the test takes a long time the battery doesn't need to be used for the test, a power supply that emulates the battery voltage under load would be OK.

If the components were badly damaged then it seems unlikely that any verification could be made after the event. TBH, I doubt very much that an ebike that "looked" like it might meet the limit would be subjected to testing anyway, unless the accident was very serious, which isn't that likely, I'd have thought.

The question really comes down to what an individual is prepared to accept in terms of risk. This is a balance of probability (i.e. how likely are you to ever get caught, given the area where you ride etc) and impact (i.e. what might happen in the event of getting caught).

These are both going to depend to a great extent on how the ebike looks (I'd suspect something with a battery the size of a suitcase and hub motor the size of a large frying pan might make getting caught more probable) and where and how you ride (riding somewhere like central London probably increases the risk of getting involved in an accident than riding around the rural lanes of Wiltshire, for example).

Personally I'm quite happy to accept that my low power folder is very unlikely to ever get me into hot water. It's sufficiently discreet in appearance, and I ride in areas and at speeds such that a major accident is unlikely, and for me the risk is acceptable. Others may have a different tolerance to risk, though, like the lady down the road who I'm certain wouldn't tolerate an illegal ebike at all.

Back to the Speedict then. The new version has reurrected the panic button, which instantly limits the current and speed to whatever limits you've set. Nobody knows you've done it because it's on your phone, not the bike. The Speedict would be part of the controller.

If your happy with the deception, then yes, it probably would work well. You could also just use one of the speed switch settings as a legal one, too. I did this on a bike built for a friend, using a reed switch inside the controller wired to one of the programmable speed limit pads. This was activated by a small magnet, that when pulled off switched the controller to "legal" mode, with no indication that there was any such switch.

 

[Marctwo]

How are you going to know what the output power is though? Being able to adjust the power output by adjusting the controller is fine, but without a defined method for actually measuring the power at the motor you aren't going to have a clue as to what to set the controller current limit at, are you?

Simple arithmetic. If you control the voltage and the current then you control the power.

Of course, you already know this which is why you could describe how easy it is to set the controller to pass the BS test.

 

[Jeremy]

Simple arithmetic. If you control the voltage and the current then you control the power.

Of course, you already know this which is why you could describe how easy it is to set the controller to pass the BS test.

OK, I see where your coming from and where the misunderstanding is. The current and voltage going into the controller from the battery (which is the only reasonable place to make such a measurement) determines the input electrical power to the whole system, motor and controller. The law requires a definition of the mechanical output power from the motor only. The two are different, and highly load and temperature dependent.

Take a look at the Tongxin data I posted earlier. Look at the green efficiency line. This is the ratio of electrical input power to mechanical output power. Note that it changes massively with load, so unless you know the motor torque load and rpm it is near-impossible to determine mechanical output power from electrical input power.

Another complication is that this test data comes from a short duration motor test where the motor load is smoothly increased and the data logged. The motor doesn't get time during the manufacturers test to get warm. When motors warm up, as they do under load, the efficiency gets worse, so you get less mechanical output power. One cause of this is that the winding resistance increases quite a lot with temperature and higher winding resistance means greater I²R losses, hence poorer efficiency.

 

[Old_Dave]

Back to the Speedict then.

Which is a very cool device... (apart from it only using android and not iphone, but that's another story)

But I'm inclined to think that requiring user input on a cellphone to enter 'panic mode' is not the best method.. the far safer proximity detection method with a default low power on fail.

Jeremy's suggested old technology reed switch is a good basis / starting point for something that's cheap and simple.

Of course the 'rated output' would still a mystery, lol

 

[Jeremy]

Which is a very cool device... (apart from it only using android and not iphone, but that's another story)

But I'm inclined to think that requiring user input on a cellphone to enter 'panic mode' is not the best method.. the far safer proximity detection method with a default low power on fail.

Jeremy's suggested old technology reed switch is a good basis / starting point for something that's cheap and simple.

The reed switch did work well. The magnet was stitched inside a bit of velcro and sewn to a small strap a couple of inches long. There was a small bit of mating velcro stuck where the reed switch was. pulling the tab just switched the controller to "legal" mode and was easy to do. We did contemplate fixing the magnet to one of those safety lanyards you get on jet skis, so getting off the bike automatically switched it to "legal" mode, but felt it was a bit too messy.