What makes an efficient on road e-bike?

[coops]

If we get a high speed 20/25 mph class for e-bikes, gear drive really would be better then to cover that range.

I'm curious as to how the existing "derestricted for offroad" bikes fit into this picture? I'm guessing that the Torq would (hypothetically) benefit from gears to allow slower speed climbs, whereas the f-series are the opposite & would benefit by using derestricted power more carefully and prevent accidental offroad "power drains" from their massive 700W output? i.e. one to help climb hills, the other to help possibly extend range in cases of throttle miscontrol...

Stuart.

 

[flecc]

With the powerband on those needing to cover a 17 mph range, both would benefit from variable motor gears Stuart, subject to the gear system not losing too much efficiency, though as you've intimated, less so in power usage terms with the high powered F series, though more so in energy conservation terms.

This of course is very different from the 8 mph effective powerband requirement of Euro-legal bikes where drive through gear gain is very small now.
.

 

[coops]

I'm thinking an offroad "boost switch" would make more sense on an f-series than the Torq for instance, then, since you'd have sufficient power and legal speed limit when limited, and no excessive power usage worries so good range, and if you need the speed boost temporarily you can flick the switch! What do you think?

Either that, or very good throttle control, or the EAF pedelec cruise control system... .

 

[JohnInStockie]

Flecc - just a bit curious about the Panasonic motor. As I understand it, and please correct me if Im wrong, but that motor is particularly reliable as its was 'over-engineered'?

You have surprised me with your comments here. I agree that if your only going to travel at speeds between 0 - 15 mph, then whatever the incline, 5 gears is enough. However, if you then want to pedal alone upto higher speeds without assistance, then having those extras gears is a requirement.

So as I understand it, in theory it is better to have drive through gears. Yes if they break then obviously not and there will be more wear on the gearing mechs, but 2 bikes with everything working, large gear range irrespective of any national laws. In this situation I would expect the drive through gear system to be more useful and efficient.

Would it be possible to take a hub motor and attach it in such a way as to drive through the gears, maybe by attaching a second crank in the same way we have seen done with the Currie? Wouldnt that then be the most effective option?

John

 

[flecc]

That's a totally different thing John, you've introduced the rider and I was speaking only of the motor as you'll see if you check back. Of course the rider should have as many gears as is necessary for them. When I mention motor gears, those don't have to be the one's the cyclist uses, they can be incorporated within the motor.

BUT, I absolutely stick to what I've said above on today's motors, with a new drive through gear design, the extra complexity, the higher production cost, the reliability implications of the complexity, and the efficiency losses of gear systems are not overcome by the VERY small efficiency gain over the 8 mph effective power band necessary for the motor. The rider has no part in this.

Although as I said above, a very highly efficient two gear only system for the motor might just gain a tiny bit in theory, if the rider is introduced to the equation as you've done, I can show that makes that extra gear even more unnecessary and superfluous!

The Panasonic engineering wasn't in what I said above, I was referring to the hub gear reliability with both rider and motor power fed through it, and that extra complexity. However, reliable as they are, the Panasonic units have had failures within their complexity that would be impossible with a hub motor which has a much simpler design.
.

 

[flecc]

I'm thinking an offroad "boost switch" would make more sense on an f-series than the Torq for instance, then, since you'd have sufficient power and legal speed limit when limited, and no excessive power usage worries so good range, and if you need the speed boost temporarily you can flick the switch! What do you think?

Either that, or very good throttle control, or the EAF pedelec cruise control system... .

Definitely with the F series, but once we get into the higher performance areas more power will be used and so there's higher consumption. I don't think efficiency really matters then. They are opposing objectives. Best to choose which is important to you and pursue that with what's best. Save the planet or let it sink!
.

 

[flecc]

To John:

A further reply to my above, this time a summary of the position as I see it.

The Panasonic system bikes were miles ahead of the poor offerings elsewhere when they were designed at the start of the decade. They are still ahead, but only by a smaller margin now, so new drive through gear systems aren't worth designing for these reasons:

Wheel hub motors have advanced and do the whole job well and quite efficiently.

The small efficiency difference is offset by all the complexity costs, in design, manufacturing, reliability etc of drive through gear systems.

Cycle gears are not strong enough to take both rider power and the higher motor powers customers want.

The legal restrictions mean the usage band is from 5 mph to the legal power phase down point of 13 mph, only 8 mph in width, so the torque curve with a peak at 8 mph covers that easily.

There is a possibility that two gears for the motor with maximum power/torque at either 6 mph or 10 mph would provide better for steep hill climbing, but for the following:

a) Less power is needed to climb hills at 6 mph than at either 8 or 10 mph on other maximum torque points.

b) These are hybrid vehicles, human and electric. The rider power through their own gears does not have a power curve for our purpose. The rider's 100 watts or whatever is constant and just as available at 6 as at 8 mph. That evens out the supposed loss of power of the system as the motor moves from it's maximum on the curve, and in fact the riders contribution is a major proportion of the hill climb power at 6 mph on any hill. What's left for a reasonably powerful wheel hub motor to do is well within it's capabilities.

The efficiency loss that does still remain is offset by the increased performance of the powerful hub motor in most riding conditions, higher average speeds and climb speeds on moderate hills.

Therefore the only valid reasons to design a cycle motor drive through gear system now are for either heavy towing, or a high speed use off road.
.

 

[coops]

Since I've found the power used at 20mph is twice or more that used at 15mph, I've discovered two things: it does make good sense to match your top electric bike speed not too far above what you can pedal anyway, in order to contribute a larger proportion of the power needed & optimally improve motor & hence battery energy efficiency, and secondly that my own "happy medium" falls around 17-18mph tops, which with pedalling still gives me a reasonable range, with power consumption a bit higher than 15mph.

I agree with what's been said here about hub motors vs transmission up to 15mph, some good points there.

I think with the f's, with care on the throttle, especially on hills, and an eye on the speed (or your own chosen speed and/or current limit using something like this) you could still get good range & power consumption? After all, if you limit speed to say 18mph and throttle down for hills (setting the current limit to lower its 20A limit would reduce low speed torque...) why would your power consumption be much more than, say, a derestricted Torq at same speeds (so long as the motors are similarly efficient), and you could climb steeper hills too .

Stuart.

 

[flecc]

I think with the f's, with care on the throttle, especially on hills, and an eye on the speed (or your own chosen speed and/or current limit using something like this) you could still get good range & power consumption? After all, if you limit speed to say 18mph and throttle down for hills (setting the current limit to lower its 20A limit would reduce low speed torque...) why would your power consumption be much more than, say, a derestricted Torq at same speeds (so long as the motors are similarly efficient), and you could climb steeper hills too .

Stuart.

I agree, though I wouldn't want to ride around studying a DrainBrain, I'd just ride the thing.
.

 

[coops]

Sorry, I should have said set your speed/current limit with something like the drainbrain - you don't have to watch or study it flecc, that was my point - set the speed/current limit you want & go! I think at worst (so long as it works) its a more accurate battery meter (if one wants one - can be useful to avoid cut outs miles from home ) and a speed/current limiter to economise on energy use. What do you reckon?

 

[flecc]

Yes, that's different and I think you're right. On the Torq I found that riding speed choice of about 18 mph was also best for me, though naturally it's up a couple of mph now with the T bike. Presetting makes sense.
.

 

[coops]

Changing tack slightly.... efficient braking ...
The rear roller brake on the Torq was never as strong as the front V-brake, and with the T-bike you were able to attach rear V-brakes to the unused "brake fixing points" on the chainstay.

I've read your comments about disc brakes here flecc, and I could understand why you wouldn't want to comment, but if you can say, what are your thoughts about not having such a fixture for a rear V-brake alternative, or is there no problem? I'm concerned about disc brakes' stopping ability and more particularly their maintainence needs...

Stuart.

 

[flecc]

The cable operated ones are notoriously poor Stuart, and there's widespread criticism of the hydraulic ones as well.

Personally I think they can make sense on a rough terrain bike which is going to go though deep mud, since V and caliper brakes can clog up and even bind up a wheel in those conditions.

But on the road I see little point other than saving rim wear. The most important factor in disc brake efficiency is the disc diameter, and the rim is the biggest diameter disc brake possible. Although rim heat is a problem on long descents, heat is also a big problem on discs, particularly the thin cycle ones which can buckle easily. The heat isn't good for the hub bearings either.

I've never used them personally and have very little experience with the cycle ones, so have no idea on the maintenance demands. There seems to be a vast and ever changing range on the market though, so I wouldn't be surprised to find that locating spares and pads could be a problem sometimes.

I investigated the possibility of a disc on the motor in it's rear wheel position in the T bike and solved how to do it with a cheap Avid system costing only £35 using the standard cable. This was in case I used a different rim size in the end. The main problem was in welding on the IS fixing point that rear discs usually demand, front ones having a choice of two fixings usually, but I'm not familiar with the standards.

So overall I favour V brakes for all on road applications, though shoe based hub brakes have some appeal for a low maintenance utility bike.
.

 

[JohnInStockie]

Hi and thanks Flecc. As you can imagine though, I have just a couple of questions on what youve said there..

The Panasonic system bikes were miles ahead of the poor offerings elsewhere when they were designed at the start of the decade. They are still ahead, but only by a smaller margin now, so new drive through gear systems aren't worth designing for these reasons:

Personally I would prefer that they keep trying to improve on the designs, I just like us moving forward on a wide front. I realise the implications of all the extra cost of this, but we are still in early days, and so I would like to see even more unusual designs be tried rather just everyone reverting to 1 design.

Cycle gears are not strong enough to take both rider power and the higher motor powers customers want.

The Twist has done it quite well without a major adaption. Is the power too low on a Twist, and that's Ive not seen a thread of a broken hub on a Twist? It would be interesting to see exactly what the manufacturer tolerances are for different gearing types

The legal restrictions mean the usage band is from 5 mph to the legal power phase down point of 13 mph, only 8 mph in width, so the torque curve with a peak at 8 mph covers that easily.

I dont understand this. My Twist applies power from the first push on the pedal, and it should be possible to have a speed sensor that cuts out assistance completely at 25kph without phase down shouldnt it?

There is a possibility that two gears for the motor with maximum power/torque at either 6 mph or 10 mph would provide better for steep hill climbing

I dont understand this one Flecc. Surely if you are using you power efficiently with the correct gearing its always better? Surely its simply the physics of the gradient vs the mass and the force applied divided by the turning circle (or something like that )?

I would have thought that it would be one of Newtons laws to calc the energy needed, and then the efficiency of the motor, or is efficiency nothing to do with it?

The efficiency loss that does still remain is offset by the increased performance of the powerful hub motor in most riding conditions, higher average speeds and climb speeds on moderate hills.

I think there is something I am missing here. Does all this relate to power bands in some way, and not motor efficiency? If we are saying that the newer hub motors are more powerful in most conditions, then is that because they simply dont have a power band in certain areas?? This doesnt make sense to me. The Torq for example allows you to apply power immediately doesnt it?

Sorry Flecc, me no comprende


John

 

[coops]

So overall I favour V brakes for all on road applications, though shoe based hub brakes have some appeal for a low maintenance utility bike.

Thanks flecc

Sorry to interrupt the flow of the thread John, just one other quick thing re brakes, hope you don't mind :

If V-brakes cannot be fitted, is there another alternative to discs i.e. can a roller brake or the like be fitted in place of the disc, if necessary/wanted? Or does that depend on the bike setup?

 

[flecc]

To John

Personally I would prefer that they keep trying to improve on the designs, I just like us moving forward on a wide front. I realise the implications of all the extra cost of this, but we are still in early days, and so I would like to see even more unusual designs be tried rather just everyone reverting to 1 design.

Fully agreed, but economics rule.


The Twist has done it quite well without a major adaption. Is the power too low on a Twist, and that's Ive not seen a thread of a broken hub on a Twist?

There have been quite a few. Many of the 4 speed Nexus hubs failed so the SRAM that you have replaced it, and quite a few of the 3 speed as well.

I dont understand this. My Twist applies power from the first push on the pedal, and it should be possible to have a speed sensor that cuts out assistance completely at 25kph without phase down shouldnt it?

Yes but you don't cycle at zero mph. You cycle from 5 mph upwards to the phase down point, and thats an 8 mph band. It's the only motor drive area where gears could be used, and it's too narrow for that to be useful. Cars go from 0 to 30 mph on one gear for example, and they are comparable in this respect. The 13 mph power phase down you have is an absolute requirement in European law, so designing it out isn't an option.

I dont understand this one Flecc. Surely if you are using you power efficiently with the correct gearing its always better? Surely its simply the physics of the gradient vs the mass and the force applied divided by the turning circle

It's difficult to deal with this since I've already explained it and the points just aren't being understood for some reason. Yes, the theory is fine, but are you seriously suggesting it's necessary to change gear four or five times between 6 and 12 mph? Of course you aren't, or I hope you aren't, for you'd always be changing gear and never driving the thing!

Lets take an example. At 8 mph on the peak power and slowing down on a hill, we move away from the power peak. At 7 mph we are still within 5% of that peak level of power, but if we use a gear hub to bring the peak power to the 7 mph point, that adds a 15% efficiency loss, so we are at least 10% worse off by using the hub. In other words, until the power falloff away from the peak is sufficient to exceed the gear inefficiency loss, gears won't be worth using. Because modern Hall effect motors have such a broad and fairly flat top torque/power curve, you have to move a long way off the peak for the loss to be serious. But since we only have to accomodate an 8 mph wide band, how far can we move off it? Obviously not very far. Therefore, as I've said, only two gears can be accommodated with any efficiency in that narrow band, one with the peak at 6 mph for good hill climbing in the 4 to 8 mph region, and one with the peak at 10 mph to handle the 8 to 12 mph region before phase down commences. But since the two gear system adds all the disadvantages from complexity I've already expressed, and the power curve on Hall effect motors is so wide, it's still not worth bothering for any reason, environmental or practical.

I think there is something I am missing here. Does all this relate to power bands in some way, and not motor efficiency? If we are saying that the newer hub motors are more powerful in most conditions, then is that because they simply dont have a power band in certain areas?? This doesn't make sense to me. The Torq for example allows you to apply power immediately doesnt it?

Because a wheel hub Hall motor covers the load requirements at all speeds, it's made a bit more powerful than needed to ensure the power at the ends of the user band and off the peak is sufficient. Therefore it has a built in inefficiency. But that surplus built in power makes it faster on hills, in headwinds, in acceleration and even on the flat when its anywhere near the peak power sections of it's power curve. That's the payback which offsets the supposed inefficiency. I'm forever preaching this. Theoretical efficiency in isolation is worthless, the work done has to be included in the equation, and when that's included, the supposed inefficiency largely disappears, cancelled by the superior performance of the more powerful motor.

If that wasn't true, only the tinyest and weakest of everything would be efficient, clearly a nonsense.
.

 

[flecc]

Thanks flecc

Sorry to interrupt the flow of the thread John, just one other quick thing re brakes, hope you don't mind :

If V-brakes cannot be fitted, is there another alternative to discs i.e. can a roller brake or the like be fitted in place of the disc, if necessary/wanted? Or does that depend on the bike setup?

Yes any brake is ok, the hub determines these, so it means a wheel build with a suitable hub. Roller brakes are mounted on a short spline edge of the hub, but the shoe type brake drums are integral to those hubs. Both use a torque arm and a clip around a frame tube to prevent rotation.
.

 

[JohnInStockie]

@ Coops - never a problem mate

@ Flecc - thanks for your 'perseverance' on this. I have noted you may have mentioned this once or twice before Its almost there, the sponge is beginning to absorb some liquid...but can I please clarify just a couple of little points please..

15% efficiency loss using the gears...is that right? Is it that much using a chain drive and Hub gearing? I had thought that the maximum power loss through chain and gearing was about 5% as that is why we ruled out shaft drive cranks that lose about 8%??

Changing gears between 6 and 12mph....erm, yes I might. I understand what you are saying there of course, but I thought that it was good practise to try and maintain an average cadence throughout the gears, and to have a sufficient gear range to do this (Ive been trying to teach myself this )?

Hub motor a bit more powerful than needed....Ah-ha, thats what I now understand. So that they have thier 'excess power' in the main asist zone for where most would need it (or want it) e.g. 11-13 mph in the Torq?

Thanks

John

 

[flecc]

@ Flecc - thanks for your 'perseverance' on this.

No problem John

15% efficiency loss using the gears...is that right? Is it that much using a chain drive and Hub gearing? I had thought that the maximum power loss through chain and gearing was about 5% as that is why we ruled out shaft drive cranks that lose about 8%??

Derailleur gears can reach the high 90s in efficiency, but are often lower. The highest efficiency claim for a hub gear ever was on the SRAM 3 speed at 92%. But that's a bit of a con, since it included the middle gear in the tests which is direct drive, so not geared at all! Hub gear efficiency is generally about 85% overall, but a bit worse in the lower gears, so you see I was being generous.

Changing gears between 6 and 12mph....erm, yes I might. I understand what you are saying there of course, but I thought that it was good practise to try and maintain an average cadence throughout the gears, and to have a sufficient gear range to do this (Ive been trying to teach myself this )?

Yes, but again that's you, the rider. I'm speaking of the motor, and it's pointless that using lots of gears between 6 and 12 mph. After all, with each gear change the drive is lost. That cannot be efficient! Routing the motor through the hub introduces the hub gear inefficiency, and also exposes the motor to the whims of the rider changing gears at moments inappropriate to it. My guess is that the average rider will achieve less motor efficiency than if the motor is just left to drive continuously across it's powerband, since rider cadence requirements and power constant are not in any way the same as a motor's power curve. If a motor is to be geared, it's best it has it's own gears.

Hub motor a bit more powerful than needed....Ah-ha, thats what I now understand. So that they have thier 'excess power' in the main asist zone for where most would need it (or want it) e.g. 11-13 mph in the Torq?

That's it John, high for the Torq as you indicated, for speed and iffy hill climbing, but at about 8 mph for best all round 5 to 15 mph performance on most legal e-bikes.

A lot of this didn't apply for older brush hub motors and other oddities when the Twist was introduced, but as said before, things are changing.
.

 

[prState]

I've learned quite a bit in the last couple weeks.

So...if cost were not an option, how close are the current hub drives to being the best current technology can come up with? In other words, are they already 80% there, 90, or more in your best guess?