Category Archives: Technical

Brompton Folding Bike 10 Speed

Brompton Gear Range

Brompton Folding Bike 10 Speed

Brompton 10-speed! Sturmey 5-speed hub, Brompton 2-speed shifter and adapted Highpath sprockets

Brompton Gear Range ‘Could the Brompton be ‘easily’ redesigned to take wider,more conventional hubs? The rear drop-out spacing is under 115mm,whereas ‘normal’ hubs are 135mm (with 145mm normal in the USA).As the Brompton folds to the right, the extra width might only be 10mm or so. Could the hinge be widened by 5mm? Are there other design elements that would have to change?’ Simon Avakian, Palo Alto,California USA

‘I have a new 2×6-speed S-type Brompton with a Schlumpf Speed Drive which I am very satisfied with. But the gear range is only just acceptable – my wish would be gears from 2m to 10m (25″ to 125″).I can’t help dreaming of a Brompton with a Rohloff Speed Hub without widening the rear frame. I have seen a German MTB-site where the Rohloff hub gear is mounted in the frame instead of the rear wheel.’

Jon Dreyer Rensmoen,Norway

Several engineers have stretched the Brompton rear frame to accept wider hubs, such as the Nexus 8speed and even the 12-speed Rohloff.There’s no need to alter the hinge,and as Simon suggests,the bike emerges little wider than normal.It might even be possible to fit the hub elsewhere,but these are heavy and expensive conversions,and for most purposes it’s debatable whether either would be worthwhile.Surely we can come up with a lighter, cheaper and more elegant solution?

Let’s look at the proprietory equipment that can easily be fitted to the Brompton and try mixing and matching these bits and pieces to produce a widerange system.The widest range hub that can be fitted without serious engineering is the 5-speed SturmeyArcher,which is now back in production, although not fitted as standard by Brompton,as it once was.This hub has a limited 225% range,but it’s light and cheap,and we can extend the range by fitting dual chainrings (which don’t work well on the

..Is this the way Brompton will go? A10-speed would certainly satisfy the critics…

Brompton) or dual sprockets,which do.Fitting Brompton’s own 2-speed derailleur to a Sturmey 5-speed widens the range to 260%,which is useful,but hardly ground-breaking. For the really big gears we need something like the Highpath 12/18 tooth conversion.This is designed for use on the roomier SRAM 3-speed hub,but with a bit of machining it is possible to squeeze one onto a Sturmey 5-speed.

This combination gives a gear range of 337%;more than any hub gear except the expensive Rohloff,and without the weight and complication of fitting a crank-mounted Speed Drive.The idiosyncrasies of gearing mean that the bike is really a 7-speed,because three ratios are almost identical: it’s best to think of the conversion as a 5-speed with two extra gears at the bottom (or the top).Using the Brompton 44-tooth chainring,the gears emerge as: 41″,49″, 62″,79″, 93″ in the high range,and 27″,32″,41″,52″ and 62″ in the low.A smaller 40-tooth chainring would give gears of 25″ to 85″.We haven’t produced Jon’s 500% range,but we have done quite well for a weight penalty of only 200g or so.

To my surprise,the prototype system works more or less without grumbles or adjustment.In practice,one tends to leave the bike in the low range or high range for quite long periods, using the 2-speed changer only when a more extreme gear is needed.

The bad news is that a 10-speed would be quite difficult to make at present. Highpath has wisely opted not to produce a thinner and weaker sprocket,and it’s unlikely Sturmey will offer to produce a wider ‘driver’ on its 5-speed hub!

Is this the way Brompton will ultimately go? A 10-speed would certainly provide enough range to satisfy the critics,and the company could introduce one very quickly.

A to B 55 – Aug 2006

Hub Gear Conversion

Hub Gear Conversion

Hub Gear ConversionIf you think about it, the quality of a child’s bike is really important. If a child grows up with a heavy, impractical bicycle, he or she starts life with the impression that bicycles are heavy impractical machines.The evidence from the current generation is that apart from dabbling with BMX, the vast majority stop riding bicycles just as soon as they can, and most never return. Rather disturbingly, there’s growing evidence that many in Alexander’s generation will never learn to ride at all.

When Alexander was old enough for a ‘proper’ bike we chose a German-made Puky. Children’s bikes with dynamo lights, a rack and mudguards are common on the Continent, but only the Puky is easily obtainable here, thanks to importer Amba Marketing. By the spring of 2005, our 18-inch wheel example (see A to B 41) had given great service for 18 months and 400 miles, but at six, the boy was growing rapidly, and trying longer rides.The time had come for an upgrade, but not yet to a bigger bike.


…any competent cycle engineering should be able to upgrade…to hub gear operation…

How would we define the perfect bike for a small child? For obvious reasons, it needs to be reasonably fashionable. If everyone else is riding Death-Squad BMX UXB MTBs, with unobtainium gussets and nobble-tooth mud pluggers, pushing the sensible, weedy option can be hard work.The bike also needs to be suitable for road use in all weathers, plus some modest off-roading, and come equipped with user-friendly hub gears, brakes, mudguards and lights. Quite a tall order, really.

Puky sell a range of fully-equipped 20-inch bikes, and a few 18-inch bikes, but none of the smaller machines have gears.The answer was to upgrade what we had, adding a Sturmey Archer S-RC3 hub to Alexander’s Puky 18-1B, producing, one assumes, an 18-3B. The beauty of using this rare hub is that it also comes with a back-pedal operated ‘coaster’ brake. Fitting something like this might sound complicated, but any competent cycle engineer should be able to upgrade a single-speed or derailleur-geared bike to hub gear operation.

Three-speed hubs used to be almost universal in Britain, but the arrival of cheap, sexy-looking derailleurs changed all that, and enclosed hub gears are now generally confined to roadsters and small-wheeled bikes. As we point out on a regular basis, this is most unfortunate. Few adults understand the principles of riding with close-ratio derailleur gears and for children, three gears are more than enough to think about.

Hub gears can be changed whilst stationary, making them ideal in traffic (or for those of a forgetful disposition at any time) and although the number of gears might sound modest in this number/size obsessed age, even the most basic hub provides a decent gear range. The range – for those who aren’t quite sure – is the difference between top and bottom gear. A wide range of gears enables the bike to nip along under a wide range of circumstances.

With Alexander’s friends acquiring MTBs with five or six derailleur gears, we found ourselves trying to explain that a SRAM, Nexus or Sturmey three-speed offers a gear range of around 180%, which is about the same as a cheap six-speed derailleur.There’s a widespread belief that hubs are less efficient, but a three-speed should return efficiency of 94-95%, a figure that a cheap derailleur would be pushed to achieve after a few weeks’ youthful abuse. It also comes with bullet-proof indexing and is almost immune from throwing its chain off.

Hub Gear Conversion


Coaster brakes have never really caught on here, but having seen a child grow up using one, we’re converts, and most parents on the Continent would probably agree.When you’re learning to make hand signals and keeping an eye open for traffic, there’s a lot to be said for controlling the primary The coaster hub looks brake with your feet. For as if it was made for dad, there are 33% fewer the bike. Note the ‘extra’ cables to adjust and spoke holes and rather lubricate.We hope you’re avant–garde spoke pattern convinced.


We won’t bore you with the fitting process – if you know what you’re doing, it’s easy, and if you don’t, we’d recommend visiting a good bike shop. Most of the shops that advertise in A to B can carry out this sort of work, but as usual, the real experts are Bicycle Workshop in Birdy rim will fit the larger 355m (18- West London, who regularly upgrade children’s bikes (and adult cruiser bikes) to this sort of spec. If doing the work yourself, the hub costs £65 with a lever changer, plus £8 for the twistgrip. Expect to add around £35 if the shop does the work for you.

The new hub weighs 400g, so with cables and twistgrip, the weight penalty for upgrading from a single-speed has been about 1kg. Starting with a derailleur, you’ll be removing sprockets, cables, a brake lever and a brake caliper, so the weight will be about the same. Gearing depends on circumstances – we fitted an 18-tooth rear sprocket, giving gears of 26″, 35″ and 47″. Broadly speaking, that’s one gear for the flat, and two hill- climbing options. Bottom gear will tackle 12.5% (1:8), taking care of most of the local hills.

Is the boy pleased with his gears? What boy wouldn’t be pleased with a TSS32 shifter, shiny S-RC3 hub, 178% range, 18-tooth sprocket and a host of other part numbers? At six, life is all about numbers. For Alexander, the back pedal brake is familiar territory of course, but the gears took a few days to get used to. Cycling mileage has since rocketed to about 60 miles a month, and the unusual machine, with its novel lights, rack and gears, seems to be much admired.

You can’t win of course. Alexander knows a thing or two about hubs, and he’s already applying subtle pressure for a five-speed. Sturmey doesn’t make a coaster five-speed, but the indestructible SRAM P5 is available in coaster form…The perfect 20-inch bike?


The Sturmey S-RC3, like most hubs, comes drilled for 36-spokes, but children’s rims – including our rare-in- the-UK 355mm rim – are usually drilled for 20. Its unusual to find 305mm (16-inch) rims drilled for 36-spokes, but a inch) bikes. In the largest 20 and 24-inch sizes, there are plenty of rims and tyres to choose from.To make life difficult, we decided to re-drill the old rim to take 18 spokes, lacing the wheel using alternate spoke holes, braced with a single 13G spoke to prevent the wheel ‘winding up’ under braking.This arrangement would be too frail for an adult bicycle, but for a child weighing 22kg, a new rim and 36-spokes aren’t really necessary. On our single-speed bike, we also needed to stretch the rear drop-outs slightly, but it’s more likely that adjustment would be needed in the other direction on a derailleur bike.

Hybrid Motor Gearing

Electric Bike Hybrid Transmission

Professor PivotThe time-honoured bicycle drive system of pedal cranks, chain drive, and hub or derailleur gears has been in vogue for more than a century, and shows no sign of going away in the immediate future. But new thinking is starting to make an impact, and I am indebted to engineer Frank Moeller for his thoughts on the future of bicycle transmissions.


Frank’s inspiration came initially from a desire to produce a more efficient electric-assist bicycle. Electric bicycle motors can be more than 80% efficient, but a bicycle is a demanding environment, and motors generally work efficiently over a limited speed range.To keep the motor spinning close to this ideal speed, it needs to be near the pedals, as on the Panasonic drive system fitted to the Giant Twist.

This arrangement ties the motor speed to a comfortable pedal cadence, cleverly utilising the human engine to select a comfortable gear for both leg muscles and electric motor. Generally this works well, but the motor speed still varies a good deal, and with transmission losses, overall efficiency probably doesn’t exceed 60% in most cases. In other words, a typical power-assisted bicycle carries around a heavy, expensive battery, yet turns 40% of its capacity into worthless heat.That’s a better performance than most internal combustion engines, but on a bicycle – where weight is a serious issue – this poor performance is inexcusable.

Most electric vehicles throw power away on descents too.The motor could run as a generator, putting power back into the battery, but with most older designs, the noise and friction involved generally outweigh the potential benefit of recycling a little of the energy. How can matters be improved? it seems a number of avenues are worth exploring.

New Technology

Without going into too much technical detail, large ‘switched reluctance’ and ‘induction’ motors are already available, and they’re much more efficient than older types. Combine the best of these designs and there’s no reason why a new smaller ‘hybrid’ couldn’t be produced – light enough and efficient enough to fit inside a conventional bicycle hub, turning quietly and with little friction whenever the wheel turns. Such a motor would give assistance up hill and braking down hill. None of these new motor designs have yet been optimised for light electric vehicles, but engineers are working on the problems.

Today, motors and gears are usually so crude and noisy that the system necessarily spends much of it time disconnected.The new hybrid motor would be fitted inside the front hub, driven at speeds of 3,000rpm or more through a single-stage stepped epicyclic ear.This permanently engaged motor would run reasonably efficiently across a broad speed range and provide both assistance and braking.

It’s a neat idea, but Frank Moeller’s key conceptual breakthrough is to take this arrangement a step further and design a completely new bicycle drive system, based broadly on the transmission of the ‘hybrid’ petrol/electric Toyota Prius, but in this case combining and regulating human and electric power inputs.

Toyota Prius

The Prius has won many awards since its launch in 1997, and it’s easy to see why.This outwardly conventional car draws power from a small and relatively efficient petrol motor linked to the planetary gears of an epicyclic gearbox.

Hybrid Motor Chart

This is the same compact, efficient device used in hub gears, but in this case its purpose is to split the motor torque into two streams; one leaving through the sun gear to turn an electrical generator, and the rest going via the outer ring gear to the car’s wheels. On some designs the shaft to the wheels can pick up extra torque from an electric motor, but on others the motor is mounted at the other end of the vehicle to give four-wheel-drive.

This ‘Hybrid Synergy Drive’ might sound complicated, but it does away with the conventional clutch and gearbox, and performs most of the braking functions, because the clever ‘torque splitting’ arrangement functions as a continuously variable transmission. At low speed the petrol engine is turned off, and power is drawn from the batteries, via the electric motor. As the vehicle accelerates, the petrol engine is started, feeding power to the epicyclic. At first, the output shaft and wheels are turning slowly, so most of the power is diverted via the generator and converted to electrical power to feed the motor. This might be described as ‘first gear’. As speed rises, the ‘braking’ effect of the generator is progressively increased by electronic means, transferring more and more of the available torque directly to the wheels, with a smaller percentage being turned into electrical power by the generator. Above a certain road speed, the generator is given a high electrical resistance and the unit is effectively in ‘top gear’ As driver input and road conditions vary, the amount of power flowing from the generator to the motor is continuously adjusted, keeping the petrol engine turning at an optimum speed under almost all conditions.

Hybrid Motor Hub

Patent diagram

Hybrid Motor Gearing

Bench test rig In both cases, input power is spilt into mechanical and electrical components, which are recombined in the output stage

For hard acceleration, extra power is supplied from the batteries, which will be recharged from the motor/generator when the going gets a little easier, or under braking. Back in low-speed stop-start urban driving conditions, the inefficient petrol engine is turned off, and the battery/electric motor combination takes over again. The torque split allows this clever transmission to make the best use of the very different characteristics of the electric motor and petrol engine, and it can be arranged to do the same with an electric motor and human ‘engine’. Ignoring the on-board battery for he time being, rider effort would be applied to the pedals as normal, and conveyed to the rear hub via a chain drive.

…the hub actually contains fewer parts than a typical five-speed hub…

In the hub, some of the human input would proceed direct to the wheel, with a proportion being diverted via a generator/motor electrical circuit, as on the Prius, but in this case the low pedal speed is geared up to give a high generator speed.When climbing a hill, the wheel would begin to slow, causing an increasing amount of torque to run via the generator.The electrical output from the generator would run just a few centimetres to an electric motor/generator, with the mechanical output being fed to the hub shell through a second epicyclic gearbox.

It might look complicated, but the hub actually contains fewer parts (and far fewer wearing parts) than a typical five-speed hub, and it’s fully automatic. And of course the system really lends itself to electric-assist, requiring just a battery and a few control circuits to turn the human-powered vehicle into a hybrid. As on the Prius, the battery would provide additional power for acceleration and hill-climbing, and absorb ‘waste’ power when coasting downhill. It would also be possible to recharge the batteries with pedal effort under favourable conditions.The major difference is that in stop-start town traffic, where the Prius petrol motor would be turned off, the primary input would be from the human ‘engine’ rather than the battery.

A New Era

A ‘torque split’ transmission would provide an HPV or assisted-HPV with a number of hitherto unattainable attributes. For the human ‘motor’, the hub gives a foolproof continuously variable transmission, with no gears to worry about.The hub could be set to provide either a constant input torque, constant pedal cadence, or a combination of the two, maintaining this optimum level under all conditions.The rider would simply point the bike in the right direction and start turning the pedals.Without the shock-loads imposed by frequent gear changes, maintenance would be reduced, and it might be possible to reduce the size and weight of components such as the crank and chain. And with the output motor able to function as a generator, recycling some of the power normally turned to heat by the brakes, the bicycle could be fitted with a small battery, or even a ‘super capacitor’ to store braking energy.This reduces the need for powerful brakes, reducing the size and weight of the braking system too.

Although bristling with technology, the hub would be simple, and easily fitted to a conventional bicycle.The one unit would replace the conventional gears, and reduce the weight and complexity of the transmission and brakes. On an electric bike, it would also replace the electric motor, wiring and control equipment.

Fact or Fantasy?

How close is this vision to fruition? In motor car terms, the engineering is quite simple, but scaling the technology down will present many challenges. Prius consumers were initially nervous about the reliability of the novel electrical components and especially the batteries, but Toyota had sufficient confidence to offer a 100,000-mile warranty on the hybrid drive, which has proved extremely reliable.The same rugged simplicity and fully enclosed transmission would suit a bicycle very well. Moeller is already working with business partners in Taiwan, with the intention of mass producing hub units for just a few hundred dollars, to be fitted to new bicycles or sold as after-market accessories.The future may arrive sooner than you think.

Our grateful thanks to Frank Moeller. For further information, Frank can be contacted at

A to B 49 – Aug 2005

Lower gears on a folder?

“Could you tell us the options for putting very low (sub 30-inch) gears in a folding bike? Which folding bikes will accept a Rohloff hub gear? It’s very hilly around here.”

Margaret Lunnon, Coulsdon, Surrey

Alot depends on whether you are prepared to put up with a general lowering of the gears, or are hoping to keep the high gears as they are.The simplest and easiest solution, common to most small-wheeled bikes, is to fit a smaller chain ring.These are cheap and available in a variety of sizes. An even cheaper answer with hub geared bikes, is to fit a larger sprocket. Most small machines come with 13- or 14-tooth sprockets and most will accept a much larger sprocket, giving an instant and cheap gear reduction. Bear in mind, though, that these solutions lower all the gears and have no effect on the gear ‘range’ – the difference between first and top gear.

Schlumpf Mountain Drive

Schlumpf Mountain Drive - the assembly fits into the bottom bracket

There are many options to increase the gear range, some more practical than others. On a derailleur bike, larger sprockets and long-arm derailleurs tend to be impractical with small wheels, but the other extreme of fitting a small top-gear sprocket and smaller chainring is fine, if a little complex.The Shimano Capreo system is specifically designed for small wheel bikes and gives a gear range of almost 290% from a very neat little unit. Assuming a just practical 80-inch top gear, the Capreo would give ratios down to 28-inch, just inside your requirement.This system is available on the Birdy White and the new Mezzo D9, and can be retro-fitted to most derailleur machines at a price.

An easier solution open to just about every folder is the Schlumpf Mountain Drive.This is a little hub-style gear set that fits inside the chainring. In direct drive, it has no effect, but click your heel on the little control button and the bike changes down to a low ‘hill-climbing’ set of gears.The Mountain Drive gives a reduction of 2.5:1 and the similar Speed Drive a more reasonable 1.65:1. Assuming a top gear of 80 inches, and a 3-speed hub, the Speed would give a bottom gear of 27 inches and the Mountain an almost unusably low 18-inch gear.With a wider range hub, the effect is even more pronounced. The nice thing about this system is that there are no cables to worry about, but it’s heavy, quite expensive and somewhat inefficient in the more extreme ratios.

Brompton Wide Ratio Conversion

Brompton Wide Ratio Conversion - 12/18t sprockets squeezed onto a standard 6- speed hub

My personal favourite is a wide-range conversion of the Brompton 6-speed.The standard 6-speed has a range of only 215%, so try as you might, you won’t get anywhere near your 30-inch target and keep a decent top gear. But fit 12- and 18-tooth sprockets to the Brompton hub and the range increases to 282%.This is wide enough to give a range from 80 down to 28 inches, with good efficiency. Unfortunately, there are technical issues involved, and no one produces a kit at present, although I understand this omission is being rectified. More in a later issue.

Particularly fashionable at the moment are the 8-speed hubs from Sturmey Archer and Shimano.The Sturmey is designed around very large sprockets and is thus too cumbersome for most small bikes.The Shimano is much more practical, and its gear range of 307%, would give a bottom gear of 26 inches with our 80-inch top gear. Several small- wheelers are available with this hub, including the Airframe, and conversions are available for most small-wheelers. Even the Brompton can be adapted, although the extra weight and folded width would be a disincentive for regular commuters.

Rohloff 14-speed hub

Rohloff 14-speed - the widest range of any hub gear

Finally, we come to the Rohloff 14-speed hub. This offers an enormous gear range of 526%, but it’s a big, brutal device, and more or less doubles the price of a typical high-end folding bike.The range is as much as most people could ever use, from a wobbly walking pace, to a flat-out down hill top gear. Using our example of a bicycle with an 80-inch top gear, the Rohloff would reach Margaret’s 30-inch criteria in gear 4, and go as low as 15 inches in bottom gear! The hub is fitted as standard on the £1,880 Birdy Grey, the £2,000 Brompton- based SP and can be shoe-horned into most other bikes if your budget extends to that sort of thing.

Elsewhere: Puncture-resistant tubes A to B 35 Brompton Wide-Ratio Conversion A to B 31 Brompton Mountain Drive Conversion A to B 21 PBW/Rohloff Speedhub A to B 31

Schwalbe Marathon Plus Tyres

Puncture-Resistant Tyres

No More Punctures, Please!

Professor Pivot“I have had three punctures on my Brompton since Christmas.The first in the front Marathon, the second in the rear Brompton tyre, prompting me to change it for a Marathon.The third was in the front Marathon again, causing me to regret spending on the rear tyre change. Just to make it worse, Mosquito Bikes fitted a 5/8in (16mm) Schwalbe 4a AV tube that is supposed to stretch to 13/8in (37mm). When it’s stretched that thin, how long will it last?

What BSI tests, if any, apply to kevlar-reinforced tyres? If there are none, what are the closest applicable to motorcycle tyres?”

Mike Hargaden, London

Punctures are a big problem for some people, under some conditions, while others hardly get to see a flat tyre these days. In general, any solution that prevents foreign bodies penetrating the tube (and there are many) will increase rolling resistance, because tough or springy extra layers don’t like to flex, as a tyre must.They obviously increase weight too, and extra weight in the tyre can result in a less sprightly ride and slower response, making the bike feel turgid and heavy. So, even when they work, it’s not all good news.

As for 16-inch (349mm) tyres, the A to B readers’ tyre survey (see A to B 40) suggests that the standard Brompton tyre punctures about every thousand miles, the ‘Green Flash’ kevlar-banded Brompton tyre a little more frequently, and the kevlar-banded Marathon about every 860 miles.This tends to back up my own observations, so taking weight, price and poor rolling resistance into account, I simply wouldn’t recommend kevlar tyres. It may well be that certain types of band perform better than others, but I have yet to see published research on the matter.

Puncture Resistant Inner TubePuncture-proof, at a terrible cost, are the so-called ‘solid’ tyres.These may be just tolerable on a 26-inch wheel, but at 16-inch the high rolling resistance and ‘wooden’ feel make these things more trouble than they’re worth.The same goes for ‘solid’ foam inner tubes, about which the less said the better.

Raleigh extra-thick Puncture-Resistant inner tubes provide a good low-tech compromise, but these may no longer be available in small sizes. I fitted a pair to a Brompton two years ago and I’m still waiting for the tyres to deflate. Schwalbe uses a similar technique on its Marathon Plus tyre, which has a 5mm thick flexible india rubber belt under the tread. These tyres are heavy and only produced down to 20- inch (47x406mm), but a 349mm variant should be available soon. I am currently testing a pair of 20-inch examples and will release data in future issues.

Schwalbe Marathon Tyres

Schwalbe Marathon Plus - the thick rubber band prevents penetration

Many proprietory tyre liners are available, but if wrongly fitted these can cause more trouble than they’re worth. As for the tube, Schwalbe does indeed offer a simplified line- up, with typically four or five similar tyres sharing the same tube. However, you should have been offered the 4AV, which is designed to stretch from 28mm to 37mm. I can find no record of the 16mm Schwalbe 4a, which sounds very tiny, and would be quite unsuitable for stretching to 37mm.

Motorcycles do, indeed, puncture less frequently than bicycles, but they also have a lot more power available to roll the tyres, which can thus be made a lot thicker. As minimum horsepower machines, bicycles will always require lightweight and thus vulnerable tyres.

Brompton tubeless tyres

Small Tyre Design

Professor PivotThe subject of tyre technology comes up rather frequently on these pages, mainly because their inherently higher rolling resistance tends to put small-wheelers at the cutting edge.

In the mid-1990s, small tyres were at a considerable disadvantage against their bigger cousins in terms of rolling efficiency, but this was much reduced with the arrival of the Primo and Brompton tyres, whose paper-thin sidewalls flexed more easily as the tyre rolled, reducing rolling resistance. Ever since, the boffins have been burning the midnight oil searching for further gains, with the primary work being carried out at Greenspeed, the Australian recumbent manufacturer, and Brompton. Folding bikes need small wheels for reasons of folded size, of course, but recumbent designers are showing an interest in the same tyres, primarily looking for a small frontal area and reduced wind resistance. As we saw in A to B 39, Greenspeed is starting to adopt the 16-inch (more correctly 349mm) tyre on its recumbent trikes for just these reasons, and I am indebted to the company for access to its latest research in reducing the already small rolling penalty inherent with these tyres.

The Theory

Why does a tyre experience rolling resistance? Most of the energy is absorbed around the contact patch, the crucial zone where the doughnut-shaped tyre and inner tube mould themselves briefly to the flat road surface. If tyres were 100% springy, this wouldn’t matter (although the rider would probably fly off on the first bump), but rubber exhibits a useful self-damping characteristic known as ‘hysteresis’, which effectively means that not all the deformation energy is recovered when the tyre resumes its shape. This damping effect turns motion into heat, and the process takes place continuously as a tyre rolls. In really bad cases, the tyre will feel warm after a hard ride. Small tyres offer greater resistance than larger ones, because the more sharply curved tyre has to bend more acutely to become flat, and visa-versa when reassuming the curve.

As every cyclist, motorcyclist, and indeed motorist, should be aware, the easiest way to reduce the size of the contact patch, and thus the rolling resistance, is to put more air pressure in the tyres.This seems to work in two ways – firstly by reducing the circumference of the contact patch – the ‘battle front’ of rubber doing the work – and secondly by reducing the angle through which the rubber has to flex when it hits the road. Watch an old chap ride past on a Raleigh Shopper with half-inflated tyres, and you will see all the negative factors at work: a small diameter tyre, large contact patch and extreme angles of flex. One is sometimes tempted to offer a few pump-fulls of air.

There is, however, a limit to the improvement that can be made through air pressure alone, particularly on a bicycle without suspension. Pneumatic rubber tyres have been so very successful because they absorb lumps, bumps and vibrations from the road surface. Inflate a tyre really hard and it begins to act like a rigid disc, which would only be good news if the road were as flat and smooth as a mirror. In practice, roads are more or less corrugated, and a solid tyre will pass these surface irregularities to the vehicle and rider. This is not only uncomfortable, but wasteful, because energy is thrown away as the bike vibrates – effectively lifting and dropping the mass of the bike and rider.

Wider Tyres

Greenspeed Scorcher tyres on a Brompton

Greenspeed 40mm tyres on the Brompton - note the slick tread and very tight clearance around the rear tyre


Tyre shape seems to be worth looking at. Conventional orthodoxy has it that a narrow high-pressure tyre rolls better than a wider low-pressure example.We need only compare the performance of the original Moulton, with its narrow high- pressure tyres, and the frightful Raleigh RSW, equipped with wide low- pressure tyres. Narrow tyres do have advantages – low weight primarily, plus reduced frontal area (and thus wind resistance) – but do they really roll better than wide tyres?

Whatever theoretical disadvantages wide tyres might have, it seems that when we compare like with like (the Raleigh tyres were not only low-pressure, but had heavy, stiff sidewalls) they actually perform rather well. Regular readers may recall my slight disappointment with the narrow high-pressure Schwalbe Stelvio, launched in the 349mm size in early 2003.This 28mm wide tyre rolled slightly worse than the ‘cooking’ 37mm Brompton tyre and gave an inferior ride. In that case, might a wider tyre not roll even better?

Observing that wide tyres, even cheap ones, sometimes rolled better than narrow tyres, Ian Sims of Greenspeed decided to develop his own. Like the Primo and Brompton, the tyre has thin, supple sidewalls, but with a completely slick tread and a width of 40mm (against 37mm).Weight is 280g, against 200-250g for the 37mm tyres.

Greenspeed was kind enough to supply a pair of these new ‘Scorchers’, which I fitted to a Brompton – easy enough on the front, but a rather complex operation on the back, due to the tight clearances. After a period of running-in, the tyres proved surprisingly fast on my standard roll-down test, beating the Primo and Brompton tyres by a small but identifiable margin. Intriguingly, they were also more comfortable than the narrow tyres, under identical conditions.Why?

Some Experiments

Greenspeed Scorcher tyre contact patch

At the same pressure and carrying the same load, the contact patches are of near identical length, but the width varies broadly in proportion to the width of the tyre

It is widely assumed that – for a given tyre pressure, loading and wheel diameter – the area of the contact patch will always be the same, irrespective of tyre width: a long and thin patch on a narrow tyre, and a short fat one on a wide tyre. As the crucial dimension is generally assumed to be the patch circumference, it seemed to make sense to aim for a round contact patch, with the shortest possible circumference. Hence the move towards wider tyres.

Perhaps surprisingly, this turns out not to be case, or at least, not with tyres of conventional construction in the 16- inch size. Comparing the 28mm Schwalbe Stelvio, 37mm Primo Comet and 40mm Greenspeed Scorcher, with the same loading and tyre pressure, I found the length of the tyre contact patch to be a function of tyre diameter, irrespective of tyre width. But the width of the contact patch varied according to the tyre width.Thus, the most free rolling 349mm tyres have the greatest contact patch circumference and tyre/road contact area, and those with the highest rolling resistance have the shortest circumference and smallest contact area.

That the wide tyres should be more comfortable seems easy to explain.The extra width is bound to ‘average out’ the pits and bumps in the road, and we now know that the ‘point pressure’ is less with the broader tyres, presumably allowing the tyre to mould itself around obstacles, rather than deflecting. And as all the tyres share the same aspect (ie, height to width) ratio, the broader tyre is also taller, putting a greater expanse of rubber between the road and rim. All these factors might be expected to iron out bumps, but they don’t really explain the improved rolling performance.

Two possible answers spring to mind. Looking again at the illustration, it’s clear that the narrow tyre comes almost to a point at front and rear, suggesting a fairly acute degree of flexure at the front and rear of the contact patch as the tyre assumes the flat shape then springs back. On the wide tyre, the more gently rounded tyre shoulder suggests that the rubber is bending more easily to assume the flat contact patch. As one observer commented on seeing the illustration, the wider tyre displays a ‘cleaner’ ellipse, and this cleaner shape results in lower hysteresis. Presumably too, the improved shock absorption of the wide tyre reduces vibration, and thus rolling resistance.

Whatever the explanation, it looks as though a new generation of small tyres is on the way. Can we expect to see broad, slick designs on everyday bikes? Another widely held belief is that tread somehow improves grip. Obviously, this is true enough on soft or loose surfaces, but on tarmac, a slick tyre can be expected to grip better, roll better, and shrug off debris, reducing punctures, compared to a similar treaded tyre.Whether slicks gain widespread acceptance with the general public remains to be seen.

Unfortunately, most small-wheeled machines are designed for narrow tyres, and weight is important too, so it’s unlikely that wider (and taller) designs will be adopted, unless a considerable performance advantage can be demonstrated. For recumbents, on the other hand, the only downside seems to be the slightly increased frontal area.

Tubeless Tyres

Brompton tubeless tyres

Experimental tubeless tyre - note the cut inner tube protruding around the tyre. This surplus can be trimmed off when inflated

As we have seen with the Primo (see Folders 17 & 18), another solution is to make the tyre sidewalls more flexible, thus reducing the effort needed to overcome rolling resistance without compromising (probably improving) the shock absorbing characteristics of the tyre. But where do we go next? One long overlooked solution is to eliminate the inner-tube.There’s little point in fitting a tyre with paper-thin sidewalls backed up by a stiff, inflexible inner-tube. In practice, most good quality tubes flex quite well, but Greenspeed rig tests have found a reduction in rolling resistance of around 20% by eliminating the tube, so in theory, it’s well worth doing.

Greenspeed Scorcher Rolling Resistance

State-of-the-art. Running tubeless, the Greenspeed tyre rolls well - better than most 20-inch and some 26-inch tyres

If you want to experiment yourself, simply slice open an old inner-tube around the circumference, splay the tube out flat and fit it to a wheel, followed by a tyre. Inflate as usual (not easy) and trim off any excess tube. Provided the tyre is in good condition (you may need to add some sealing gunge), this home-made solution should work well enough.

Tubeless bicycle tyres are not a new idea, although most designs have required a special rim and/or tyre profile, none of which have caught on.The advantages include easy puncture repair (a soft pencil of rubber can be inserted into a hole from the outside, without disturbing the wheel or tyre), lower weight and lower rolling resistance. On the negative side, tubeless tyres are probably more prone to puncture, more difficult to inflate when off the rim, and they require some sort of sealing system Black Primo Comet around the spoke holes.

But what of the future? Greenspeed is currently fine- tuning the composition of the Scorcher tyre, a process that will no doubt yield another small performance gain.The first production examples should be available early in 2005.

Further information from Greenspeed


Are trains really ‘green’?

Professor Pivot“The current issue of Modern Railways magazine has an interesting article by Roger Ford on car, train and plane energy efficiency (June 2004, pp30-31). Ford’s analysis, ‘suggests, and I expect that this will generate some howls of protest, that a family of four going by car is about as environmentally friendly as you can get’. He has obviously forgotten the bicycle, but then he is talking about long distance journeys. Given that in theory ‘nothing can equal the steel wheel on steel rail for environmentally friendly transport’, what has gone wrong? A new state-of-the- art Virgin Super Voyager weighs 40% more per seat than an Intercity 125. Second, faster trains use a lot more energy – cutting the London-Edinburgh time by 30 minutes increases energy consumption by one half. Is half an hour worth it? Third, new trains are badly engineered.The new Pendolino intercity trains use 14 times as much energy for lighting as the trains they replace. How can this be?”

Dr Tim Leunig (daily commuter)

There’s no doubt that energy efficiency has been largely ignored by the railways since privatisation. Some of the last British Rail commuter trains were designed to use 20% less power than their (already efficient) predecessors, through lightweight construction, and by using AC motors, which can more easily provide ‘regenerative braking’ – putting electricity back into the supply when slowing down. For various reasons, this system was never made operational, and conventional brakes remain in use today. Meanwhile, the railway power supply is being completely revamped in the southeast to allow even more power-hungry German machines to enter service. Another odd modern practice is the tendency to put a diesel loco at either end of a train, because it’s cheaper than paying Network Rail to operate the points for the locomotive to run round to the front on branch lines! And although I have not seen the figures, I don’t doubt that the new Virgin Voyager is less fuel-efficient than the wonderful Intercity 125 trains (another British Rail achievement, incidentally). As with cars, extra weight through increased crash-worthiness, power-hungry air-conditioning, and greater acceleration are beginning to make inroads into the inherent efficiency of rail vehicles, although as we shall see, the figures stubbornly indicate that both modes are becoming more fuel-efficient.

When comparing road with rail, we must try not to lose sight of the bigger transport picture. Road transport has indeed become slightly more fuel-efficient in recent years: average vehicle consumption improving slightly, from 25.2mpg in 1993 to 28.2mpg in 2002, largely because of the introduction of fuel-efficient small diesels. Incidentally, these figures are drawn from total vehicle mileage and total fuel consumption, so they include buses and HGVs, which might sound unfair. On the other hand, only 5.8% of traffic is HGV, and the figures also include mopeds and motorcycles. Cars and light vans account for an astonishing 92% of total mileage.

In broad terms, the fuel consumption of road vehicles has hardly changed in 80 years because the increase in efficiency has been obscured by increased weight, bigger engines and so on. Only in North America have cars genuinely become more economic (from a very low base, of course).Throughout the developed world, vehicle efficiency is thought to be on the fall again – presumably because of the recent growth in gas-guzzling 4-wheel- drives, and as a side-effect of increasing congestion. And irrespective of the fuel efficiency of individual vehicles, the growth in UK vehicle mileage has caused an increase in overall consumption, from 39.5 million tonnes (petroleum equivalent) in 1993 to 41.5 million tonnes in 2002.


Virgin’s new Pendolino (left). Is it really less fuel-efficient than it’s predecessor (right)? PHOTO :

Meanwhile, the amount of fuel consumed by the railway industry (mainly diesel fuel and electricity) has fallen dramatically, from 0.93 million tonnes to 0.72 million tonnes (petroleum equivalent) in the same period, even though rail passenger/miles have increased by almost a quarter.There are many reasons why this might be so – scrapping of older thirstier freight locomotives, reduction of heavy coal traffic, and (hopefully) better vehicle utilisation, being the obvious ones.

If we look more closely at the 2002 figures and remove the fuel used to move freight (about 10% of the total), we find that passenger rail vehicles consumed some 648,000 tonnes of fuel and covered 443 million train/km.This could be expressed as 684 km/tonne or .581 km/litre, or even more conveniently, as 1.7 miles per gallon. As the average passenger loading in 2002 was 89.6, we can deduce a rough figure of 148 mpg/passenger for rail.

Yes, passenger rail vehicles are getting heavier, thirstier and faster, and they’re doing more miles, but because they’re faster, they’re attracting a lot more passengers, which helps to explain why the mpg/passenger figure is holding up so well.

Roger Ford suggests that a family of four can travel long distances more efficiently by road. In theory – provided their vehicle was a little more efficient than average – this would be possible, but as we all know, the problem with road transport is a passenger loading per car that hovers frustratingly around one. In other words, cars are usually carrying one person to work, or worse still, undertaking ‘positioning moves’; running driver-only to pick up passengers. In ‘cradle to grave’ terms, intensively-used rail vehicles do much better. As always, the answer is to make better use of public transport.

But, as Tim rightly observes, rail could do better and could make more effort to build on its many other environmental advantages.With the right technology, reduced track congestion and even better vehicle loadings, an improvement to 300mpg/passenger or more would be quite achievable.

Hope Mono Disc Brake

Which bicycle brake system?

Which brake system?

Professor Pivot“A to B 41 was interesting, as always, but I’m a bit confused by Martin Fillan’s comment (Letters, A to B 40) about grease leaking onto brake drum linings and his suggestion about using a roller brake.What exactly is a roller brake? Is it better than a drum? And is it easier to replace? I can’t find one in the Sturmey catalogue, so whose is it and how much extra will it weigh?”

John Burnett

Brakes are a fascinating subject, long neglected by this august publication. Broadly speaking, the problem faced by engineers since the invention of the wheel has been to produce a simple, tough device capable of transposing a large but weak hand or foot movement into a small but powerful force to push against a rotating body and slow its progress.The brake then needs to dissipate the considerable amount of heat generated – something that few bicycle brakes are very good at.The following devices are all available today to retard the progress of bicycles… Some more successfully than others:

1. Caliper brake

Alhongo Dual Pivot calliper brakes

The Alhonga dual-pivot caliper. More complex than some, but the principle is the same - two arms, one attached to the brake cable and the other to the cable sleeve.The wheel rim is clamped between the two brake shoes

Long outmoded on motor vehicles, a caliper brake (sometimes called a side-pull) consists of a pair of curved arms or calipers pivoting somewhere beneath the headset bearings, with ‘blocks’ of friction material at their lower extremities. By the action of a pull rod, push bar, or more usually a flexible cable these days, the friction blocks are moved towards each other, squeezing the two outer faces of the wheel rim in the process.

The caliper is light and cheap, because the rotating element is already in place, but being completely exposed to the elements, it is badly effected by rain, grease, oil and grit. Different calipers and brake blocks are affected in different ways, but the most important element is the frictional co-efficient of the wheel rim material. Chromed steel lasts for ever, and works very well when dry, but loses most of its stopping ability in the wet. Aluminium is less effective in the dry, but relatively good in the wet, making it a safer material overall. Unfortunately, aluminium rims can wear away quite fast, especially on small-wheeled bikes.

The quality of the brake ‘feel’ depends largely on the friction material and the construction of the caliper. Poor calipers bend and distort when the brake is applied, giving a rubbery feel at the lever and/or judder or squeal.

Calipers are notoriously difficult to centre correctly, which can leave one brake block rubbing against the rim, and a wobbly rim will cause one or both blocks to rub intermittently. Generally, the rim disposes of heat quite successfully, but heat build-up can be a problem on long descents, particularly for heavily-laden or small-wheeled bikes. Excessive heat in the rim can cause tube failure and a catastrophic blow-out.

2. Band brake

Band Brake

Band brake principles. Friction between the tethered band and rotating drum tends to slow the drum’s progress

A long-outmoded Edwardian technology, the band brake is nevertheless worth mentioning because these devices do still turn up in the rear wheels of Chinese bicycles once in a while. A band brake utilises a flexible band of friction material, firmly fastened at one end and wrapped loosely round a rotating steel drum.When the band is pulled tight by a lever, it wraps tightly around the drum, slowing its progress.This tendency for the brake to apply itself without undue effort from the rider is known as ‘self servo’.The bad news is that the effect usually disappears in reverse, so a band brake will not stop you running backwards down a hill…

Band brakes are cheap, low-tech devices, but the negative aspects go on forever.With the drum inside the band, there’s nowhere for heat to go, and being difficult to protect from the elements, water can slosh straight in, resulting in a near total loss of braking effort. Brake bind, shriek and squeal can be a problem too, especially after a good soaking.

3. Drum or hub brakes

More correctly an ‘internal expanding shoe’ brake.This was the standard motorcar and motorcycle brake for most of this century until superseded by cheap reliable discs, and remains a favourite on hard-used and/or heavy bicycles. The general layout is two curved aluminium blocks or ‘shoes’ faced with friction material, both pivoting at the same point, and pushed outward at the other end by a cam of some kind, to make contact with a metal (usually steel) drum. Like the band brake, drums exhibit a self servo action – the leading shoe tending to be drawn harder into contact with the drum, whilst the trailing shoe tends to be pushed away, and visa versa in reverse. A variant common on older motorcycles and cars was known as ‘twin leading shoe’ – much more effective going forwards, but virtually useless in reverse (see band brake).

Drums can be heavy, although much of the weight penalty is negated where the brake is combined with a hub gear, and modern manufacturing techniques can reduce the weight significantly.The shoes are largely immune from contamination, but internal and external sealing arrangements can be a bit crude. Sealing problems between the gear and brake components of a hub can result in grease contamination, which can ruin the shoes, and poor external sealing can allow water in, although this generally requires total immersion.Wet friction shoes lose most of their effect, and as they start to dry, a violent self-servo action can result in brake snatch and squeal.

Hub brakes are not good at dissipating heat, but they make do by transferring it into the substantial mass of the hub in the short term, where it can safely radiate away when the descent is over. If a good quality drum does overheat, it should gently ‘fade’, or become less effective until cooled. Adjustment is rarely required once the shoes have bedded in, and the progressive action and ‘feel’ of a hub brake beats most other types.

Drum brake exploded diagram. Left to right: back plate (fixed pivot above, moving pivot and lever below), brake shoes faced with friction material and steel drum.This is the Sturmey Archer ‘BR’ of 1932 – it survived, effectively unchanged, for 70 years

4. Back-pedal or coaster


Simplified coaster brake diagram. Left to right: brake arm and hub dust cap (fixed to bicycle frame), brake shoe segments, brake actuator, rotating hub shell. Back pedalling drives the actuator into the shoes, forcing them against the inside of the hub shell

Rare in Britain, but common elsewhere, the coaster is usually combined with a hub gear. Pedal backwards and a metal cone slides through the hub, pushing metal brake segments against the rotating hub shell. Operation can be a bit insensitive, with a lack of feel, although hubs vary. Being grease filled, a coaster brake is more or less immune from contamination. It also has no exposed levers and cables to go wrong, requires no adjustment, and lasts more or less for ever. Like the drum brake, heat dissipation relies on warming the hub on a descent, then allowing the heat to escape. In extremis, localised heating from the metal-on-metal contact can boil away the grease or even weld parts together, although I have never heard of this on a bicycle.

5. Cantilever and V-brake

Similar rim-squeezing action to the caliper, but the force is provided by two vertical arms, pivoting at the bottom and brought together by a cable pulling the top of the arms together.The only real difference between the cantilever and V- is in the way the cable pulls the arms. In both types, the brake blocks are mounted some way down the arms in order to gain a degree of mechanical advantage.

V-brakes have become the braking system of choice because they’re light and crudely effective in operation. Problems are as for the caliper brake – water and oil contamination, rim wear, difficulty with centring and heat build-up on long descents. Cheap V-brakes can be unpredictably fierce in operation, so many feature pressure limiting devices of various kinds (usually fitted in the cable) to prevent the front wheel from locking up. Other problems include squeal on cheaper units, judder, and frame or fork distortion when the brake is applied. On the positive side,V-brakes are simple to maintain.

6. Roller brake

Nexus Roller Brake

Diagrammatic view of Nexus roller brake. Force from the brake lever is applied through the operating lever (top) to the central cam.The cam pushes the rollers out against brake shoe segments, forcing them into contact with a rotating drum, integral (in some designs) with a cooling disc

Like so much else in the bicycle world, the roller brake is a Shimano invention, or re-invention. Combining elements of the drum, coaster and disc brake, the friction effect comes from steel rollers which are forced outward by a cam, pushing metal shoes against a rotating steel drum. Heat build-up can be a problem, but most designs incorporate a cooling disc, just like a ‘proper’ disc brake. Brake feel can be unpleasantly ‘woolly’ and vague compared to other types, and the shoes can make nasty metal-on-metal noises unless well greased.That’s also the good news, because water and oil won’t make much headway into a grease-packed unit.

7. Disc brake

Hope Mono Disc Brake

The Hope Mono Mini bicycle disc brake. High pressure fluid pushes a piston against a friction pad, forcing it into the disc. Most brake assemblies contain two (or more) opposing pistons, but here the caliper ‘floats’, allowing a fixed friction pad to contact the back of the disc

This brake generally utilises a pair of friction pads, which are forced against opposite sides of a steel disc. Heat does build up in the disc (they can glow cherry red on a hard-driven racing car, or after stopping a high speed train), but the disc is well exposed, so heat is rapidly dissipated and fade is rare. Disc brakes have become the preferred means of de-acceleration on just about every wheeled vehicle, from aircraft to trains, cars and motorcycles. Progress in the bicycle world has been limited, mainly because the disc brake lends itself to hydraulic operation, which can add weight and complication. Pads may also bind slightly in the ‘off’ position, which can be frustrating on a fractional horsepower vehicle. Early bicycle discs were heavy, ineffective in the wet and noisy, but these problems have been largely eliminated by reducing the disc to a delicate tracery of struts.

In answer to the question, it isn’t easy to convert a bicycle to roller brakes because the Shimano Nexus system will fit only a Shimano hub. But if you have a Sturmey Archer or SRAM-equipped bike, there’s really no need – an upgrade to hub brakes is generally quite easy (not on smaller bikes like the Brompton, unfortunately). In practice, few roller, drum, coaster or disc brakes work badly enough to be worth changing.


Electric bikes – How Many Watts?

Professor Pivot“There seems to be some confusion with regard to the legal power of electric bikes today. I am hearing that the maximum power permitted is now 250w and not 200w. Is this correct?”

Alan D. Shaw

Once upon a time, the law was the law and we all knew where we stood, but today things are a bit more complicated, thanks in this case to Europe. Basically, we now have both British law, allowing 200-watt motors for bicycles (and 250 watts for tandems and trikes), and European law (250 watts throughout) on the statute books at the same time, resulting in confusion as to which should take precedence.The same applies to the legality of so-called ‘e-bikes’ that allow the motor to be operated without pedalling.These are legal under British law, but illegal under European law. Neither the Department for Transport or Department for Trade & Industry seem willing or able to answer these questions, even suggesting rather unhelpfully that the matter should be settled in the courts!

To make things even more complicated, electric motors are rated on a ‘continuous’ basis, so that 200-250 watts rating bears little relation to the actual power at the wheel. A continuous rating is, crudely, the power that a motor (or indeed, a human) can produce all day. It can be precisely defined, but I have yet to find a British, European or ISO definition, although one, or perhaps several definitions, must exist.

Study the graphs in A to B road tests, and you will see that most electric bicycles produce in excess of 400 watts, and peak outputs of 1,000 watts or more are not uncommon.These powerful motors were designed for the US market, where some States allow motors of ‘up to 1,000 watts’ and a top assisted speed of 20mph.With the collapse of the US electric bike market, they’re now being sold elsewhere, and it’s a measure of how confused the situation has become that these machines are being openly traded in the UK. Personally, I do not feel that a normal bicycle has the lights, brakes and other equipment to deal with these higher speeds, and with long-range fuel cells not so far off, I would prefer to see a new hybrid bicycle/moped vehicle class, allowing greater speed, but with compulsory insurance and so forth. [As in Switzerland and elsewhere, see News, Eds].

For now, provided your electric bicycle does not exceed 15mph under power (24km/h in Europe!), or burn rubber at the lights, you can buy and use just about anything you like. At some point, an innocent bicyclist will be stopped on a US or European machine and hauled before the courts. Barring appeals, this will give a clear precedent for cycle trade, police and public to follow.

One wonders how the Department for Transport sees its role? It is quite absurd that the official transport bodies are unable to settle this matter, but as we know all too well, bicycles are hardly a top transport priority.

Methanol Fuelled Bikes?

Direct Methanol Fuel Cell (DMFC)

Miniaturisation of the DMFC has been extremely rapid, from the 2.5kw lab plant ...

I ask these questions because it seems to me that a sensible use for bio-methanol would be as fuel for extremely low-mpg transport, such as 28cc engine-powered bicycles. Surely this would be a truly sustainable form of low cost transport and without the battery range limitations of electric bikes (speaking as a Heinzmann rider).”

Simon Rayson
Dorchester, Dorset

There has been no suggestion to my knowledge that recent changes in electric bicycle legislation will bring harmonisation with Europe where small internal combustion- powered bicycles are concerned, but legal matters of this kind are so labyrinthine that change of this kind is quite possible.

Direct Methanol Fuel Cell (DMFC)

.. and a one watt mobile phone supply.

Speaking theoretically – as we must – methanol is, indeed, an excellent fuel. Not quite as ‘green’ as it might sound, because a considerable amount of fossil fuel is used in its harvesting and distillation from grain or plant waste, but a largely renewable fuel all the same. Current production amounts to some 30 million tons per annum – a tiny spec in motorcar terms, but enough, if we chose, to provide 300 million individuals with 4,000 miles per annum of renewable transportation indefinitely… assuming they could all be persuaded to ride internal combustion-assisted bicycles, of course!

Direct Methanol Fuel Cell (DMFC)

An electric bike would require around 100 watts from today’s 20 watt package PHOTOS: Aalborg University, Denmark and Toshiba

A better option, eliminating the noise, smell and inconvenience of the engine would be the DMFC, or Direct Methanol Fuel Cell. Smart money is already pouring into DMFC research for low-power devices, and the first (rather bulky) lap-top power cell, with maximum output of 20 watts, is about to go on sale, with smaller, more powerful cells just around the corner.

Several problems exist – methanol is poisonous, and unlike hydrogen cells, the DMFC produces as much carbon dioxide as an internal combustion engine. However, there is a promise of greater efficiency, and of course the raw fuel could be produced sustainably, making the operation potentially carbon-neutral.

A DMFC cell would be lighter than any current battery, give silent power, unprecedented range and instant, relatively safe, liquid refuelling. Predictions are dangerous things, but I would be very surprised if the Japanese do not launch a DMFC fuel-cell bicycle on their domestic market within two years.

Recharging an Electric Bike on the Move

“Many years ago I cycled to Nice from home in the Birmingham area. Due to a road accident, I am not able to do such a silly thing again, though I would like to.Would it be possible and legal to tow a small generator behind an electric bike, leaving it running whilst riding?”

Anthony Skidmore
Halesowen,West Midlands

…we’re talking hybrid petrol/electric/human power… [but] the roar of a petrol engine would be quite out of keeping…

This question should really be broken into two parts – legal and technical. Under UK law (the Electrically Assisted Pedal Cycles Regulations 1983) an electric bicycle is defined as a machine where ‘The motor assistance must be provided by an electric motor.’ Which sounds fine, except that, ‘Propulsion by an internal combustion engine is not permitted’. Now, of course, this wording is intended to outlaw direct drive from an internal combustion engine.What you propose is, effectively, a mobile battery charger, topping up a conventional battery.This would be another interesting matter for the courts to decide. One assumes the legal debate would centre on this concept of ‘propulsion’, thus I suspect the trailer would make no difference – if this form of charging was adjudged to be within the law, it would also be legal carried on the bicycle.

Technically, there are no great complications:We’re talking hybrid petrol/electric/human power here. Such a vehicle could run on human power in flat urban areas, human and electric power on hills, and top-up the battery from the internal combustion power source as and when required. Range would be quite considerable, because fuel consumption would be nil on many journeys. But on the open road, the roar of a petrol engine would be quite out of keeping, so it’s hard to see the advantages over a small motorcycle for longer journeys.

Once again, one suspects that fuel cells may soon render such dramatic steps unnecessary. Provided an internal combustion engine is not involved, the regulations seem to allow any form of electrical power generation, from solar to nuclear. Neither of these extremes are very practical, but a small fuel cell probably would be.

Crank Length

Bicycle crank length - road speed vs rider effort

Heart rate versus effective road speed for different crank lengths. Cadence was adjusted to suit each crank length: ie at 15mph, 175mm were spun at at 79rpm, 125mm and 100mm at 89rpm. Peak power of 780 watts was recorded with 175mm cranks, but both 155mm and 125mm managed 770 watts, and 110mm, 705 watts.

I too used to belief firmly that 150mm cranks were only for kids, or adults with very short legs. In fact, after much experimentation, I came to the conclusion that extra-long 185mm cranks suited me best, and that the formula put forward by Kirby Palm [see] was correct.

Then two of my customers, Rob Hague, and Mark Mueller started experimenting with short cranks. Rob found an improvement in his race times by going from 170mm to 150mm, even though he has long legs! Mark found that after training for a month with 110mm cranks, he was back to the same performance he was getting with 170mm cranks. My son Paul tried 100mm cranks on his trike and found so much improvement over normal length cranks, that there was no way he was ever going back to normal ones.

Then a customer, Irene, who had hired a trike with 165mm cranks for three months, came to take delivery of her new trike. I measured her legs against the formula, and discovered that according to Palm, she needed 150mm, so I fitted a pair of 150s for her.

After a few days I got rave emails from her saying how smooth the trike felt, and how much better than the demo it was.This seemed rather puzzling to me, as the hire trike was a little lighter.Then she told me that with the new trike, she did not get leg cramps any more, which had been a problem for her with the old trike.Then the penny dropped. It was the cranks which had made the difference.

Next, Paul built a trike to race at the 24hr Pedal Prix at Wonthaggi.Without time to train all the team on 100mm cranks it was decided to use 150mm cranks. Despite losing a few laps following the collapse of an experiment wheel which had been fitted by mistake, they won the race by three laps! Paul then did some tests on our new Computrainer machine and came up with the following results. In all it seems than the higher cadence gained by using short cranks, has no negative effect on power output, and gives an improved leg function, with reduced stress.

On a recumbent, short cranks also help to get a 69rpm, 155mm more efficient aerodynamic were spun at fairing over the rider’s legs.

Solar Rickshaw

Professor Pivot“I was interested to see the solar electric recumbent in A to B 38. It occurred to me that if you could power a recumbent tricycle with two solar panels, one could probably power an electric rickshaw with three or four such panels?”

Felicity Wright

With the oil supply/demand crisis on the way, this may well become a fascinating question. At present, rickshaws are being replaced by internal combustion machines of various kinds, and that will have to change. But is a solar-powered rickshaw viable? We have all the data we need from past A to B magazines – issue 37, on solar power, and issue 39 featuring an electric-assist rickshaw.

As a rule of thumb, the sun supplies around one kilowatt of energy per square metre, which just happens to be the roof area of a typical rickshaw. Unfortunately, another rather more depressing rule of thumb suggests that commercial panel efficiency is somewhere in the region of 10%, reducing peak power to 100 watts. In practice, our Unisolar flexible panels did even worse with British solar radiation, achieving a peak output equivalent to 65 watts per square metre, or 390watt/hours (or Wh) per day.


Cycles Maximus solar- powered pedicab

The roof of the Cycles Maximus rickshaw featured in A to B 39, measures approximately 100cm by 110cm, giving a practical collection area of about 1.1 square metres, or 7.2 times the size of the panels in our previous experiment.Thus, provided our rickshaw was always in the open – something hard to arrange in a cityscape, for example – we could reasonably expect to see peak power of 72 watts, and daily output of 430Wh.

…future developments will depend on the price and availability of oil…

One immediate problem – in UK climes at least – is that our roof-mounted panels were angled towards the south at 20 degrees because at our latitude, a horizontal panel would never be fully effective. Another problem is cloud cover – even in June, the average panel output was cut almost by half, to 36.6watt/hours per day.Taking these factors into account, it’s unlikely that our rickshaw would generate more than 200Wh per day in fine weather, and virtually nothing on an overcast winters day.

The powerful Lynch motor on the Cycles Maximus rickshaw had a battery of 828Wh, giving about two hours power-assistance, well loaded in hilly country. So again, rather depressingly, we are brought to the conclusion that four days, or 40 hours, constant charging in mid-summer would give only two hour’s use…With the less power- hungry Heinzmann motor, the equations become a little more favourable – somewhere in the region of ten hours charging per motor/hour.

Of course, the power consumed depends on the terrain. In a relatively flat city, an hour’s power-assistance per day might well be sufficient, but in this case, we are really looking at a human-powered vehicle with occasional motor assistance, rather than the other way around!

Naturally, once we get close to the equator, we have a great deal more sunlight to play with.With more powerful solar radiation and no cloud cover, we could reasonably expect to generate peak power of up to 100 watts, or perhaps 600Wh per day, giving some three hours motor-assistance.There are other incidental advantages – the solar panels would help to keep the occupants cool by absorbing (and reflecting) solar radiation, and although a battery would be required to deal with the peaks and troughs of production and use, it could be quite small, itself reducing the weight of the machine.

Several experimental solar rickshaws have been trialled in India, one impractical monster weighing 300kg, or nearly three times the weight of the British-designed Maximus! Some have been built with access to prodigious overseas aid, such as the rickshaw claiming to carry panels with an output of 850 watts – clearly either satellite- grade technology, or a surface area of a great deal more than one square metre…There are also a number of ‘exhibition’ solar machines around, but these are a long way from earning a living day-to-day. As with so many other things, future developments will depend on the price and availability of oil. State of the art panels, with an efficiency of 30% or more, would make a lightweight solar vehicle perfectly viable, but only in a sunny area!

Solar-Powered Bicycle Lights?

Professor Pivot“I think A to B should do a review of existing options for solar powered bicycle lights. I found one at and the Alternative Energy Centre in Wales [] sells a solar-powered 6 LED red emergency light which is sold with a bracket to attach to a bicycle.”

Jonathan Pattison

cateye-el300Clearly, in this case, we’re looking at a very much smaller power requirement, from a correspondingly small solar panel. First let’s look a bit more closely at the power demands of a bicycle front light. A typical conventional or halogen filament bulb for a bicycle front lamp consumes two watts.That’s manageable, but realistically, few people would want to carry around a solar panel large enough to provide that sort of power, so we need slightly cleverer technology to reduce the energy demand. Fortunately, as regular readers will be aware, Light Emitting Diodes are advancing very rapidly and can now provide an adequate light output with much lower power consumption.

The only front LED lamp currently available to German (and by default, British) standards is the Cateye EL300G, although the situation is changing rapidly, so this may already be out of date.When we tested its predecessor, the EL300 (A to B 33), we found power consumption of 0.58 – 0.89 watts, depending on battery voltage. Consumption is claimed to have fallen since, so we’ll generously base our calculations on the lower figure.

“…the solid-state LED solar torch is perfectly viable and will no doubt be available soon…”

To provide 0.58 watts for, let us say, two hours each evening would require 0.58 watts x 2 hours = 1.16 watt/hours. If we reckon on twelve hours charging time each day in summer, the charge rate would thus need to be 1.16 watt/hours divided by 12 = 0.097 watts.With peak power of 65 watts per square metre of solar panel (see above), each square centimetre will provide 0.0065 watts, so a panel of 15 square centimetres would more than cover our requirement of 0.097 watts.

This raw data suggests that a panel measuring just four square centimetres would provide two hours of light output per day – if only it were thus! The problem, of course, is that our panels provide a peak output of 65 watts per square metre. In reality, the mean figure over the course of a bright sunny day is about half this, increasing the panel requirement to 30 square centimetres.

The Cateye EL300 is one of the few cycle lights efficient enough to run from solar power

Unfortunately, even in June, the sun is often partially or completely overcast, and during my experiments last summer I recorded an even lower mean figure over a 16-day period of 20 watts per square metre, giving a panel size of nearly 50 square centimetres to work our low-consumption front light.

Nevertheless, the figures suggest that a square panel measuring just 7cms across would provide up to two hours light each night throughout a typical English summer, and a panel of double that size would be physically possible, making a light of this kind practical for most of the spring and autumn too. One proviso – the panel must be angled towards direct sunlight throughout most of the day, something that’s hard to arrange in practice, as we discovered during our experiments.

But what of the Energy Capture cycle light Jonathan found on the internet? Armed with our raw data, we can tell quite a lot:The solar panel measures a reasonable 88 square centimetres, and is rather optimistically claimed to have a 1.1 watt output.That may be attainable in a test lab, but working on our 20 watts/square metre average figure, I would suggest 0.17 watts in practice.That would certainly power an LED front light, but unfortunately the Solar Light is fitted with a filament bulb drawing four times as much power as the Cateye EL300!

For occasional use (say 30 minutes per day) a lamp of this kind might suffice, but for nearly £50 you really would be better off buying a conventional light with rechargeable batteries, and recharging a spare set each day by mounting a small solar panel on a south- facing window sill. However, the solid-state LED solar torch is perfectly viable and will no doubt be available soon. Regrettably, it seems Energy Capture has ceased trading, demonstrating that this example might have been an idea ahead of its time.

Electric Bike Range Claims

“I would like to query the basis of the electric bike range statistics you quote. I realise that you are ‘A to B’, but for many people, their use of such a bike for leisure purposes would be ‘A to A’; that is to say starting and finishing at the same location. If one assumes a hilly terrain, and ignores any flat areas, at the completion of a journey the uphill sections will have equalled the downhill. If therefore, one only used the electric-assist on the uphill sections, and turned the motor off on the downhill sections to freewheel, would the ranges you quote notionally be doubled?”

Michael Bartlett, Shoreham-by-Sea

Unfortunately, life is rarely that simple. For one thing, where a bike has the capability, the A to B testers usually opt for speed over range, which is why tests always quote the average speed.To keep speed up, the motor is often used for long periods on the flat, as well as climbing hills. Secondly, you never get back all the energy expended climbing a hill going down the other side! This is partly because motors tend to be inefficient when climbing at low speed, but primarily because most of the energy is dissipated in fighting the wind.This effect is barely noticeable at low speed, but descending a hill at 30mph will scrub off much of the kinetic energy stored on the climb…This is why ‘regenerative’ braking (recharging a battery or other storage device on a descent) is hardly worthwhile on a bicycle.

Think of the hill as a battery:Whether you ride a conventional or power-assisted bicycle, you store kinetic energy on the way up, and discharge it on the way down. A heavy freight train crossing the Swiss Alps will store a great deal of energy, which can usefully be returned to the electricity grid going down the other side, but a bicycle stores a tiny amount, and on such a small lightweight vehicle, wind resistance has a comparatively large effect. On a switchback road, the stored energy may enable you to get half way up the other side, so it can be useful, but in most cases, mechanical or electrical storage devices would be of little help. Now, where were we?

The A to B 17.6 mile test route starts and finishes at about the same elevation, but climbs and falls almost continuously in between, so the motor tends to be used for a high proportion of the ride.The figures published in the magazine are always lower than the maximum achievable, although they seem to equate fairly well to the sort of range a typical rider might expect, making shorter trips in heavier stop-start traffic.

In terms of maximum range, in level country, you might expect to exceed the A to B figures by 40% or more, even when using the motor much of the time. For example, the Powabyke or Ezee Forza can manage 50 miles relatively easily on the flat.

Most bikes complete the test course in 75-80 minutes, but the Forza did the run in 62 minutes – exactly 17mph. How can a bicycle limited to 15mph maintain such a high speed? The reason is that the more powerful machines cruise at close to 15mph and rarely fall below 10mph on hills.Add a few downhill bursts, and the average speed can exceed the maximum assisted speed…This is why electric bikes work so well in hilly areas.

Brompton 5-speed Upgrade


Old mechanism (front) and new (behind). They are effectively identical, except for the longer axle on the new item. Note the old and new locknuts.

Once upon a time the finest hub gears in the world were churned out from a factory in Nottingham, many finding homes in the Raleigh bicycles made just across the Triumph Road.The sad demise of Sturmey Archer in the autumn of 2000 is a subject we’ve returned to many times. For A to B newbies, this healthy company was brought down by a combination of corporate greed and downright stupidity, resulting in Sturmey being handed to a bunch of asset-strippers, who bled the company dry in a few weeks and chucked out a life-less corpse, the remnants being shipped to Taiwan by Sunrace.

One of the biggest potential losers was Brompton, a company with a full order-book for a folding bike designed specifically for the Sturmey 3 and 5-speed hub gears. In the final weeks, some hastily arranged heavy transport brought a reasonable stock of hubs down to London, giving a breathing space for the bikes to be adapted to accept the German/US SRAM hubs.

The excellent 3-speed SRAM worked well, but the 5-speed was much too wide for the bike. Brompton was forced to become a 3-speed manufacturer for a while, until its own 2-speed derailleur entered production in May 2002 and was fitted alongside the SRAM to produce the elegant 2×3-speed L6 and T6 Bromptons.

…the Sturmey 5-speed wasn’t the most reliable of hub gears.. thanks primarily to poor adjustment…

Meanwhile, Sturmey 5-speed spares were running low – a potential problem for the estimated 20,000 5-speed Bromptons, and countless Moultons, Bickertons and Micros in regular use worldwide. Most bikes could be adapted to take a SRAM 5- speed hub, but the Brompton could not. Conversion to 2×3 spec is not viable either, because the parts alone (3- speed SRAM and wheel, rear frame, cables, changer and 2-speed derailleur kit) cost more than the bike is worth.

In early 2001, Sunrace put the Sturmey 3-speed back into production in Taiwan, and things began to look a little brighter. In late 2002 the 5-speed hub returned too, and although very different on the outside, it was almost identical internally to the ‘ball- locking’ hubs fitted to the Brompton in the last year or so of UK production. Not quite identical, but we’ll come to that.

Can this hub – produced three years later, and 10,000 miles from Nottingham – really be fitted to an elderly 5-speed Brompton? It turns out that it can, and for a reasonable cost, new Sunrace-Sturmey ‘internals’ could revitalise your old folder.

It has to be said that the Sturmey 5-speed was not the most reliable of hubs – a fair number of machines losing one or more gears, thanks primarily to poor adjustment, made worse by a rather woolly gear-shift action. Broken axles were relatively common too.

The ‘ball locking’ mechanism was introduced to hold the hub more positively in gear. Ironically, it came into production just a year or so before the company was dragged under, but the tooling and expertise were kept together, and when production restarted in Taiwan, it was natural that the latest system would be used.

Buying the Bits


All the grime will have to be removed. Carefully check the parts for wear - this sprocket is marginal. The thin lock washer and locknut will be re-used

We’re going to describe the process of swapping old-style Sturmey hub internals for Sunrace-Sturmey parts. For the mechanically-minded, it’s all quite straightforward, and the only specialist tool required is a 16mm cone adjusting spanner. If you’re unsure about mechanical bits, contact Sturmey and Brompton specialist, Bicycle Workshop in west London (tel: 0207 229 4850).

Step one is to buy the necessary bits, and once again, unless your local cycle shop is particularly keen and knowledgeable, we recommend Bicycle Workshop.The hub gear internals will cost £70, and the indicator chain £4.50, plus postage, together with a labour cost of about £15 if the shop does the work for you.

Other parts may be needed though. Carefully inspect the inner and outer gear cables, gear shifter and Brompton cable guide. If the action of any parts is sticky or rough, replace ‘em! The same applies to the sprocket and chain. If they were fitted more than a thousand miles ago (always replace these parts together), you will also need a new 13 or 14-tooth sprocket and chain from a Brompton dealer.The new nickel-plated chain works perfectly well, but it must be 1/8″ and not the 3/32″ derailleur-type chain, fitted to the new 6-speed bikes.

Getting Started


Removing the right-hand bearing cone. The next step is to unscrew the internals from the other end

You’ll need to remove the rear wheel, and as we delight in saying, if that proves tricky, this might be the stage to hand over to an expert.With the wheel off, thoroughly wire brush both the sprocket assembly and associated (right-hand) cone and locknuts. Repeat this operation for the opposite, left-hand bits and pieces too. It’s essential that grit doesn’t find its way into the hub, and you’ll be re-using these locknuts.


Screw the new internal assembly into the shell and tighten. Note the old locknut and lock washer.

When the hub is reasonably clean, mount the wheel in a vice and remove the sprocket assembly, right-hand locknut and thin lock washer, putting the parts carefully aside in the order in which they were removed. Remount the wheel in the vice and remove the left-hand locknut and bearing cone.

You’re now ready to spin the internals out of the hub shell. If you don’t have the special ‘C’ spanner, and a hammer and punch fails to shift the ring, clamp the assembly in a vice and spin the wheel anti-clockwise.


Final adjustment to the left-hand bearing using the cone spanner

With the old internals removed, carefully inspect the inside of the shell for foreign bodies, rust or signs of water damage. If all’s well the new unit can be screwed straight in and tightened with a hammer and punch. If not, the shell will have to be carefully degreased, cleaned and regreased – tedious, but very necessary.

You will now be in a position to screw on the left-hand bearing cone and the old thinner locknut. On the right (sprocket) side, carefully hold the bearing cone and spin off the right-hand locknut without disturbing the cone. If the adjustment is upset, the cone should be hand-tightened and backed off half a turn before being locked in place. Finally, put back the dust shield, spacer and sprocket, checking sure that the circlip beds down correctly in its slot.

…the most common cause of failure is misassembly following roadside repairs…

With the axle back in the vice, it’s time to adjust the bearings.The sprocket is an important part of the adjustment process, because the left cone should be tightened until (quoting Sturmey) there is ‘minimum’ free play at the wheel rim, but noticeable play at the sprocket, with no tightness or roughness. Lateral rocking of the sprocket looks alarming, but it’s quite normal.When satisfied with the adjustment, hold the cone nut steady and tighten the locknut. Don’t spend hours fiddling – it will need checking after a few weeks riding anyway.

The hub should now look exactly the same as the old one, but with a longer axle stub on the left-hand side.Again, this looks alarming, but it has no harmful effect. If you really want to, you can angle-grind this (very hard) axle back, but we wouldn’t recommend it.

Sturmey produced a special short axle hub for the Brompton, but the difference is merely cosmetic.The only other Brompton-specific parts are the thinner locknuts and right-hand washer that just enable the 5-speed hub to squeeze into the Brompton’s narrow frame drop-out.


Assembly details. Note that the gear selector guide is followed by the lock washer and nut. The gear selector guide must point directly at the cable guide roller on the rear frame

It’s now time to refit the wheel to the bike, taking great care that the stepped anti- rotation washers sit comfortably in the drop-outs, and that the gear selector guide support (the washer with a funny bent bit sticking out), is followed by the lock washer and hub nut.The most common cause of 5-speed failures is misassembly following a roadside puncture repair, so do take care to get it right, or you could damage £75 worth of shiny new bits.The tab on the guide support is easily bent – it should stand at 90 degrees to the washer face – and the tab must point towards the cable guide roller assembly. If any of these parts are loose or poorly aligned, the gear cable will stick or wiggle around, causing missed gears. Finally, fit a new ‘blue’ ball- locking gear indicator rod, and refit the tensioner and chain.

Final adjustment

Adjustment is critical and needs to be carried out in second gear.You can either use a mirror to observe the blue band on the indicator rod, or very carefully turn the bike upside down, taking care not to move the gear shift.What you must never do is adjust the cable with the bike partially folded, which will give a completely false reading. In 2nd gear, the blue band should be entirely visible, but only just, if that makes any sense. As with the hub bearings, it’s a good idea to reset the adjuster after a few rides.

Upgrades with new parts can be unpredictable, but our conversion worked perfectly, and yours should too.With careful adjustment and maintenance, the revamped Sturmey hub should last for many years and revitalise a tired Brompton.

Bicycle Workshop tel 0207 229 4850 . Brompton tel 0208 323 8484

Schwalbe Marathon Plus Tyres

Avoiding Punctures

Professor Pivot“Can you tell us how to deal with or prevent punctures? I don’t mean sticking a patch on, but what is the best strategy? I believe that the hassle of punctures really does put people off using a bike regularly for important journeys. In Peterborough the cycleways are edged with thorn bushes like pyracantha.”

James Haugh

Astonishingly, some local authorities continue to plant thorn bushes next to cycle paths and have no policy for sweeping debris from paths – something normally done by motor traffic on the road. Not surprisingly, cycle path usage continues to languish in these places.

James is quite right, of course.The (often irrational) fear of punctures does put many people off cycling, but these days the threat can be largely eliminated.With the right equipment and some basic preparation, roadside disasters should become rare events.


Choose good quality tyres and tubes, and replace them before they’re life- expired. Cheap and/or worn tyres are much more likely to puncture.The range of tyre brands is bewildering, and some are only suited to particular uses, so I can only give a few pointers: I’m not convinced that kevlar-banded tyres really do give much protection, but many swear otherwise.They certainly increase rolling resistance, as do puncture-proof liners, which can also cause tube failure if incorrectly fitted. An interesting, but heavier, option (as yet untested by A to B) is Schwalbe’s Marathon Plus, constructed with a special spongy lining to ‘bounce’ objects back out of the tyre.This is now available in the 47- 406mm (20-inch) size, with 18″ and 16″ on the way.

Surprisingly enough, slick and lightly-treaded tyres often work quite well, presumably because they are less likely to trap and hold road debris. But paradoxically, you shouldn’t write off treaded tyres either, because some are very puncture resistant. As a second line of defence, thick puncture-resistant inner tubes seem to help, at the expense of increased rolling resistance.

For a given bike/tyre size, ask regular users what they’d recommend. Much of the advice will be pure hearsay, but you should begin to get a picture of what works and, more importantly, the brands to avoid.


Keep the tyres correctly inflated. Under-inflated tyres are good at picking up debris and can also suffer ‘pinch punctures’ on bumps and kerbs. It seems reasonable to assume that over-inflated tyres might be vulnerable too, although I have no specific evidence for this. A simple rule of thumb is to inspect the tyres with the bike loaded.They should bulge slightly around the road contact patch: No bulge, or too pronounced a bulge, and you’re asking for trouble.

Secondly, it might seem obvious, but if you can avoid thorns, roadside debris or glass, you will more or less eliminate punctures.Try to stick to the well-swept part of the road – wobbling along in the gutter is a recipe for tyre failure. If in doubt, jump off and push – or even carry – the bike through. Once back on a clear stretch of road, check the tyres quickly for foreign bodies and decide whether to remove them. If an object has barely begun to penetrate the tyre, remove it. If in doubt, break it off flush with the tyre and carry on – the inner tube will sometimes grasp and seal an object with very little loss of air. If you hear or feel a rhythmic bump, always stop and investigate.This sort of procedure takes seconds, but can save hours of unpleasantness.

Traditional cycle tourists use all sorts of clever tricks to prevent debris penetrating the tyre, including blades positioned close to the tread to knock objects off.These sort of things are best left to the experts, because they can cause more harm than good if poorly fitted or adjusted. However, it’s probably worth pouring in one of the proprietary leak sealants, but don’t forget those regular inspections.


Some cyclists take pride in their instantaneous roadside repairs, but a large proportion of mine seem to fail, particularly in miserable weather.These days, I very rarely bother to carry a puncture repair kit, following instead a few simple rules:

a) Holts Tyre Weld is a mixture of sealant foam and compressed gas that repairs smaller punctures and re-inflates the tyre. Designed to inflate one tubeless car tyre,Tyre Weld is arguably more effective with bicycle tyres, providing up to a dozen repairs, depending on the canister and tyre sizes.This product will usually get you home, and can even be treated as a permanent repair if you’re lucky. It won’t work in every case, but it seems to be wholly or partially effective in about 75% of punctures in my experience.

b) If sealant doesn’t work, it may be possible to pump the tyre up, ride on, re-inflate, ride on, and so forth, especially with front tyres, where less pressure is required. It’s not much fun, but easier than wrestling with the tyre, tube and glue by the roadside. Sealant foam can take a few minutes to work, so this technique may gradually provide a cure.

c) Ride a folding bike! If the worst happens, you can complete your journey or reach a cycle shop by train, bus, taxi or hitching a lift.This last resort is rare, but a folder does give peace of mind that you will be able to get home quickly and easily by other means.

Bicycle Brake Lights

Professor PivotI’ve been trying to locate a rear LED light incorporating a brake light.The only ones I have seen recently are those gimmicky types with built-in indicator lights. I’ve managed to find a few products on the internet, such as Mavic, whose ML-273EZ seems to fit the bill.

Gethin Sheppard
Cowbridge, South Glamorgan

The Mavic design shows two LEDs in rear light mode, but five under braking, which would certainly prove visible at dusk, or during the night. Surprisingly though, the system is activated by a rather crude switch looped over the rear V-brake. Useless if your machine doesn’t happen to feature V-brakes, and rather dubious if it does.The gimmick sounds like the B-Seen, a rather expensive lighting, indicator and brake light set widely advertised in the UK cycling press in the last few months.

However, a light weight and effective brake light system utilising ultra-bright red LEDs is perfectly feasible these days and could be made to function automatically using a ‘tilt switch’.With all the componentry inside the rear lamp housing, there would be no wires, no switches and very little extra weight over a conventional rear lamp. Sensitivity would not be as good as a wired system, but probably adequate. I believe there is no legal difficulty with fitting a brake light in the UK, but I don’t think a device of this kind is currently on sale. If anyone knows better, please do get in touch.

Solar PV Panels

Solar Powered Bicycle – The Results

Professor PivotSolar Power Results

Readers may recall the launch of my solar vehicle experiments in the last issue. At the time – having made a number of purely theoretical calculations – I concluded that a 20 watt solar panel might just be viable on an electric bicycle. In the event, for reasons of panel size and cost, it was necessary to settle on a pair of five watt panels, reducing the potential gains, but keeping the apparatus down to a manageable size.Why two panels? It’s simply that most small solar panels provide a 12 volt output, but the Lafree (the subject of my experiment) is, like many other electric bicycles, 24 volt, requiring two smaller panels connected in series.

Fitting solar panels to the bicycle brings a number of advantages: greater range, a less traumatic life for the battery (thanks to the constant trickle charge, reduced peak loading, and ‘shallower’ charge cycle), plus greater efficiency, because of the direct motor supply. Batteries are around 90% efficient when charging and much the same when discharging. That’s pretty good as these things go, but a direct supply will still give a bonus.

The disadvantages are the weight and cumbersome nature of even the smallest panels, and the potential screening effect of roadside buildings and trees, the rider, and of course vehicle orientation. In other words, if the panels are not angled towards direct sunlight for most of the journey, there is little point in carrying them around.

Static panels are remote from the vehicle, so range is unimproved. On the positive side, the panels can be positioned to capture most of the available solar radiation from dawn to dusk and they’ll charge all day, every day, and even in the rain, when the bicycle might be under cover. A spare battery is not obligatory with this system, but it does mean charging can be undertaken continuously – more effective if the bicycle is away for most of the day.

The Roof Option

After a few measurements, and long hours weighing up the options, it became clear that even the smallest solar panels would be unsuitable for use on a conventional bicycle that has to be wheeled through narrow gaps, leant against shop windows, dumped on the ground, and so forth. A recumbent trike would be a different proposition, of course, and I hope to investigate this option in a later issue.

The A to B long-term test Lafree spends most of its life towing a trailer.The roof of a child trailer would make a good location for solar panels, but one would need to plug and unplug the panels whenever the trailer was connected, and charging would not be possible while riding solo.Trailer mounting could make sense for a long hilly touring holiday, or for a commercial bike/trailer outfit left connected all day, but not for typical ‘school-run’ or commuter applications.

…a roof of this size receives a steady eight kilowatts of power from the sun…

Taking all these factors into account, it became clear that static ‘base station’ panels were the best option, and such was the persuasiveness of this argument that no attempt was made to rig the panels to the bike or trailer, even on a temporary basis.

In this case, we were fortunate to have a building aligned east-west, giving eight square metres of south-facing roof at 20 degrees to the horizontal.That’s a good compromise in British latitudes, although a steeper angle would be more productive in winter, and a flatter roof slightly more efficient in summer.

A roof of this size receives a steady eight kilowatts of power from the sun on a bright day, which could generate around 800 watts, working on our 10% efficiency formula from the last issue. But would our tiny five watt panels provide enough power?

solar-powered-bicycle-1The answer depends on the time of year, the amount the bicycle is used, and the cloud cover.The solid line on the graph indicates solar charging on a typical June day. Recorded just a week before the summer solstice, daylight hours are obviously very long, although one should not assume that solar radiation is necessarily stronger in June than at other times in the summer.

Note that the panels begin to provide a small charge well before sunrise, and continue to work after sunset, albeit at a low rate of charge.The real current flow begins when the sun first strikes the panels at 9am, climbing rapidly to peak at 2pm, before falling equally rapidly until 6pm. Note the pronounced troughs in production during the afternoon, when the sun is obscured by passing clouds. Readers will hardly need reminding that this is a fairly typical pattern in the United Kingdom, as clouds bubble up in the heat of the day.

Solar PV PanelsIn ideal conditions, we can expect our two 5 watt panels to produce about 70 Watt/hours (Wh) of electricity, and in practice, our representative English early summer day produces just under 60 Watt-hours. One could expect to do a little better, or much, much worse, according to conditions. Angling the panels to follow the sun would produce a little more power, but certainly not enough to justify the complication involved.

Sixty Wh is not much, but it’s enough to propel the efficient Lafree for about seven miles with gentle pedal-assistance. In other words, if your daily range does not exceed that amount (or, for example, 14 miles every second day), a ten watt power station should keep you moving in fine summer weather without the need to plug into the mains supply.

Our test bike travels further than this – about ten miles four days a week towing a child trailer (an effective solo mileage of 13.3 miles per day), plus smaller variable distances on the other three days, so a conventional supplement is required.

And the weather is not always fine! The figures below cover a typical 16-day period in late June. Early summer 2003 will be remembered as a very dry period, but cloud cover was about average, so the figures can be regarded as typical:

Average Daily Mileage Daily Conventional Charge Daily Solar Charge Solar Charge %
9 miles* 28Wh 36.6Wh 57%

* The actual mileage of 6.8 miles per day has been multiplied by 1.33 to make allowance for the trailer. Nine miles would be a more realistic solo mileage.

In the summer, and with modest mileage, this basic system works well. By doubling the number of panels, we could either produce 100% of our power requirements, or double our daily mileage. Conventional rigid roof-panels are somewhat cheaper than the flexible panels we have used in this experiment, so for example, £300 should buy two 12 volt, 15 watt panels, providing a surplus of power under most conditions.

Rather than directly charging an expensive spare electric bike power pack, another option is to charge a cheap-and-cheerful 12 volt car battery direct from a single 12 volt panel, topping up the bike power pack through an inverter, which converts the 12 volts from the panel to 240 volts to run the conventional charger, which converts it back to the voltage used by the bicycle. One large panel is cheaper than two small ones; a second bike battery is unnecessary; and the car battery provides capacity of 500 watt/hours or more to carry the system through gloomy days. However, the efficiency of this multi-stage operation can prove alarmingly low, and there’s the extra cost to consider.

A 30 watt solar power station will cost around £300 for the solar panel and regulator, plus £100 for a basic car battery and inverter. In fine weather, such a system would easily generate enough power to give a daily solar range of 20 miles or more.This would also be the best solution for charging a 36 volt machine such as the Powabyke, for which three 12 volt solar panels would be too expensive and cumbersome. Note, though, that the Powabyke is a less efficient machine, so mileage would be rather less.

The Bottom Line

Sceptical readers may have noticed the accent on weather conditions throughout this article.The problem is that a ten watt panel will indeed provide a peak output of ten watts and a mean figure of five watts or so on a sunny day in June. But on an overcast day, these figures can be halved, and in really grim weather – even in high summer – output can fall to 20% of the rated figure, or even less. Note our average daily figure of 36.6Wh, which equates to about 3.6 watts over ten hours. And don’t expect to save any money: our ten watt system might produce eight kilowatt/hours per year, but at a cost of over £2 per Kwh (compared to £0.06 from the mains), assuming a ten-year panel life.

But if you have a hankering to run a solar vehicle, the means is clearly available.The Lafree/solar panel combination is relatively cheap, simple to use, and should provide 1,000 solar miles a year, even in temperate Britain, although we have yet to verify the winter figures… Despite all the talk of sustainability, no other vehicle or power system can match this sort of result for such a modest outlay.

Solar panels, inverters and other equipment are available from a number of specialist suppliers, principally yacht chandlers, such as Compass Watersports: A guide to watts, amps and volts can be found on our web site:


Solar Powered Transport

Professor Pivot“I’ve always been interested in the idea of solar-powered transport, but no-one seems to have built anything practical yet. Is a solar vehicle a practical proposition in the UK?”

Jonathon Crouch
King’s Lynn

Professor Pivot replies:

Clean inexhaustible solar power has been a transport dream since photovoltaic cells first began converting light directly into electricity, but the reality seems as far away as ever. Solar cells have many applications these days, from lighting remote telephone boxes to powering satellites, but for transport, the problems are two-fold: cost and energy efficiency. At £5 to £20 per watt (bigger panels are much cheaper), the cost has changed little for some years: It’s the classic Catch 22 of high technology produced in low volumes, but increased demand for other high technology products has seen prices tumble, so the same is bound to happen to photovoltaics eventually.

Energy efficiency is a more taxing problem. Photovoltaics are improving rapidly, but most convert only 10 to 15% of the light energy hitting their surface into electrical power. Specialist cells of 25% efficiency are becoming available, and 35% or more is possible in the laboratory, but to avoid disappointment, we should work on a performance of a little over 10%.

Solar Cars?

Cover the horizontal panels of a typical car in photovoltaics, and you might cram in six square metres, trapping around 3 kilowatts of energy, but giving a peak output of only 0.4 kilowatts of electricity.To reproduce internal combustion performance, you’d need 30 kilowatts or so.Thus, with today’s technology, under ideal summer conditions, we could expect to generate around 1% of the peak power required. In practice, cars spend far more time sitting in the sun than moving, but even with the best panels charging a battery all day long, range would be very limited, and practically nil in winter.

Solar cars have achieved some amazing feats, of course, but the successful machines utilise a large surface area of priceless aerospace-grade panels to optimise power input, with sophisticated motors and lightweight construction to optimise performance. And it’s no coincidence that the annual solar challenge takes place in the Australian desert… But as we all know, cars are notoriously energy-hungry machines, and we can do a great deal better with other modes. Motorcycles and planes are even worse in the power requirement/surface area stakes, but low-speed motor boats are feasible, and a few have been produced. Best of all, though, is the humble bicycle – true, it offers a modest surface area, but it has an even more modest power demand.

…bicycles, like cars, spend a lot of time sitting in the sun, going nowhere…

The power requirement of electric bicycles varies a great deal, but we know from experience that the Panasonic-equipped Giant Lafree is the most efficient currently available, with a mean consumption of less than 100 watts under typical conditions.We could generate 100 watts from a panel measuring about 1.5 square metres, which would certainly be feasible on a faired recumbent. However, with a battery on board, there’s no need to generate all the power, particularly as bicycles, like cars, spend a lot of time sitting idly in the sun, going nowhere. In practice, a much smaller panel generating 20 watts would more or less recharge the Lafree battery during a long sunny day, giving a daily solar range of nearly 20 miles. If one were to start the day with a full battery, the solar boost might extend the non-stop range from 20 to 24 miles, or anything up to 40 miles spread over the course of a day, provided the bike was left in the sun between rides. A solar charger of this kind might also enhance the battery life, thanks to reduced current drain on hills and a steady trickle charge, rather than a daily boost. Suddenly the technology looks more practical.

In Practice…

uni-solar-usf-5-solar-pvThe bad news is that solar panels are generally unsuited to use on bicycles. Although the panels themselves are light, they’re fragile too, so they’re usually housed in a heavy rigid frame.A few lighter, flexible panels are produced for boats and mobile homes, and although these are not particularly space-efficient, we must concentrate our search in this area.

Taking weight, space constraints and price into account, a pair of Uni- Solar flexible USF5 panels look like a good compromise.The two panels weigh 1.1kg, measure 44cm x 54cm and cost around £180 in the UK. Rated output is 300mA at 33 volts, or 10 watts – only half the power required to refill the battery during the day, but enough to provide some reliable data. By comparison, a spare battery for the Lafree would give twice as much mileage (in all weathers, of course), but cost a little more and weigh three times as much.

I’ll be fitting the panels to A to B’s long-term Giant Lafree in the next few weeks, and reporting back in subsequent issues. In terms of practicality, this is cutting edge stuff:We The Uni-Solar USF5. The border is unproductive, but gives aim to find whether panels are the panel some protection from inevitable knocks practical in our gloomy climes, or whether the weight and bulk cannot be justified for everyday use.

Perhaps carefully angled panels on a south-facing roof would be more effective? Or a combination of the two systems? More in the next issue.