Speedo-Driver Problems

August 7, 2013


The front wheel with axle and speedo-driver assembly removed

My bike has just celebrated its 3 year birthday, and with it, it’s now required to have an annual MOT inspection (UK), therefore much of the last couple of weeks was spent fixing a few things that needed doing to make it roadworthy. There was a broken weld on the centre-stand and a nail stuck in the back tyre. Most importantly, the speedometer hadn’t worked for a good while now, ever since the speed had first gone ‘off the clock’. I’d been putting the job off because it meant dismantling the instrument display to check on possible causes there.

The speedo works via a little driver attached to the axle by the front wheel (above right, and below). As the wheel rotates, this turns little cogs in the driver, that turns an impeller, which hooks into the base of the speedo cable. At first, I’d come to the conclusion that since spinning the wheel made the little impeller (the slot poking out of the speedo-driver) rotate, then the fault must be with the cable or behind the instrument display. The cable checked out fine, and oddly, so did the instrument display, so I backtracked to the speedo-driver again.

P1020577 [1600x1200]

The speedo-driver assembly (attached)


Wheel with speedo-driver coupler removed

It turned out that though the impeller was indeed turning, any light resistance would make it slip. This slippage, it turned out, was due to a worn coupler. The piece – a bit like a bottle cap that fits in a recess at the centre of the wheel – anchors itself to points in the driver via a couple of little tabs at either side.

The tabs on mine were worn away and mashed-up looking, and no amount of bending them would make them engage. Eventually, though, a local bike shop found one identical that got me back in action for the MOT.


The coupler to secure the speedo-driver assembly

A week later though, and the day before the MOT, the speedo failed again. It turns out that the tabs on this new one had gotten mashed down, so I plied them back up again as a repair. I’m not sure why it failed so quickly – maybe the relatively high speed of the bike, and the fact that the cable could do with some grease, resulted in it sticking as it rotated. I’m going to cover the whole length of the speedo cable in grease next, to see if that fixes it. I might also extent the tabs further up by sawing or slitting extra length for the tabs from the base. They don’t seem to go up far enough to make a firm connection with the driver assembly so a bit more bodging might be in order.

The good news, though, is that the bike got through its test just fine. A few days previously I’d had my local bike shop fix the tyre (I took tools with me and took the back wheel off myself), and so all I had to do was hope the speedo would last the duration of the test, which it did.  Despite my concerns (and others’) about my custom-built, LED headlights with their imitation, “dip-effect” function, there was no mention of this. Neither was there any quibble about the ever-so-slightly-too-small licence plate, either. He just tightened the rear brakes up a bit and left it at that. 🙂


Charge Point Success!

April 4, 2013

P1050069 [1600]

Success! My town now has a public charge point for electric vehicles

In early 2011, after I’d enthusiastically set about putting my bike through just about every possible upgrade, it soon dawned on me that provision for public charge points in the UK is pretty non-existent. Though they have a park-and-charge scheme in London, it is a struggle to find any public charge points anywhere else.

I met with my local MP (member of parliament), Sir Alan Meale to get some advice on how to go about getting more of these things available to the public and – as a campaigner with a long history of pursuing environmental issues – he had lots to say about what needs to be done to get more of this type of service put in place. The key, he said, was to target planning proposals for new developments and file an objection to plans for any car park that does not offer a charge point. Planning applications are posted in everybody’s local newspapers, and the process of local consultation is usually well organised, with council websites offering full electronic details of applications and plans, with the option offered for any members of the public offer their objections to any part of that development. These objections are all then addressed in a committee meeting between council officials and developers, with issues being discussed, and modifications being made to plans accordingly.

The big supermarket development

Practically every town in the country has some kind of big, commercial development in progress, with supermarkets like Tesco, Sainsbury and Aldi (here in the UK) rolling out huge new sell-it-all style warehoues of the kind we’re becoming increasingly accustomed to. The building spree of these growing commercial empires, though, offers perfect opportunities for electric motoring enthusiasts to take matters into their own hands by simply hunting down the relevant planning applications and complaining about lack of provision for motorcycles in general, and for electric vehicles specifically.

So when my MP reminded me of a massive, joint Sainsburys/Aldi development that was scheduled to replace a smaller existing Sainsbury’s, I did as he instructed and tracked down the plans for the development. The plans incorporated an impressive, 400 bay car park. So I formally filed an objection on the grounds that a) the parking development did not make adequate provision for motorcycles, and b) that no charge points were available for electric vehicles. I also wrote a letter to Sir Alan asking him to support my proposal for the plans to be amended to include these provisions, which he duly did, sending a letter to the committee dealing with the proposal.

I thought nothing more of it, and had totally forgotten about it when a few months ago a family member reported seeing a charge point in the new development.

P1050070 [1600]

Here it is! It’s activated by a little card that you get from staff in the supermarket. You wave the card across a little panel on the unit to release a flap on either side, each of which houses a regular, 240V power point. If you look closely at the picture you can make out a little display counting down 3 hours, which is the amount of time I’m allowed for a single ‘session’.

P1050073 [1600]

You can see the little green LED strip lit up indicating it’s live

Here you can see the charge point in action with my custom charger top-box. It just so happens that my bike will charge from flat in 2 hours 50 minutes, so 3 hours is more than I’ll need for a good top up while shopping.

P1050072 [1600]

So my little town is now host to the second of only two power points that exist within about a hundred miles of here!

Want one in your town too? If my story is any guide to how easy it is if you go through the proper channels, then all you have to do is a little homework, and some letter writing. These things aren’t going to just pop up all by themselves, YOU can get involved and be part of the electric motoring revolution. I appeal to all electric vehicle owners in the UK and elsewhere – both as motorcyclists and as electric vehicle enthusiasts – to actively seek out and file objections to local developments with “car” parks (just the fact that they are called “car” parks speaks volumes about the prevailing attitude towards motorcycles, any pedants here might also suggest that the authorities start referring to them as “parking areas”).

P1050071 [1600]

So please, folks, we CAN make a difference. If you want to see one of these in your local area, just do the following:

1) Check the planning applications in your local newspaper for anything promising a large car park. In particular, look out for a new commercial development housing a supermarket or big retail outlet

2) Go to the link given in the ad to file an objection. Write a polite and reasonable sounding request for provision for a charge point, and for better facilities for motorcycles and bicycles

3) Forward a copy of your request to your local MP asking him/her to support your application in writing.

A meeting is usually called to discuss the proposal, and the public is invited. I didn’t make it to mine, but it evidently didn’t make much of a difference.

Scooter Rebooted Pt. 2 – Heavy Duty Controller Heatsink

January 6, 2013

HS12 1600

A custom, copper heatsink and mount are bolted to the baseplate, extending it to the rear to accommodate the larger, 18-FET controller


  • 4 pieces of – 100 X 260mm 0.3mm copper sheet
  • 3 pieces of – 60 X 260mm 0.3mm copper sheet
  • One piece of – 60 X 220mm 3mm copper sheet
  • One piece of – 60 X 220mm 0.3mm copper sheet
  • 6 3.5mm countersunk head bolts

Having upgraded both the hub motor and the controller, I thought I’d take the opportunity to incorporate a custom heatsink into the rear mounting plate. The current 1500W hub motor is basically maxed out now, with the 16-FET controller more than enough to give it all it can take (about 4KW), so the controller doesn’t run very hot at the moment, but in order to clear an upgrade path for a more powerful hub motor, as well as address an issue with mounting controller that’s too big for the baseplate, I decided to design a heatsink custom made to maximise contact with the controller casing.

HS03 1600

Most controllers are either 180mm or 210mm from end-to-end of the baseplate. This one, though is 260mm, and as you can see from the picture above it overhangs the baseplate by 50mm or so. My solution to this was to build a heatsink that would also double as an extension for the baseplate. By stacking cut pieces of thin copper sheet of alternating width and bolting them firmly to the base, I get good wide fins that help with heat dissipation. The air being funnelled to the controller area, and the good thermal contact with the base plate should both help draw heat away from the controller.

HS02 1600

Below you can see stage one of the heatsink which acts as the main base for the controller.

HS04 1600

As any of you who have dabbled with controller will know, however, the controller casing has an odd recess underneath that prevents most of the controller case making good thermal contact. Only the brackets at the end, and a section of case running the length either side make contact with the baseplate when the controller is mounted. The recess is about 3.2mm, and I had two special pieces cut to act as a seat that would allow the base of the casing to make near-complete thermal contact with the heatsink.

Below you can see the main item, a 3mm thick slab of copper, and a thinner 0.3mm sheet of the same size that brings the plate almost flush with (actually about 0.1mm proud of) the controller base.

HS05 1600

The tricky bit in mounting this on top of the first stage of the heatsink is to do it in such a way that the heads of bolts are sunk so they are flush with the base of the controller case. Any bolt poking up will stop the case from making good thermal contact.

HS06 1600

After some fiddling with different drill bit sizes, and very gentle drilling I eventually ended up with a second stage plate with six suitably  recessed bolt holes.

HS07 1600

HS08 1600

Finally, six corresponding holes were drilled through the heatsink base sheets and the baseplate. From the underside, you can see both parts of the heatsink now firmly bolted to the baseplate. I cut off the excess lenghs of these bolts with a grinder. You can see how the heatsink adds the extra required length to the baseplate from below.

HS14 1600

The finished product! A nice big heatsink that doubles as a lengthened baseplate for the controller.

HS11 1600

HS12 1600

A final touch before mounting the controller was to remove some instructional stickers that were on the base. Thus, the aluminium of the base sits directly onto its heatsink mount.

HS13 1600

And there you have it. A solid heatsink that helps keep things running nice and cool. You can see a thermocouple I attached so I could monitor its performance. The most striking difference is that though the controller still gets warm when thrashed, it cools down much quicker, with the rate of cooling directly proportional to how hot it’s getting. Definitely a must for people who like to push their controllers to the limits.

Scoota Rebooted Pt.1 – The Big Makeover

September 7, 2012

The rebuilt scooter with Lyen 18-FET 4110 controller, 1500W motor, big Bopper 120/90 tyres and extended centre-stand

While I was dismantling and fixing the motor and had so much of the back end in pieces, it was a good opportunity to give the bike a proper makeover. With all the commuting I’d done in all weathers, and the amount of time it had sat outside my workplace in the pouring rain, it was getting a bit rusty round the gills, with the swing-arm, battery box and other frame parts in need of a clean-up and respray. There was also a broken bracket on the battery case where a crater in the road had jarred me hard enough to almost bring me off the bike.

Swing arm removed

Swing arm and stand removed

There was also the matter of the centre-stand. I’d recently decided to go for even bigger tyres than the 3.5″ K-62s, and decided that the 4.2″ equivalent Michelin Bopper would fit the bike, give me a more comfortable ride, and also increase the amount of distance covered per turn by an extra 8.5% compared to the K-62 tyre, which would aid top speed with a sufficiently powerful controller. However with Big Bopper tyres front and rear, the bike stands at over an inch higher, with the centre stand barely able to touch the ground.

I’d ordered a custom, 18-FET controller from Lyen, and while I was waiting for it to arrive I got to work on cleaning up the bike and giving it a respray. You can see the swing arm below looking a bit worse for wear from the weather.

Likewise the stand, but before the stand got a respray it would need some surgery to make the legs longer. To extend the stand, I decided to do a ‘cut-and-shut’ with two inch-long pieces of steel tubing. Fortunately my local metal worker had some of exactly the same gauge and diameter.

After sanding off the paint around the cuts with steel wool, I set to work on it with an arc-welder.

A bit of grinding to tidy up the welds and it’s an inch higher and ready for its respray. I also took care of the broken bracket on the battery case.

The nice, shiny, resprayed swing arm goes back on with the repaired motor.

After a respray, the back end is ready to be reassembled

This is where I finally get to put into commission my custom-made, heavy-duty torque arms. Because of the length and the longer adjuster slots, the wheel can be slid back a further inch or so keep it well clear of the battery box. Standard torque arms are especially susceptible to wearing out when people use regen, as accelerating and then decelerating under regen alternately turns the axle one way and then the other, which can lead to it gradually chewing its way through the retaining slot. Hopefully these much sturdier pieces will let me use regen with impunity. I might even double up with a couple of extra ones just to be ultra-safe…

While I was at it, I removed the plastic panels to get better access to the frame and battery box. While the battery bank was having a couple of weak cells replaced, I had resprayed the battery box with hammerite inside and out. Once everything was reassembled, the wiring was re-secured with fresh cable ties. To accommodate the larger tyre, the mudguard needed a couple of inches of the forward end trimming off as the tyre was rubbing against it. However that part of the mudguard serves little purpose as it’s below even the base of the battery box.

Sprayed and back in place, the extended stand gives the extra inch that the larger tyres need to allow the bike to be properly parked. The rear wheel therefore has about the same ground-clearance as it did before – about an inch, so that the wheel can be spun while the bike’s on its stand.

The stand – extended by an inch so it works with the bigger tyres

All reassembled and ready to go, there’s just one thing missing: The silver panels that came with the bike had been getting steadily more battered, and one finally broke. While I was buying my motor from Steavan, I also got a couple of fresh panels that he had for sale.

The best match he had for the bike were black ones, which I think look fine.

The Road Test

Once I’d got Lyen’s controller up and running, it was time to run it in, steadily increasing the rated and phase currents until I found a setting that gave me the power I needed, but without getting the controller to run dangerously hot. I eventually settled on 60/150 rated/phase, though a little less than that would probably have made no difference. It seemed that any setting at all above 55A or so made no extra difference to the power. But since I’m putting 4KW plus through a 1500W motor, this is hardly surprising – it seems that the motor just won’t draw any more than that. The next stage on the upgrade path is an even bigger hub motor. The controller, though, runs pretty cool, topping out at 60°C only after a great deal of extended thrashing.

The performance though, is fantastic. Very torquey, with phenomenal acceleration that’s more than adequate to beat most things away from traffic lights – I’ve recently noticed shocked boy racers in BMWs or Audis chasing me to try and make a point.

Though the acceleration is excellent, I didn’t end up with quite the top speed I’d hoped for. With the Lyen 12-FET controller I was getting 43-46 mph satnav. Now I’m getting 48-50mph satnav, topping out at about 52mph when the wind’s in the right direction. That’s the equivalent of about 55mph clock-speed though, and more than adequate for short hops between towns. I was hoping for a little more than this, as I thought my 120/90 tyres would leverage me a bit more speed, but different motors are wound for different types of performance and it’s just pot-luck what you end up with. I believe this one is better on acceleration because it’s wound for ‘torquey’ rather than ‘fast’. Still, I’m more than happy with the results.

With the new motor/controller combination, regen is a little harsher than it was before, and I might tweak the resistor value to make it a bit lighter. I’m also thinking of wiring a button into the regen self connector so I can switch it on and off without plugging my laptop in and changing the settings.

As part of my upgrade, I also built a custom heatsink for the controller. Though heat is not a particular problem at the moment, that might change if I get a 6KW hub motor or something. More about that next time…

Testing the Hub Motor

July 21, 2012

Though hub motors tend to be fairly well behaved and can run for years without problems, they do occasionally go wrong. If the bike suddenly stops working or develops problems, you often end up in a position where you need to find out if it’s the controller or the motor that’s at fault.

Many enthusiasts who’ve done a lot of tinkering with their bikes usually end up with at least one spare controller lying around, in which case it’s just a case of swapping it out to see if that fixes the problem. Most people, however, do not have spares available and may need to do some specific testing to rule out any issues with the motor.

1) Access the Phase Wires and Hall Connector

The first thing to do is to remove the seat to get access to the junction where the phase wires are connected (bottom) and the hall connector socket can be found (top left). The phase wires are the three, thick blue, green and yellow wires. The hall sensor wires are five thin wires: black, red, blue, green and yellow that run from the motor cable to a 6-way mini-connector. On some bikes the phase wires are part of a plug instead, in which case you’ll have to adapt your test accordingly. For this test you will need to detach the phase wires on the motor side from the connection block so that they are isolated from the controller. Leave the hall connector connected, though.

Phase wires and Hall Connector

The phase wires are the wires that provide power to the motor, magnetising sequences of coils that generate the rotational motion. The hall sensors are three little transistor-like devices embedded along the inside edge of the motor, which send information about the motor movement back to the controller.

2) Test the Phase Wires

First, with the bike on its stand (and the bike switched off), test that the rear wheel spins freely. If it doesn’t you most likely have a short between the phase wires, though this is rare. The next thing to do is to remove the phase wires from the block and short them one pair at a time (e.g: yellow/green, green/blue, blue/yellow). As you hold them shorted together, try turning the wheel again. This time, you should find resistance. If you don’t get resistance when any pair of wires is shorted, then there is a break somewhere between the phase wire and the motor.

Phase wires and hall connector, detached

3) Test the Hall Sensors

For the next test keep the phase wires disconnected but make sure that they are kept properly separated. Preferably put a bit of insulation tape on the ends to prevent mishaps. Again, keep the hall connector connected.

Next, you will need to turn the ignition on, and take voltage readings from the wires in the hall connector. Since the connector is plugged in, you will need to access it from the rear, sticking the multimeter probes into the recesses and touching the crimps holding the wires in place.

The hall sensor wires

Take voltage readings across the black wire and – in turn – the blue, green and yellow wires with the back wheel turning slowly. It’s best to get someone to help out with turning the wheel for this part, while you take readings. As the wheel turns, the signal should flip between 0V and about 4.5V, at a rate of about 23 cycles per rotation of the wheel. If you get this reading from each of the wires, then all is well, otherwise, you have a blown hall sensor and will need to open the motor to replace it.

If you have a bit of pocket money spare, and want to make life easier, you could invest in one of these very handy motor/controller tester gadgets that will do it all for you. You just plug in the hall connector, attach clips to the phase wires, and the blinky lights give you the answers straight away! The controller-tester part of the unit, however, is only compatible with controllers up to 60V, though, so it can test the motor – but not the controller –  if you have a 72V system.

Replacing a Hall Sensor

July 12, 2012

Hall effect sensors sometimes fail, but are easy enough to fix

Taking the hub motor apart and replacing a hall sensor might sound like a pretty daunting task, but it’s really not that difficult. The first thing you need to do is get hold of a suitable, replacement SS41 hall-effect sensor. One can usually be found on ebay here, or from RS components (but with a pretty steep postage price) here.

Next you remove the stator (the hub motor section of the rear wheel) from the rear wheel itself (see blog entry) so that you can get at the motor’s inner workings. This done, the hall sensors and wiring are exposed and ready to work on.

The sensors are held in little shaped grooves in the metal treads on one side of the stator’s perimeter. Now you need to identify which sensor holds the wire that you detected a fault on. Each sensor has three wires, – two of them are the 4.5V live (red) and GND (black) which serve the sensor, the third is the hall sensor wire which returns a signal to the controller based on its rotational position in the hub motor casing. The third wire for each hall sensor is yellow, green and blue respectively, as shown below.

In my case, it was the yellow wire that was showing no signal, and this one goes to the middle hall sensor on my stator.

The component itself is held in with a bit of epoxy resin and sometimes has bits of silicone sealant gummed about. The epoxy can be softened by warming it a little with the flame from a cigarette lighter.

A gentle tap with a hammer and screwdriver knocks it loose with ease.

Note how one side of the hall effect sensor has bevelled corners. This side faces outwards (on my motor, at least) and is designed so that it can only fit in its socket the right way round. This is important, as it makes it hard to accidentally reverse the wiring sequence.

Here’s the replacement sensor in situ.

You can more clearly here how it fits with the bevelled corners facing outwards.

Next it’s a simple case of soldering the new sensor in. Snip the head of the old sensor off, strip the hall sensor wires, and desolder the legs from the old one, then solder on the legs of the replacement sensor. Don’t forget to put fresh sections of heatshrink on the wires, and push them way up the wires where the soldering iron won’t prematurely shrink them. Then pull the heatshrink down into place over the soldered legs, like with their neighbours.

This done, it’s a case of fitting the sensor back into its socket. Use a little blob of epoxy resin to hold it firmly in place.

All done! Finally, carefully refit the stator to the rear wheel and test the wiring from the hall connector. If you did everything right, then you should once more have a working hub motor.

Torque Talk

July 4, 2012

The final product! – The Zenid, custom, 12mm, heavy duty torque arm made from laser-cut, 6.3mm steel (alongside the original)

I’ve taken a bit of a break from the bike in recent weeks to catch up on work and to await the delivery of some key parts for my next upgrade. I ordered a nice 18-FET version of Lyen’s controller to replace the 12-FET one I already have, and also had the current one repaired after the FET blew from the torque arm mishap.

Figuring out how to come by a suitable replacement for the rubbish torque arms that come with the hub motors turned out to be something of a journey. First I followed up a link on Endless Sphere to a guy who had custom made some torque rings that looked like they might be suitable, however they were the wrong design for the type of swing arm that the Ego and other electric scooters use.

After enquiring with a local machine shop about the cost of custom machining some suitable parts, and doing some exploring online, I eventually decided that it would be more cost-effective to simply custom design and order my own, and sell the surplus in my shop. The wonders of the internet now mean that there are manufacturers who have pretty much automated the whole process so that you can design, explore cost options, and order whatever you want with little or no human intervention. So I duly figured out how to use some web-integrated CAD software, came up with a design for my dream torque arm, and placed my order.

The CAD design, in 3D preview mode

The only thing that wasn’t automated though, were shipping options. My manufacturer of choice was a US firm (I couldn’t find any UK company who did this, alas), and the only option was for a ridiculously expensive express UPS delivery that cost almost as much as the order itself, so I emailed them and asked if I could have it shipped by a much cheaper (and slower) USPS tracked service. The said it was fine, and told me to just add a note to the CAD drawing for the

I spent the last three weeks occasionally checking the tracking to see how the package was progressing, and it seemed to take a dog’s age to crawl across the planet from from its east-coast origin. When it finally arrived in the UK it spent nearly a full week sitting in customs before finally clearing and continuing its journey. I was none too happy to be landed with an import duty and ‘clearance charge’ of nearly £40 for the privilege of this inconvenience, but duly paid it and eventually ended up with a batch of freshly minted parts on my doorstep this morning.

The real thing, next to the original that came with the motor

Here it is, the real thing, shown next to the inferior original. They’re 62mm x 24mm, and manufactured from 6.3mm laser-cut steel, and have 12mm torque slots, with a 6mm hole for the adjuster bolt, – suitable for the 1400W/1500W motors. The torque slots are slightly wider as the 1500W motor has a 16mm axle on the cable side, as compared to the 14mm on the 1400W motor, though the slot diameter required (on mine at least) is 12mm on both. I’ve also made mine longer, with a wider adjustment slot so that the rear wheel can slide back further to accommodate larger tyres. I got a batch of them, so there’s plenty left to go in my little shop in due course.

There was a moment of panic when I put it on the axle for the 1500W motor – it’s a tight fit and needed a gentle tap with a hammer to get it over the end, but then it slid down the shaft nicely, an almost perfect fit with very little free play! Here it is next to the even worse torque arm that came with the 1500W motor (these are the ones that failed).

The bike is in bits at the moment, while I repair a bracket on the battery box, respray the swing-arm, service the battery pack and fix the wiring on  the hub motor. I’ll get round to this in due course and keep you posted.