Busted!

April 30th was not a happy day for me. In addition to the torque arm failing – mangling the phase wires on my nice new motor and blowing a FET in my controller – I also got snapped by a speed trap. It was apparently one of those automated signs that tells you your speed as you approach and then says “Thank you!” or “Slow down” depending on the verdict. Whle I was taking the bike for a spin on the new motor and testing its responsiveness, I ended up wellying it rather more than I was planning to. The sign duly told me “38mph – Slow down!”, which I promptly did, but I discovered to my dismay that these signs are apparently now fitted with the same automated cameras that inhabit the more menacing gatsos.

[Edit: I’ve just been told by someone on Endless Sphere that the photo was actually taken with a Lidar gun, probably by someone sitting in a parked car nearby]

This was of course my own fault: My speedometer died the first time it went off the clock, and I’ve not got round to fixing it. The speedo/tacho cable and the front wheel assembly that drives it are fine, but clearly something behind the dash isn’t right. Fortunately, I’ve been given the option of 4-hour lecture on road safety in place of a fine and points on my license, and I’ve duly signed up for a class in a couple of weeks time.

Disassembling the hub motor

As I reported last time, the phase wires to the hub motor got a bit chewed up when the torque arm failed and the axle began to creep around in its fitting. Though the repair to the cable isn’t in principle that difficult, I knew next to nothing about hub motors and so was a bit nervous about attempting this job myself.  So at first I took it down to a motor specialist at a nearby industrial estate and asked the guy there if he could take it apart and tell me what needs doing. If it wasn’t too expensive, I figured, I’d just have it done proferssionally.

However after a week or so and two follow-up visits, it was clear that he wasn’t especially keen on actually doing anything and had just left it sitting in a corner until I lost patience and made an excuse to take it back. But in the mean time I’d been given some advice on how to get the thing apart myself, and it turned out to be a lot easier than I thought it would be to at least get the stator (the main motor assembly) out of the wheel.

Just a bit of pressure on the tyre and stator pops right out!

Once all the little alan bolts have been removed from the edge of the housing, it is just a case of resting the left-side end of the axle on a piece of wood and applying a bit of force to the tyre by pushing down firmly on it. The stator popped out with very little fuss.

Though I now had access to the inside of the unit, there was still the hub plate (right) that needed removing, and it wasn’t at all obvious what it was that was securing this to the rest of the assembly.

Again, I was unsure about how to proceed, so just thought I would drop it off at a local mechanic’s shop to see if they could tell me how to remove the plate. They said they weren’t sure just from looking at it it, but to leave it with them and they’d “take a look at it”. Unsurprisingly,  two phone calls, two visits and nearly two weeks later, it was clear that they were every bit as lazy or disorganised as the previous taker, so I made an excuse and took it back again. I’ve been given some more advice on how to get this plate off, and will give it a go when I find time.

In recent weeks I’ve been indundated with other work, and so for now the bike is in pieces waiting for a rebuild. The LiFePO4 pack has a couple of weak cell-pairs, and is in pieces while I get replacement cells (I have two spare, but still need another couple). The premature decline of these cells is probably due to a couple of times where I ran the bank dangerously low when I really needed to be somewhere even though I was low on juice. I never wired the LVC portion of the BMS into my throttle so it would automatically cut power to the controller if individual cells went to low. Instead I just made regular checks with my bank monitor and kept the bank regularly topped up, seldom letting it run down significantly. The times where I let it run way low would have hurt the weakest cells, and shortened their lifespan, which is one of the risks you take if you operate ‘without a net’ like I do. The cells, though, can still be put to use in non-bank related projects, but are making enough of a dent in the range to warrant replacement.

The lithium pack – two weak cell pairs need replacing

I’m also awaiting the repair to my controller, but can run just fine on one of my cheap spares until that comes back from Lyen’s little workshop.

Maintenance on the battery box, swing-arm and centre-stand

While the battery box is empty, I’ve decided that this would be a good time to sort out a broken weld on one of the brackets securing it to the frame (courtesy of a massive pothole that nearly floored me), and a split along one edge – again  from a weakened or poor weld (Chinese quality control at work). I have an arc-welder now, so I’m going to attempt this myself when I can find time.

I’m also going to take care of the centre-stand and swing arm. They’re both a bit tatty looking and rusty round the gills, so I’ll take them off and give them a respray. The stand also needs extending by about an inch to accommodate my bigger tyres, so another weld-job is in order there too.

The Big Bopper

Work in progress – fitting the “1500W” motor with its formidable new Michelin Bopper tyre

As some of you may have noticed, I’ve spent quite a bit of time these last few month trying to get hold of a hub motor that was more up to the job of handling the extra power that my performance controller can provide. The weak link in the system as it stood was those feeble phase wires on the 1400W motor that comes with the Ego Scoota. I’d trawled the internet and posted messages to forums in an attempt to find somebody who could supply me with a 2000W motor, but to no avail. Chinese vendors either simply didn’t have the motors that they were advertising, or else insisted on the deal-breaker payment method comprising of “T/T” payments, which comprises simply transferring money directly into their bank account. I never deal with anyone on that basis, as – as far as I’m concerned – it’s tantamount to handing over a paper bag of cash to a stranger in a park.

Other scooter owners, however, had reported going instead for a 1500W motor. Users like Hohi on the Electric Motoring Forum reported that even though the nominal rating of these motors was only a fraction more that the original, the phase wires that come with them are quite impressive, and that these motors were actually quite significantly more powerful. The nominal rating evidently had to be taken very much with a pinch of salt.

Hohi got his from a forum member Steavan who occasionaly buys up and breaks these bikes, so when Steavan said that he could supply me with a similar motor, I jumped at the chance, especially since he was only a reasonably short 50 mile hop away in neighbouring Leicestershire. Two weeks late and one jaunt to a scenic cottage in the countryside later, I was the proud owner of a 1500W motor that formally belonged to a Xinben Ambition. This bike had apparently only gone about about 20km from new before the Lithium battery died. It was therefore in near mint condition, with nice fresh brake pads. Though the case was stamped with “1500W 48V”, it looked identical to the one housing my ego motor, but as reported by Hohi, the thickness of the cable was worlds apart!

Wow! The phase wires on the 1500W motor shown alongside those of my existing “1400W” one

As you can see the difference is impressive. However it also made me wonder about whether the gauge of the wiring on the internal coils and the magnets inside would also be similarly improved. One clue to this, I figured, would be the difference in weights between the two motors, so when it came to changing wheels I made a point of giving them both a weigh-in. The results:

“1400W” Ego motor  – 24.2Kg

“1500W” Xinben motor – 29.4Kg

An impressive 5.2Kg extra weight. Clearly there’s a lot more to the difference between these motors than the nominal ratings suggest, so what the motor actually gets called needs to be taken very much with a pinch of salt.

Tyre options

At about the same time as my search for a better motor, I was also thinking of ways of squeezing a little more top speed out of the bike. Though the performance controllers that are available for these bikes are capable of delivering as much power as the wiring on the bike’s system is capable of handling, they are still limited by the fact that motors have a maximum RPM for a given voltage. The Lithium banks help in this respect as the 72V SLA ‘equivalent’ has a nominal voltage of 76.8V for a 24s pack, with a fully charged pack delivering as much as 84V. As I reported in my blog entries on the K-62 tyres, though, I realised that simply increasing the diameter of the tyre could be one way of upping the bike’s top speed. Bigger diameter tyres mean more distance travelled for each turn of the motor, so providing your system can deliver the power, some of that torque can be traded in for a little more top speed.

In the case of the K-62s, the difference in diameter was fairly modest. The 3.0″ rim tyres that came with the ego mean a total diameter of about 16″, while the 3.5″ K-62s up that figure to about 17″ – a modest 6.25% improvement. I had, however, heard report that 4.0″ tyres could be acquired which would up this value even more. However to my disappointment all of the 4.0″ tyres that I could find required an inner tube, which to my mind added unnecessary expense and complexity to what should be a simple swap.

Say hello to the Michelin Bopper!

As I learned a little more about tyre specifications and figured out how to decode their specifications,  I realised that I did indeed have more upgrade options when it came to the tyres. Figuring this out was not helped by the fact that there appear to be two different formats, one imperial and one metric, for describing the dimensions of tyres. In the one that I was familiar with from my K-62s, the height of the tyres is given in inches, followed by the diameter of the rim and then the load rating (how much it can carry) followed by the speed rating. Hence 3.50-10 tells you in a straightforward way that the tyre height (from the outside to the inside) is 3.5″ and that it is intended for a 10″.

The other format, though, is slightly different and more complicated. Here, the width of the tyre is given in mm, followed by the profile or aspect ratio of the tyre, then the rim size. This second figure is easily confused with the height, but that it is not. The aspect ratio is the ratio of the tyre height to the width of the tyre. Hence to get the actual height of the tyre you need to multiply the width by the percentage given as the aspect ratio. So for a tyre designated 120/90, you multiply the width of 120 by 0.9 to give the height, = 108mm. Then you turn this into centimeters (=10.6cm) and divide this by 2.56 to get the height in inches, in this case 4.22″.

Armed with this knowledge, I had a look at what tyres were available for a 10″ rim. It turns out that tyre sizes as big as 130/90 are available, translating to an impressive 4.57″, however I was concerned about whether tyres of such a size would actually be able to fit on the Ego Scoota without rubbing against the mudguards or the brackets at either side that hold it in place, and – more importantly – offer enough clearance from the battery box taking into account the movements of the swing-arm that occur as the shocks compress in response to bumps in the road.

The Bopper – nice and chunky with a greatly improved 4.2″ height

After doing a bit of measuring up, and thinking about the trade off in power that would result from increased tyre height, I decided that the 130/90 was too risky. It would come perilously close to rubbing against the surrounding structures or jamming up against the battery box over a bump. In the end I settled for the 120/90 Michelin Bopper, and ordered one from Oponeo.co.uk. I was more than happy with their service: The bopper was a substiture order for a 4.00″ K-62 that I ordered by mistake not realising it required an inner tube. Their freephone support people just told me to refuse delivery of the K-62 and they would replace it with the bopper, which they duly did.

Even next to the chunkier K-62s the difference is impressive. The diameter has now gone from the original 16″ to a nice, beefy 18.44″. That’s a 15.25% increase on the original 3.00″ tyres or an 8.5% increase over the 3.50″ K-62s. I was delighted with the result!

The Michelin Bopper (guess which!), next to the old K-62

There was only one teething problem on fitting the new wheel with its bigger tyres.The mudguard is slightly asymmetrical, and the leading edge closest the battery box was scraping against tyre slightly while under load. This turned out to be due to the fact the the forward, hidden part of the mudguard is ragged and looks like it was poorly cut by the manufacturers, however this can easily be remedied by cutting off a part of the forward section (it doesn’t need to go all the way down below the battery box anyway). For now I’ve removed the mudguard until I get round to doing this.

Time for a test run

With the motor fitted and wired up, you can see how the phase wires now look more in proportion with the thick guage counterparts running from the Lyen controller.

A test run was very reassuring. Starting at 30A, I raised the controller settings to 40A, then finally to 45A, and was impressed at the improvement in power. However, though this was a big step forward for me, this story did not have a happy ending. The higher power combined with the bigger wheel size quickly identified the newest weak link in the system – the torque arm that secures the axle to the frame and stops it rotating.

The torque arm and axle assembly (this one from the old motor)

… and from above, showing the flattened section of thread used to hold the axle in place

A few miles into my otherwise delightful test run, the controller began sputtering and eventually died. Examiningthe rear wheel it became evident that something was very wrong. The axle had worn away the slot in the torque arm that was supposed to hold it in place, and begun to rotate. The cable to the motor had been slowly wrapping its way round the axle, until the edges of the hole where it enters the motor chewed into the wires, shorting the controller and making a fair old mess of things. 😦

Noooo! – My nice new motor cable gets mangled up as the torque arm fails

Other people, it turns out have had a similar problem, and a couple of owners have had more solid replacements custom made to replace the existing one. So for now, it’s back to the old motor and my stock controller until I can fix the cable and get the controller repaired.

The adventure continues…