The most frustrating thing for a group of enthusiastic engineers is to have their efforts and aspirations thwarted by bureaucracy. But that’s exactly what happened when our shipping company refused to give us our container or even accept any form of payment. We spent an unproductive and relaxing two weeks in Darwin without a car to work on. However, what we achieved in the following fortnight after we finally got our hands on the car cannot be overstated. The car arrived in a “working” condition (we’d done some driving back in the UK at Bourn Airfield) but getting the vehicle race-ready took a huge amount of proverbial elbow grease and midnight oil.

We successfully fitted new lights and LED drivers, tested new driver controls and telemetry, performed solar battery charging tests and re-wired a few things that we’d never got around to in the UK, including the rear-view camera. The mechanical guys chipped in with a new set of wheels, worked out how to fit the Michelin tyres (and then taught a couple of other teams how to do the same) and fitted a brand new canopy. We painted the car and stickered her up with new sponsor logos, after BA successfully lost our first logos package somewhere between the UK and Sydney.

We always knew that testing was going to be critical to race success and not having the car for two weeks effectively robbed us of two weeks’ testing time. Despite this, we spent a couple of days out on the Cox Peninsula road just driving and practising convoy communications. I like to think of testing as deliberately causing problems so that they won’t happen when it actually matters. Pretty much anything that can go wrong in a solar car at some point will, and of all the things that can go wrong, the majority are electrical. The main problem that we “caused” during this testing period was battery related. One of our 5 Cell Management Modules (CMMs) decided to fail in a short-circuit state. These boards had until this point been doing a stellar job of keeping all 80 cells nicely balanced, but this fault destroyed two cells (and their two replacements) and meant we had to replace the CMM. Fortunately we’d brought two spare CMMs with us, unfortunately one of them didn’t work, and the other had an ID that conflicted with one that was still working. Some epic software bodging by “Batt-man” Ed meant we found a working solution. The CMM boards later threw regular over temperature faults and started physically shedding capacitors at an alarming rate. We fixed these problems as they appeared, but were nevertheless quite frustrated by such issues on supposedly reliable hardware. When you build a car, you half expect your home-made components to go wrong, not the stuff built by professionals.

We began the race quietly optimistic that we’d (perhaps) finish the 3000km on solar power. The first day we started fairly low on the grid after putting in a (sensibly) cautious qualifying lap at Hidden Valley due to concerns about the suspension and the new wheels. But when you’re not competing for the top 5 positions, grid position is relatively unimportant in a 3000km endurance race. We got off to a flying start and overtook the usual first hour breakdowns on the way out of Darwin. We had an exciting but (thankfully) uneventful few hours driving. The inevitable first problem occurred when the driver controls stopped functioning. It took the best part of an hour to work out where the problem was and find a solution. We fell back on an earlier version of the controls and carried on driving. Later inspection of the circuit would reveal that an inductor in the power supply had come loose causing total power loss to the steering wheel. The result of this fault was that we were behind our ideal race pace and would struggle to make Katherine (the first control stop) in time. What we failed to realise was that this was not critical and that we could have missed this control stop and perhaps adjusted our strategy accordingly for the following day. Instead we made the mistake of thrashing our car a bit too hard, driving at 20kph faster than we realistically could sustain. We ended up with what we assumed was a flat battery and were forced to trailer.

Another problem also contributed to our first day disappointment. Our telemetry system failed due to a blown fuse in the chase car’s 12V system. Another case of off-the-shelf components failing and lack of testing left us without good battery data while we drained a bit too much juice. As it turned out, the battery wasn’t completely drained, but one or two overly-discharged cells had triggered Low Voltage Protection (LVP). This fault was quite possibly a result of having to replace cells before the race because of the aforementioned CMM short-circuit issue. When a battery pack isn’t given enough opportunity to balance itself (which it wasn’t in this case) the entire pack capacity can be limited by only a few unbalanced cells. If the CMM issue had been triggered earlier with more testing, we would probably have started the race with a well-balanced pack and had access to a bit more capacity on day 1.

By now you might have noticed a common theme in my musings – we didn’t test enough. If shipping hadn’t been so problematic, we might have used our relaxing two weeks in Darwin more productively by (money permitting) spending several days driving the Cox Peninsula road in convoy.

By day two we’d established what went wrong on day 1, and put in place the necessary fixes. This required some proper Outback Engineering and the WSC observer looked on with a concerned expression as we fired up our generator and plugged various cables into the car. After we’d explained (across a slightly difficult language barrier) that we weren’t charging our battery but just powering a couple of laptops and a soldering iron he left us to it.

The rest of the race passed without major technical incident. From this point onwards we encountered all the problems afflicting every team: haze from a large bushfire on days 1-3, bushfires closing the road on day 3, strong crosswinds, lighting storms, quite a bit of rain, and the general lack of sunlight. Every encounter with other teams at control stops began with some engineers standing in a circle looking dejectedly at the clouds and pointing hopefully at a small patch of blue sky in the distance. Towards the end of the race we found ourselves doing a kind of “inverse storm chaser” manoeuvre where we’d trailer as fast as we could away from the bad weather and then spend a few hours at a rest stop where the sky would be lighter or (if we were lucky) we’d get 30 minutes of direct sunlight.

Squeezing every last ounce of energy from a solar electric vehicle is strangely satisfying, especially at the point in a solar race when everyone is resigned to not finishing on solar power alone and what really matters is the number of solar km covered.

So we finished the race in 25th position after covering 1487km on solar power alone. It’d be a lie to say that we’re completely happy with our position. Whilst better weather would almost certainly have enabled us to cover over 2000km, it would also have similarly improved the performance of the other 28 cars who also failed to complete the full distance. We were slightly unlucky to be caught at the back of the pack after day 1 as cloud cover advanced from the north and hit the slowest teams hardest, and we certainly had our share of bad luck with technical issues. But the fact remains that we lost over an hour to an electrical problem that would have been solved before the race if we’d tested more. The resulting sprint to Katherine cost us (and our battery) dearly, and if we’d driven more conservatively we wouldn’t have been forced to trailer on day 1.

One reason for our excessive energy use on day 1 is that our car simply weighed too much. The first half of the race was much hillier than any of us had been expecting and this took a heavy toll. Even with perfect weather, no technical problems, and perfect race strategy, it’s difficult to say whether our current vehicle is genuinely capable of finishing the race on solar power.
We began the year with plans to build a new car, but also with several thousand pounds of debt. Initial optimism, excitement and inspiration gave way to frustration as the decision was made to re-use our 2009 vehicle due to financial issues. This was clearly the right decision at the time – we might never have made it to Australia at all otherwise. To build a competitive solar car requires strong financial backing, and at the start of the year we simply didn’t have this. The other limiting factor is the time required. CUER is composed entirely of Cambridge undergraduate engineers. The Cambridge MEng is demanding and time consuming and where the best solar car teams in the world have a team of full-time engineers on an 18 month sabbatical from their studies, million dollar budgets and dedicated production facilities, we’re the equivalent of a few guys with spanners in a shed. With these constraints and the best will in the world (such as the 2009 team’s: they actually built a new car) you’re never going to build a world-beating vehicle.

Our solar car remains the best in the UK, an achievement we’re immensely proud of. But we recognise that we’re currently not going to beat the top Japanese and Dutch teams at their own game. The World Solar Challenge has made a handful of engineers across the world extremely good at building and racing solar-powered cars that have very little real-world use (but are damn good fun). It’s my personal opinion that CUER should adopt a very “Cambridge” attitude and build a genuinely innovative and perhaps slightly more practical vehicle. This isn’t an easy task and several conditions would make this easier, summarised below in my own personal wish-list:
• A bigger budget (we currently operate with roughly negative money…). This would probably require a large, dedicated business team and might include financial support from the university/department itself.
• A strong dedication of time from perhaps a 10-strong team of undergraduate engineers. Ideally a sabbatical year off the master’s course to concentrate on building the car.
• Dedicated lab space and facilities. Many teams have their own space to build their car. We beg, borrow and steal space from various labs and do a lot of our work in a car park at the back of the department. Other student-run projects in the department suffer from the same lack of facilities. It’s a pretty sad state of affairs when the best university in the world can’t support its own students in such fantastic projects.

Having one or two of the items from the above list would greatly improve our chances of doing something quite special next time around, especially considering at least half the 2011 team are remaining in Cambridge with either PhDs or employment. We’ve all got some fantastic ideas and the technical experience and know-how to put designs into practice.

Having completed the World Solar Challenge is no mean feat, and CUER can be extremely proud that it has now entered twice and still has a fully-functioning solar car. The battery is still workable, we have a robust wiring system, our old and battered solar array still gives us a good power output, and Douglas’s telemetry system performs admirably. Despite the issues, WSC was the experience of a lifetime for all of us.

The battery management system (BMS) is something we’re particularly proud of. Not only does it intercept, interpret and act upon CAN messages from the Lifebatt cell management modules (CMMs), but it also monitors the current into and out of the battery via our Isabellenhütte shunt resistors and shuts off the battery from the rest of the vehicle in the event of a fault.

Ed programs the BMS

Our Homebrew BMS. In a glorious and aesthetically pleasing twist of fate, every component is black

Endeavour’s new, brighter lights are now more-or-less complete. The build process used a variety of cool bits and pieces, including rapid-prototyped positives from Jaguar Land Rover, silicone rubber negative moulds, and finally a transparent resin body for the light itself. We’ve gone for 3W LEDs this time around, so there’s no way anyone’s going to miss us when we brake at the last minute to avoid a kangaroo. We probably won’t run them at 3W in the race, but it’s good to have the option of blinding people behind…

Dan tops up his tan with an array of 3-watt LEDs

Lately we’ve been sharing lab space with these weird guys who call themselves aero specialists, but quite like to take apart brakes and stuff. We even spotted technical director Tom in a rare moment of “getting his hands dirty” taking off a wheel or something. Whatever it is they’re up to, they leave a trail of destruction and discarded tools and carbon fibre wherever they appear:

Tom and George do something mechanical

We begin with the smallest unit in a battery – the cell. This is a single lithium-iron-phosphate (LiFePO4) cell. A cell is essentially a block of chemistry – in its simplest form, it contains two metal electrodes and an electrolyte (here, LiFePO4). What we often refer to as a battery (e.g. Duracell, Energizer) is actually just a single cell. They work by undergoing oxidation and reduction reactions at the two electrodes; these reactions are complementary and require a flow of electrons. It is this flow of electrons that provides the ‘electricity’ from the cell.

Here is one of our cells:

We combine multiple cells by connecting them into parallel pairs. Wiring something in parallel means we join all the positive terminals together to make one positive terminal, and do the same with the negative terminals. This allows us to draw a higher current at the voltage of a single cell. Current is the number of electrons that flows through our circuit per second. Voltage is the amount of energy given to each electron as it flows through the cell. It is easy to see that if you multiply current by voltage (energy per electron times electrons per second) you get a power term – energy transferred per second. In our case, we want about 1.2 kW of power.

The name for four cells wired together is a ‘stack’.

In this picture, you can see two four-cell stacks wired together to create an eight-cell ‘stick’. In the battery pack itself, the physical ‘sticks’ will be ten cells long, and are also called ‘stacks’. However, in terms of monitoring the battery, they will be divided into groups of eight cells.

We bolt together three ten-cell stacks to make a thirty-cell ‘block’:

In the above image you can also see one of LiFeBATT’s cell management modules (it’s the green board). The electrical team have been getting lots of voltage, temperature and status information over CAN off it today – very good news!

We connect four stacks together to make a forty-cell “beast”.

Here is a picture of the stack, the block and the beast laid out in their brackets. Not all the cells are in place yet:

Once the stack, the block and the beast are all mounted inside the battery pack, it becomes known as “the Kraken”. These images show the Kraken partly-constructed, and mounted inside Endeavour’s chassis. As you can see it is a rather tight fit!

So there you have it!

1 Kraken = 1 beast + 1 block + 1 stack = 4 stacks + 3 stacks + 1 stack = 8 sticks + 6 sticks + 2 sticks = 16 4-cell-stacks + 12 4-cell-stacks + 4 4-cell-stacks = 80 cells.

Easy. Of course, that’s because we’re on the metric system now.

- Drill a few holes in the side walls for the top cell supports

- Drill holes to mount the contactor(s), DC-DC converter and CAN node.

- Make up as many 10 cell “sticks” of cells as we can and get these mounted. Putting them in the largest bracket set (“the 40-cell Beast”) would be best so that that part is complete.

- Produce interconnects between the blocks of cells (crimp connectors on thick wire with single Anderson connectors).

This stuff should only take about 3 hours (maybe that’s optimistic…).
When we’ve got this sorted we can try and get some BMS software tested on our prototype CAN nodes with the PicKit. We have also produced a rough outline of what we need to run past the guys at LiFeBATT so we can make sure our BMS catches all error states.

First off, let’s get the formalities out of the way. ‘MPPT’ stands for ‘Maximum Power Point Tracker’. This may have been mentioned in previous blog posts, possibly in an attempt to clear away the aforementioned fog. It’s not entirely certain why they thought it would help. It’s unlikely that the response to this revelation was “oh, Maximum Power Point Trackers – they track the maximum power point! Of course! It’s all so clear!” No, this is a PR challenge even Ronseal would struggle with.

However, unlike the average Ronseal customer (or perhaps not?) we are in a position to understand the relationship between the photons reaching a solar cell, and the amount of useful energy we can get out of it. However energy on its own is not a useful measure. A solar array could provide 1kJ of energy – in fact, they all will, if you wait long enough – but an array that produces 1kJ in 0.5s is better than one that produces it in 20s.

Power, then, is a better measure of solar cell effectiveness – it tells us how much energy a cell is capable of generating in a given time. The greater the power of the cells, the more energy is available each second to be converted into e.g. forward motion via a motor.

Electrical power is calculated by multiplying the voltage by the current. As we have seen, the voltage of a solar cell can vary depending on what material it is made from. There are also other factors influencing this, temperature being a good example.

The current is a measure of how many electrons flow through a circuit in a given time. Again, this depends on things like temperature, internal resistance of the cell, resistance of the overall circuit, and, of course, the proportion of photons that are creating free electrons in the first place.

Since both current and voltage can fluctuate quite regularly during normal operation, the power must therefore also fluctuate. Current and voltage are related to each other, so that if one changes the other will also change. In a solar cell this relationship is exponential – but the exact nature of that exponential relationship will change depending on factors such as solar intensity and even the unique characteristics of individual cells. This is more clearly expressed in graphical form:

It’s worth noting, from looking at the above graph, the specific relationship between current and voltage. At first, as current decreases gradually, voltage rises rapidly, so I*V (power) will also increase. However, once a certain point is reached, the current decreases more rapidly with small voltage increases. At the point indicated by the black line, the decrease in current outweighs the increase in voltage and the power value no longer increases. In other words, the black line marks the point of maximum power (aha!)

The upshot of all this is:

• Current and voltage vary in quite complicated but predictable ways
• Power generated by the cell therefore also varies, such that it has a maximum possible value.

The question is: can we manipulate the current and/or voltage in response to these variations, such that we are always getting the maximum possible power?

Yes we can! This is what a Maximum Power Point Tracker does. Control circuits within the device allow it to work out the maximum possible power, and what values of current and voltage are required. Both voltage and current are related to a third factor – resistance. By changing the resistance of the circuit through which the current travels, both voltage and current values can be adjusted. An MPPT is capable of artificially manipulating the resistance in order to control the values of voltage and current.

The function of the MPPT is therefore twofold:

• By measuring instantaneous values of current and voltage, it calculates the relationship between them and works out the optimal values of V and I to gain maximum power
• It imposes an electrical load resistance on the solar array, to alter the values of V (and therefore I) to these optimal values, thus ensuring that the maximum possible power is always extracted.

Because the I-V relationships of different solar cells/panels/modules vary, it is sensible to have a large number of MPPTs, each tailored to an individual module of cells. This is what happens on Endeavour – with five separate solar modules, it contains five MPPTs to monitor each one:

In a system such as that used in Endeavour, whereby the solar array feeds a battery, there is a danger that the battery can become ‘overcharged’. Using very unscientific terms:

it’s as though the battery is full up with energy, and can’t accommodate any more. Since the solar array can’t just stop producing energy, the MPPT must exactly match supply with demand so that energy is added to the battery at the same rate it is being fed to the motor. In this way the battery does not become overloaded.

This mode of functioning requires a diversion away from the maximum power point, but the MPPT still has the capabilities to do the job.

An MPPT can also provide additional advantages when used to drive an electric motor directly. Because they can adjust the voltage of the cells, they can be used like a transmission. When starting a vehicle, a high torque (low gear) is required to get it going, corresponding to a low voltage/high current situation. The MPPTs can do this, then raise the voltage to normal levels once the car has got going.

And there you have it: MPPT. Maximum Power Point Tracker.

Does exactly what it says on the tin.

——————–

*After ‘CUER’, that is.

The basic premise behind a solar cell is identical to that behind any type of chemical cell or battery – the separation of regions of different electrical potential. An electric current flows when these regions are connected to each other by an electrical circuit, allowing the negatively-charged electrons to flow towards the positive terminal. In a solar cell, this is achieved by the use of semiconductor materials – a certain class of non-metals that, under certain conditions, can conduct electricity.

The most common solar cells use silicon as their base material.

Silicon has a valence of four – four electrons in its outer shell. It can create a bond using each electron, so there are four possible bonds available. In its pure form, therefore, silicon atoms can bond with each other to create a single crystal structure in which every electron is used up in a bond. However, when silicon is doped with impurities – other atoms such as phosphorous or boron – that have different numbers of outer electrons (5 for P, 3 for B), things change.

Phosphorous forms four bonds with silicon but the fifth electron cannot be used, so with the addition of energy (e.g. heat) these extra electrons can be knocked loose of their atoms. Once this happens we have a source of free-moving charged particles capable of carrying an electrical current. The phosphorous atoms left behind become positively charged ions*.

Phosphorous-doped silicon is an n-type material – ‘negative’ – because of this oversupply of electrons. Boron-doped silicon is p-type – ‘positive’**. Boron, with only three outer electrons, cannot form the fourth bond with silicon. There is an electron deficit, also known as a ‘hole’. When an extra electron is provided to form that bond (resulting in an ‘electron-hole pair’), a negative ion is formed.

When n- and p-type materials come into contact, the excess n-side electrons move towards the p-side to bond with the holes. The electrons/holes closest to the boundary are the first to form pairs. However, once this region around the n/p boundary accumulates electron-hole pairs, it becomes more difficult for electrons to cross that boundary and reach the free holes on the other side. This is because the electrons that bond with holes on the p-side form negatively charged ions, and leave behind positively-charged ions on the n-side. The p-side ions repel electrons, and the n-side attracts them.

The result of this is an electric field that opposes the initial movement of electrons, so that electrons can only flow from p to n and not the other way about. This zone  is known as the depletion zone, and acts as a ‘barrier’ separating the n- and p-sides.

The importance of photons

Where does light come in? Light is made up of photons – individual light particles. Each photon has a fixed wavelength, and a fixed amount of energy available that corresponds to that wavelength. Photons can be absorbed by electrons and impart that energy to them. If the energy is great enough, the electrons can break free of their bonds.

These electrons may then be influenced by the electric field of the depletion zone and travel to the n-side of the material, while the ‘hole’ remains on the p-side. As a result there is a build-up of electrons on the n-side and an excess of holes on the p-side – a potential difference exists between the two sides. When an external current path is provided (i.e. a circuit linking n to p) the electrons flow through the circuit to the p-side, driven by this voltage.

This is where the electric current comes from. The photons absorbed by the solar cell constantly produce free electrons that then travel through the circuit to the p-side, where they may form an electron-hole pair before once again being liberated by the absorption of a photon. In this way a solar cell provides a potentially unlimited supply of energy.

Efficiency

The problem with this basic silicon cell layout is one of efficiency. Sunlight is made up of a range of wavelengths and therefore contains a variety of protons with different energies. A very specific amount of energy is required to release an electron from its bonds. This energy is different for every material and is known as the band gap energy for that material.

The second important point is that a photon represents a quantum of energy – a single packet that cannot be divided up – so an electron has no choice but to absorb the entire energy of a photon. Certain photons won’t have enough energy to free an electron. Any photons with insufficient energy have no useful impact. They pass through the solar cell. Those photons with an energy greater than the band gap do release electrons, but the extra energy is imparted to the electron as kinetic energy, which rapidly becomes wasted as heat.

These effects together account for around 70% of energy loss.

Using a material with a lower band gap would mean that fewer photons are totally wasted. Unfortunately, this has the result of reducing the voltage of the cell.

There are also problems connecting the external circuit to the cell since silicon is not a good conductor. We want to make it as easy as possible for the free electrons in the n-side to reach the electrical contacts and flow through our circuit to the p-side. This is done using a metallic contact grid. Nevertheless the resistance of the semiconductor material (through which the electrons must travel  to reach the contacts) and the presence of the grid blocking out some photons lead to extra losses.

Finally, not all electrons released from their bonds may travel to the n-side – they may recombine with another hole on the p-side – in which case the energy of the photon is wasted.

Multijunction Cells

The World Solar Challenge has two sub-categories of vehicle corresponding to the type of solar array they use. Vehicles using silicon cell technology (such as Endeavour) are evaluated separately from those using Gallium Arsenide cells or similar (such as Nuna). This is because GaAs cells are multijunction cells, and are therefore much more efficient.

As mentioned above, one problem with solar cells is that they have a specific band gap – a fixed energy required from photons to release electrons. This causes a lot of photon energy to be wasted, if photons have too much or too little energy. A multijunction cell attempts to remedy this by using two or three different semiconductor materials, all with different band gaps. They are layered on top of each other, with the highest band gap at the top (so that photons with lower energies can pass through and be absorbed further down).

The overall result of this is that multijunction cells can generate electrical energy from a greater range of wavelengths of light – fewer photons are wasted – so they are much more efficient. However, due to difficulties with manufacturing, they are also much more expensive, about 100 times more so than silicon.

Because of this, the WSC awards a prize for Best Silicon Cell Finish, since it recognises that these vehicles are at a disadvantage compared with multijunction cell users. However, in the 2011 race, the rules have been changed to reflect the disparity between Si and GaAs cells. Any car using GaAs cells will be restricted to half the array area of a silicon car.

Solar Concentrators – the importance of intensity

Intensity of light is a measure of how many photons strike a given area of cells in a given amount of time. In a solar cell, the number of electrons available to do electrical work is proportional to the number of photons absorbed by the cell. A higher intensity of light will produce a bigger current, so the aim is to gather as many photons per second as possible. This can be done by increasing the total area of the solar array – but this is restricted in the WSC.

Instead, a solar concentrator may be used. The Michigan solar car Continuum is a particularly good example of this. Parabolic reflectors collect light and focus it onto a much smaller area – increasing the intensity by a huge amount.

The alterations in the WSC regulations mean that some teams may decide to keep a 6m^2  car, comprising 3m^2 of concentrators and 3m^2 of gallium arsenide cells.

****

——————————————————————

*An ion is a charged particle. A phosphorous atom is electrically neutral, since it has fifteen protons in its nucleus, each with a positive charge, and fifteen electrons in orbitals around it, each with a negative charge. When an electron is lost from phosphorous in a solar cell, these charges no longer balance out and it has an extra positive charge, turning it into an ion. The equivalent happens when boron gains the extra electron, becoming a negative ion.

**It’s important to note that these materials don’t actually have an overall charge initially. The ‘n’ and ‘p’ simply refer to the fact that one has too many electrons to form bonds, the other too few. It’s very misleading. This is why I quit electronics.

———————————————————————–

Image Credits

1. Electronics Tutorials
2. Images Scientific Instruments
3. The Future of Things

Panic Level: 100-125 bpm

This week, CUER SunSpot takes you on a rather dull extremely fascinating journey inside the world of electrical and information engineering - a world where phases, magnetic fields, currents, voltages and, ultimately, sharp corners* cause a multitude of things to go nastily wrong. Or, alternatively, wonderfully right. But, this being problem-solving week, the probability that there is going to be a blog post about everything working absolutely fine is somewhere down there with you getting to see that live Michael Jackson concert next month…

Too soon? But seriously - today’s post is as a homage to that artist, who himself loved things that go round and round, such as ferris wheels, merry-go-rounds, vinyul records, CDs, occasionally himself and even questions about his sad and untimely demise. Today, dear reader, we learn about the Permanent Magnet Brushless DC Electric Motor, or PMBDCEM ‘motor’ for short.

Exploded view of motor. Please note that motor does not actually explode. Except in very unfortunate cases.

The motor casing is adapted from the Aurora design, with CSIRO electromagnetics. The basic principle (for the technically minded) is that the motor contains a toroidal coil of wire (the stator) and a ring of very strong permanent magnets (the rotor). The rotor and stator can move independently of one another. The electricity flows through the stator in three phases from the motor controller. The motor controller uses Hall effect sensors to track the position of the rotor and act accordingly. The resulting changing magnetic field induced by the flow of electricity in phases through the coil (remember school physics?) interacts with the permanent magnetic field and causes the rotor to spin. Thus the motor goes round and round.

The basic principle (for dummies) is that the electricity goes in, and the motor goes round and round.

The basic principle (for Harry Potter fans) is that the motor goes round and round by magic.

The basic principle (for the very religious/spiritual) is that the motor goes round and round by the power of faith and belief. Our current theory is that, when it worked first time, none of us could actually believe it. Then it broke down.  

The electrical team have been having some issues with the motor ever since the axle holding the stator sheared free of its holding plate and began to spin, taking the really-shouldn’t-spin-at-all stator with it. This caused the cables carrying the three phases into the stator to twist around each other as well. (By the way, the motor was running at the time. We’re not looking at spontaneous rotation or anything. Though that would have made a much more interesting post.)

The (by now) standard response to this incident was to go ‘oh no’, write it down, and then do nothing much about it. Until recently – when Jonathan finally put his foot down on Charlie’s skull and threatened to crush it and demanded politely requested that he be allowed to take it apart. This was duly done – yesterday, in fact – and CUER SunSpot was allowed a sneak peek inside…

The motor is a pretty delicate and heavy bit of kit. The magnets inside it are so strong that it needs a special separator tool just to prise the casing apart. After much careful bashing with pliers and screwdrivers, the bearing cover came off, only to fail to fit over the connectors on the ends of the cables. An inordinate amount of time was spent filing down said connectors before the cover could be removed. Many thanks goes to Michelin for this unorthodox use of their tyre soap.

Eventually it came apart and the extent of the damage was fully realised. It didn’t help that some idiot Jonathan had decided to stuff the axle (through which the cables passed) with plasticine to hold things in place. This had horribly backfired. Some green and gooey minutes later, the cables were finally untwisted, removed from the motor and could be examined more closely. As Jonathan had predicted (via some nifty electronics trick), Phase 1 was effectively grounded – the twisting had caused the heat shrink and about half the copper inside to be cut straight through by the edge of the axle.

As it turned out, Phases 2 and 3 were similarly damaged, although not to the same extent. As an interim measure, the damaged lengths of cable were cut off, some new lengths stripped and covered in heat shrink – although this is not enough to protect it in Australia. It will be properly wrapped in some abrasion-resistant stuff by then. The motor was then put back together and some torque tests run…

Unfortunately, the team then encountered further oddities within the motor controller itself, which will be elucidated in these pages at another time (i.e. when they’ve figured it out).

This was a useful learning experience, which hopefully will feed back into the third generation design. Current thinking is that, since the axle doesn’t need to turn, it’s really quite silly to make it circular, which then needs to be held in place via a friction fit. A friction fit, in fact, that requires more precision than we can currently provide. There are also several things inside the motor that can be tweaked so that the wiring is more resistant both to heat and to abrasion. SunSpot suggested that they should be made resistant to faliure, but got fobbed off with some excuse about being too vague…

Still – progress was made. Lessons were learnt. Snacks were had in abundance.

We leave you with this apt quote from Jonathan himself, which rather sums the situation up nicely:

“I wish I’d done aero instead.”

Sorry. This is the one we meant:

“I knew there was a possibility that it wouldn’t survive Australia; I didn’t expect that it wouldn’t survive Cambridge!”

Until next time, then, eco-fans!

*Well, very ultimately, it’s all electrons and quarks and gluons and the three fundamental forces that actually DO stuff to them, and so forth. At least, that’s what it is at the moment. We’re hoping to get an RSS feed from the LHC as soon as it actually works, which will keep us all up to date on how many elementary particles/dimensions there are currently making up our universe.

Chao says "mess with it and die". We think he's probably right...

So far the project has progressed smoothly with few delays, thanks in no small part to the contributions by REAPsystems Ltd and Farnell UK Ltd. REAPsystems are supporting the project by donating their battery management systems (BMS), which will be used to monitor and control the operation of the Lithium-polymer based pack for optimal performance and reliability, as well as supplying the team with additional components and design expertise. Farnell UK have kindly donated electronic components and tools to CUER, which has been beneficial not only for the battery pack, but also for the rest of the electrical team’s projects.

- Anthony