Brickthus

You could fit an axle and crank to the side axle hole of the switch, with the crank on the lower side of the axle, and pull it downwards with a belt. That would spring it back to the centre. Alternatively, with the crank on the upper side, push upwards from below with a suspension spring. Mark

Anyone who might wish for a LEGO jet engine kit with Technic functions should take a look at a new LEGO Idea that I posted in The Embassy LEGO Ideas discussion. A new Rolls-Royce jet engine with cut-away sections and internal gearbox has been submitted. It is already attracting significant support after 24 hours and should soon reach 1000 votes. It can be motorised; there is a video of it on Twitter (search for @cakesmyth) Mark

A new LEGO Idea for a Rolls-Royce jet engine with cut-away sections has been submitted. This would make a great STEM tool for education as well as a building challenge for even experienced adult LEGO fans. It includes an epicyclic gearbox at the front, which slows down the fan and speeds up the turbines, which is more efficient on the real engine and adds to the Technic functions in the model. I'd be as proud to complete such a model as the real engineers are with making the real engines. Mark

Sturdy is very important, both for a set to be played with by children (build quality) and for exhibition reliability. 45 metres of curves cost around £380 for the three pieces in quantity. Still a significant cost but manageable for a circle of curves in a set. It works out at £26 for the curves but cheaper for the straights, so with a train motor, IR receiver, AAA battery box and train handset it should meet the train-set price point. Plus there is extra value as you can make a double-bend as a second-model variation. I found that 19 2x2 sections fits on a 48-stud radius for the larger baseplates. This is the minimum feasible radius without straining the bricks (which would be an "illegal move"). The tighter the radius, the more strain it will put on the train's motor(s). I did consider the 1-3-1-3 building technique but it is less adaptable to curves because the track is 2-studs wide. The gaps would be bigger, leading to bumps. I avoided using under-studs because those would wear out the tyres quicker. I tried the original idea of tiny tyres (2 lots of 3 on Technic half-bushes under a 14-stud long carriage) but switched to the 18mm tyres because they grip well and are common in City and Classic themes. Then I found that 23 sections fits a radius of 58, 27 fits 68 and 31 fits 78. Beyond that a small adjustment is needed, such as a 3-long section. The aim was to make the original free-form curves fit a regular grid system so that I could make a set like the 1990s monorail curves but with multiple radii that has been a craving of train and monorail enthusiasts alike for many years. 48M minimum radius with modulo-16 straight lengths, but reserving the capability to go down to modulo-2 or 1 for special shapes, seems to work well. A double-bend, to slew a track across by 10, is like a 32M straight made with curve sections but it needs an extra 1 in the middle. There are no restrictions on track shapes using patterns derived from the curve and straight construction techniques. That makes it cheaper than 3D printing (re-use the bricks) and also unrestricted for the user (not waiting for TLG to make a new mould like PF railway track; they said it was more feasible to make track shapes in just plastic but we have had only the short-lived crossover and the flexi-track with check-rails). 5-plates high is tidy and allows the train guide pins to clear the adjacent studs of a brick used to support the track, or even a hoop or tunnel fixed underneath the track. Consider also that a child would like the smallest cross-section of track for easy holding in smaller hands. The hills have to have a deeper profile at the slope-change places in order to get enough strength. I have used 2 clips and a push contact at those joints, rather than just a single clip that would be flimsy. This kept up the product focus. The aim is to be able to replace the original monorail in form and function and have the system never go obsolete. Obsolescence hurts AFOLs in the pocket. LEGO products need to stick to the brick and avoid being too bespoke. Mark

Thanks David. I would just say that, in the instructions for your 16-stud straight track piece, it needs the 16-stud picture, not the 32-stud picture at the end. I have made successful straights of 48 and 64 studs without needing intermediate support. That is useful for joining and leaving a helix. I see the potential for automated control but for now I prefer a train I can control with a PF handset and PF point control with a switch panel, which is more like a signal box of a railway. At an exhibition it is best to leave trains running and talk to people. The monorail curves have a tendency to bank outwards slightly. I found the same tendency with brick-built curves from the Hidaka original design to my own experiments, as well as plate-built ones. This is OK if you build the trains with a low centre of gravity. With careful train design I have had no problems at full speed but the battery box and motor are close to the track. Even an XL-motor might be quite tall if it were to go fast. How fast does your train go? You could measure over a number of laps of a known track length and convert to metres per second. I would prefer to have more than 8 supports per circle of track because of the potential to sag between them. The train has enough power to cope but the "wow wow wow" sound might be annoying. The 4-track quarter-circle section with 16 supports per circle is better as it stays flat and without any sag. The 1x2 plate connections on the curves (1 stud overlap) are nothing like as strong as the 2x8 connections on the straights (2x4 overlap). My helical track has the potential to do the same as the tall monorail helix at the show. I wanted to prove that early, to make it worthwhile for record track lengths in order to convince those who chase such records to have a go. I have enough parts for 45 metres of curves so far; it is easy enough to buy just 3 pieces in large amounts. For gaining height I would like to do it with up to 16 supports per circle; not quite so easy to fit the geometry for all of them with a stud fix to the baseplates. I can do multiple-radius helixes but more than 1 track would need outer supports too. In reality that means a circular canyon of scenery and/or some tunnels for part of it. To make the trains go in and out of mountains is more interesting. Mark

The 2-wide brick-built monorail system is capable of hills of certain angle. Masao Hidaka made some quite steep hills but I prefer to limit the angle to a more modest and robust template, which is capable of rising by 5 bricks high in 48 studs long, or by 10 bricks high in 80 studs long (a bit longer for the same height change of the 64 stud length of the 1990s monorail hill parts). The template can extend to any length of hill with the same slope-change pieces at the ends. From Masao Hidaka's original free-form track ideas I systemised it to do all the functions of the 1990s monorail, and to fit on a grid. I even made the direction switch function work, though it is obsolete with Power Functions having IR remote control. I added a right-angle crossing and multi-point junction. I have 4 radii of curves spaced 10-studs apart (centres at 48, 58, 68 and 78 studs radius) with trains from 6-wide up to 8-wide. I have done a 2-track helix to make sure that space-saving technique can be used to facilitate world-record track lengths (562m to beat the 1990s monorail record). I also made the trains faster; with belt drive it did 0.95m/s, almost as fast as L-gauge trains (measured over 10 laps of a 12-metre track including helixes). I have slowed it down to 0.57m/s for heavier trains to climb with good reliability. The belt-drive concept ran for 3028m on one charge of the PF LiPo battery. My philosophy is that each train should have half its wheels powered. A 1-motor train can have a battery car and at least 1 more car. A 2-motor train (with the battery between them) should be able to have 4 further cars; I have tested 3 successfully ad I think there is scope for more to work. The design should bias the adhesion weight over the driving wheels by making the extra cars lighter than the powered cars. Also for bidirectional running, motors should be at the two ends of the train. I did this with my gearmotor protoype that was my first step beyond Masao Hidaka's original work. My shared-bogie system was a second step. The belt drive, use of the train motor and new shaft drive are further steps. It also has flashing lights of my own design! Masao Hidaka ran his Osaka Monorail prototype as a LEGO Ideas project and got 1707 supporters. I have a complete system with a set theme and would be prepared to do a project from my own work but would it be worth it unless we could get to 10,000 supporters? I have considered many factors and learnt a lot from my BB-8 project that ran for 18 months and took about 500 hours of building work plus a lot of promotion work on social media. That did OK at 1327 supporters but I felt like I had to fight for every one of them. With that one I didn't have an optimised technical architecture because I had to launch quickly as the film was out but there was no LEGO BB-8 set. For the monorail I have been building over 5 years so I have ironed out most of the technical issues, such that I plan to exhibit a layout next year in the UK once I have built some layout scenery. I you would like me to submit what I have as a project then I would be happy to do so; a key question is "would it make a good LEGO set for the original age group of the 1990s monorail, as well as for TFOLs and AFOLs?". I would like to create a bit more space in what should be my free time first, so that I could spend enough time on it to do it justice. I think a project needs 1000 supporters in the first month if it is to reach 10,000; BB-8 got about 260 int he first month and that has Star Wars groups available to attract support from outside the LEGO fan community. That is a second key question: "Could we attract interest from beyond the AFOL community?". If not, then that might seriously limit the potential support base. We would have a lot of convincing to do of traditional 1990s monorail devotees; they even have 3D-printed pieces becoming available as a product soon - images exist of the prototypes. They still have the snag of keeping those motors going, and (for the time being) of unreliable points. A further 3D-printed product might be possible to fix that in the future but it was one of my reasons for looking for another monorail system. Other reasons include obsolescence of track, motor and train parts, cost of 2nd hand parts, limited power supply with a PP3 battery, slow speed, single curve radius, 4-wide product that struggles to do 6-wide, and limited scope for points and crossings (TLG filed patents for a Y-point and a right-angle crossing). I think I blasted most of those limitations with the brick-built 2-wide system! I am encouraged that djm (David) has written some instructions for my straights and curves and the Hidaka points - major thanks. David, if you are OK with me linking to them from a prospective LEGO Ideas project then that is another factor that might be persuasive because it means people can have a go. I will reply to your message when I get chance. Mark

Thanks David. The 4-track switch in the linked Brickshelf folder has 3 positions: all-straight, all left and all-right. These correspond to the 3 positions of the servo motor when operated from a switch or a bang-bang remote control with IR receiver. It means a single £21 servo motor can control 8 switch points in a junction, so it saves on cost. Most other junctions are a sub-set of those 8 switch points, sometimes 2 sets, one at either end of a station. In particular a through-terminus has an island platform with a reversing track between the running tracks at one end. That would use three of the eight switch points with the servo motor. The geometry is adapted from Masao Hidaka's design. I used multiple sets of the geometry, supported the middle and made it so that two switch rails can come towards each other, rather than a single point having fixed destination tracks. The switch points on the outer tracks use belts to spring them to the straight position because they have only 2 positions. The straight position of each one corresponds to 2 of the 3 servo motor positions. I might make an improvement by fixing the round bricks and log bricks in the middle more securely for long-term running. This would add axles from underneath to manage the collision forces from the monorail train guide rollers. In the meantime an occasional finger-press suffices. Mark

In the monorail system that I have been developing on and off since 2013 (from the Hidaka monorail that uses a City prototype of the Osaka monorail), I have built a Classic Space monorail train. The train has a new drive-train and is geared down 5:3 so that the heavier train will climb well. It is still based on a train motor but I stopped using belt drive. The average speed is up to 0.57m/s. The lights on the right-hand axle will flash when the train moves. This uses the other channel of the IR Receiver. The battery box (AAAs or LiPo) is behind the Space logo and is easy to change. There is a removable module system. The module on the train unfolds to become a scooter. There is a medical stretcher module and a power module on the ground. It has proven to be reliable on my loft layout. I will do a video in due course. I need to get the lighting right as it was too dark last time I did a video in the loft. As well as a second Classic Space train and a 2-motor goods train with extra wagons, I have prototypes for Futuron, Blacktron and M-Tron trains. More pics of the system here Mark

So we have seen a car propelled by a LEGO engine, powered by a 3rd party air supply. Impressive as the largest LEGO pneumatic engine so far. We have also seen an RC plane whose flight surfaces are the smooth side of LEGO plates. Shame it uses tape for the hinges and falls apart on landing, but it's a step in the right direction. Good to see some LEGO parts flying. They do the same at LEGOLand, making a metal train chassis, putting a LEGO sculpture of the loco body on top and calling it a "LEGO train". I'm pleased to see some of the [TC10] entries are pure-LEGO pneumatic trains. The current status of a pure-unmodified-LEGO heli-spinner launch system is that, after a few LEGO flight experiments, I managed to launch a Ninjago Airjitzu spinner (without minifig or pod) by LEGO battery and motor power alone, using 2x PF train motors geared up 5:1. An earlier test using 2x M-motors was nearly there but not quite. The spinner rose to just an inch above the launcher, compared to an arm pull sending a spinner with minifig and pod up to the ceiling. This is the limit of LEGO motor power. Also 800mA at 7.4V per LiPo battery is the limit of power-to-weight ratio. The current status of a pure-LEGO helicopter with 3rd-party power is 200g of lift from a 270g model, not yet off the ground. This used 2x 5292 motors, which did not take long to warm up, and a bench power supply set at 9.15V and drawing about 2.6 amps. I suppose more is possible. So how do we move forward with a working pure-unmodified-LEGO aircraft? A fixed-wing aircraft does not use as much thrust as a helicopter in order to maintain flight, so the LEGO baseplate wings are a good idea. This puts the LEGO flight challenge in league with the generic drone challenge - how to get long range out of an RC drone. It is suggested that it could do VTOL as a helicopter but then transform into a fixed-wing aircraft as an Osprey does. The extra hardware to transform would push the weight over the limit for a LEGO aircraft, so a fixed wing aircraft is a good aspiration. The trick will be using few-enough LEGO servos or micro-motors to get the weight below the lift capacity. 1x LiPo, 2x IR Receiver (or more likely S-Brick for range), propeller motors and a few 9V micro-motors. Worth a try. Mark

Very nice. I would buy a set like this if it were developed. I would hope the 2nd model would do something automatic like 8868. Mark

That's close to being a set I would buy in multiple. Well done. Mark

Shame about the accelerator, it looks like a fun set that might be played with for longer than average. Not so many Power Functions that it would be unviable; quite a few sets have a motor and battery box but it does skew the price a bit. I see your points about the obsolete rails and the ball energy; would the latter rule out a GBC? If they did a GBC set like the train set with several different models to make, all of which conformed to the standard for passing on the balls, then I'd buy quite a few to make a circuit. Interesting GBC modules take a lot of designing. Also a shame about the Gingerbread house, perhaps my favourite of that bunch but I accept the points about TLG making their own smaller one. I wish it were not the case that so many sets have to be symbolic, giving just a cut-down flavour of something without doing it properly! If something can be done properly in a moderate-sized set then that has more chance as a LEGO Ideas set. Even the Caterham Seven is missing its steering! Magnet safety (and obsolescence) is an issue I also tackled for my larger BB-8, that and a different potential high-energy-debris issue of releasing weight blocks from a spinning counterweight! The small BB-8 was OK functionally, using train magnets with a thin shell and a single direction of movement (no turning or spinning) but a larger model needs a thicker shell (given no ready-made large sphere LEGO piece) so the old train magnets would not be strong enough. I tried some 12mm Neodymium magnets, safely encased in City dustbins with Technic cogs inserted in the top, which also provide an axle to hold it, but even that had a limited attraction distance. I might use the technique elsewhere for a display model sometime. In the end I decided to try a split-shell method for BB-8. After many trials, BB-8 finally works with his head on top and can roll, turn and spin . The head movement, "speech" light and deployable tools can be transferred from the static model in due course. This means the whole model would meet the functions-to-price ratio of Technic sets. Still within the 3000-piece limit and no modified pieces! I would like to give TLG the opportunity to accept or reject it as an example of making the best of Power Functions; it follows on from Volvo Loader 42030 and may help Technic to branch out from the usual vehicle selection. I think TLG should promote the capability of Power Functions more (whether PF1 or the forthcoming PF2). Mark

Glad to inspire you! About time more of us did more interesting pneumatic models. Maybe one day TLG will revisit sequencers; it's been a long time since the refuse truck of 8868 in 1991. According to the systems I have worked on, the way I can think of doing the sequence you need is two 2-function loops interacting. The up/down and a dummy function are one loop and the in/out and another dummy function are the other loop. Each loop does 3 of its 4 operations before enabling the other loop, so down, dummy, up or out, dummy, in. Probably 4 cylinders and 6 switches minimum. Perhaps if I had been reading here more regularly and not also building BB-8, which now works , I might have had a go but it's great to see you guys picking up on sequencers. I'm glad to help anyone to learn more about it. Most of what I knew about mechanical engineering until recently is what I learnt from LEGO building; LEGO, along with electronics kits and a bit of programming, made me an engineer. BTW what is the source of buying just the pneumatics bag from 42043? I could do with a few without so many of the other parts - only so many wheels needed for a truck that will fit in the house! Thanks, Mark

Happy to help I would usually use 25psi for pneumatic models. I had a car tyre air compressor that would produce a continuous 25psi, till I broke it . The practical limit for LEGO pneumatic air supply by hand is 6 hand pumps at 1 pump per second, which I used in my Pick and Place robot. 2 pumps per second can be done for a short time! The experiment above was to show a) the power of the 1/11 cylinder and b) the amount of air needed by the 2/11 cylinders with the consequent pressure built up in the system. Things were easier when we had just 2/6 cylinders and switch valves were not quite so stiff. Practical models would make use of the twin 2/11 cylinders to move a heavy function and the 1/11 (or 1/5) cylinder for logic only, to move a switch but nothing else. Mark

Sorry I've not been around for a while - been too busy building BB-8! Some may be 9V and others may be 12V. Research into data sheets will reveal the ratings of each one. Philo tests LEGO motors at up to 12V. He may know more. 9.15V is possible for any 9V motor (without research) because that is the voltage of a fresh set of batteries, which is the powering method recommended by TLG. If the motor is stalling then you will need an additional motor. I would suggest either using two 5292 motors, or gearing down by 3:1. If you use two motors, you might need a battery box and IR Receiver for each one. This would effectively put two battery boxes in parallel but only once the electrical power has been converted into mechanical power. I have done the same for my BB-8 Droid. In the update I will post soon, I used 1 XL motor for each half of the spherical shell, driven from a single IR Receiver and a LiPo battery. This allowed BB-8 to move a bit, up to 1 metre in each direction, but after a few tries a V1 Receiver stalled and a swapped-in V2 Receiver complained of overcurrent/overheating by making a squeak/whistle sound. In reality I had tripped the LiPo battery's 800mA overcurrent trip. I changed to using 2x XL motors each side, driving one from each of two IR Receivers set to the same channel and using the corresponding red/blue output of each receiver for each side, with a LiPo battery for each IR Receiver. This has doubled the current capacity whilst the power from the two batteries comes together only as the motors are linked mechanically; there is no electrical connection between them. BB-8 now moves with his head on and can drive, turn and spin without tripping any electrical limits - nearly time to make a video! A similar system is used on trains, where two PF locomotives may be set to the same IR channel and the power only comes together at the rail. You should not try to unify the driving components electrically but you should use a motor on each of two electric drive systems and unify the power mechanically. Then make sure the two battery sources have similar age batteries to get the same state of charge and minimise one motor pushing the other. Mark

I prefer a mixture. Helicopter 8856 was studded and rigid but too chunky for it, though some of its studless parts were great, such as the rotor pieces. Helicopter 42052 is studless and gets its rigidity from panels. Panels are expensive so you have to justify using them as a significant proportion of the cost of the set. A real weakness of studless is that a vehicle chassis is too flexible, The New Porsche uses more frames to reduce this, but that commits a model to be at least 5x7 in cross section. I found the best mix is a studded chassis with studless attachments. This can happen more in other themes, where the model is predominantly studded, with a few Technic parts added, using studded beams as the interface. I do that quite a lot for steam trains with Technic driving gear. However, I haven't yet tried a studded chassis with panels over the top. Aesthetics might drive the whole model to odd-number sizes, which favour studless throughout. For a mixture I really like this piece as a stud-to-Technic interface. Great for attaching Technic or rod-based roof ties in a large building. Mark

2 methods come to mind: 1. You could make a 3-ring single motor mechanism from the new Porsche gearbox parts. The Porsche uses 2 driving rings with a new smooth red axle extender in the middle, allowing the shift to be in any position, not restricted to 3 datum points as the previous white piece does. If you add the outer sliding piece from another set and use another red axle extender and 2 more of the red cogs from the Porsche, you can make the 3 rings. Then you have to arrange for them to be actuated with cranks at 120-degrees from each other; using 24mm pulleys like this should help (minus the pneumatic parts). The test is whether it would be OK to have 2 gears engaged at once in the rotation cycle. The Porsche avoids this because its gear selection cranks are at 90 degrees to each other, putting the non-engaged rings in the middle. With 3 rings on one crankshaft, the non-engaged ones might be very close to engagement if not engaged permanently. I suspect 3 rings on one crankshaft might be useful if it's a multi-function gear selector but would not work for a vehicle gearbox. For the multi-function selector fr 42009 it would be important whether any of the functions needed a lot of torque if its driving ring were not fully engaged. 2. A linear function selector gearbox My rail crane uses a 4-gear linear function selector gearbox for its 4 functions. One M-motor drives and the other selects. The selection motor uses a white clutch gear to cope with the end-stops, with a worm moving the selector along the axle. It can stop the selector in a mid-position and actuate two functions at once, so it can select a jib raising movement and a hook lowering movement at the same time for luffing operations. It uses M-motors for both selection and drive because 3 of the 4 functions are worm-driven so that they stay put when not selected. The jib turn function has to be free-running when not selected so that it can move along the rails. This pic shows one of the hook worm drives and the 4-cogs-in-a-row input shaft to the gearbox. The 20-tooth selector gear slides along the line. My shunting loco has a similar gearbox for 3 functions (fast speed, slow speed and decoupling). The shunting loco uses an M-motor for selection and an L-motor for drive to get plenty of torque. This will shift a huge train slowly, just like the real thing. Inside the skin of tiles, the LiPo battery and IR Receiver are at the front, the two motors behind that, one on top of the other, and the gearbox fills the cab. I suspect you would use an M-motor for selection and the L-motor for drive in 42009. The jib extension and hook raise/lower functions each have a worm drive already and the jib raise/lower uses Linear Actuators that contain worms so they will all stay put when not selected. With a linear gearbox like this it is theoretically possible to use some 32M axles to have 28 functions and also select adjacent pairs! Mark

As I post updates (now 14 on BB-8) I get a lot of likes on social media, but turning them into supporters is a black art. This is contrary to most projects that are complete as a model (or a virtual design) before initial submission. After 5 months I have over 800 supporters but I would not resubmit before reaching 1000. The snag is, I've done most of the outer shell where people can see a difference. The harder stuff is the head-holding function and the spinning function, both of which are inside and less photogenic, making it harder to gain more support. Even the melting of a piece or two didn't excite people! The project has spawned a few ideas that extend beyond itself, which I'm pleased to offer the community: - The magnet encasement in minifig dustbins as a means of preventing swallowing, potentially leading to a more versatile safe use of magnets for other themes. - Technic models beyond the usual vehicles; BB-8 is based on a sphere but I would like to see moving animals too. The ideas book 8888 featured a motorized dog. - Spinning counterweights. Spinning things fast is fun but I've found the limit of what the LEGO system will stand and it's another thing I have to make safe in order to suggest it in an Ideas project. As links to the project on the LEGO Ideas site are discouraged, here's the MOC picture gallery. Mark

In Alternative 1 it is the gap between the trails that matters, not their centres. The gap is 37.8mm, which equates to a standard gauge of 4'8.5" in 8mm scale. This makes the scale accuracy better than 'OO' and nearly as good as 'HO' if you scale from drawings of other models. I have a book of 4mm scale drawings for UK locos. The minifigs can have 1 or 2 plates added to their short legs to improve their height to width ratio. There are also some longer legs from Woody in the Toy Story sets. I use mostly an 8-wide base for UK protoypes with the ability to add multiples of 0.2M either side (4.8 inches at a time) where the book says the loco is wider, but I would build most vehicles with a 10-wide base for US prototypes. Mark

I found a local store that has magnets in 3 sizes. I tried the medium size, which is about 12mm diameter and 3mm thick. Seems to be the right diameter to put between the old-style couplers. I needed them for my BB-8 droid, to hold the head on, but I'll get a few extra ones for trains. I use Technic half beams in rakes of wagons to save on couplings, since the versatile ones are obsolete. For safety, strong magnets should be encased to prevent kids swallowing them so I found a smaller solution than the current buffer beam. Mark

I want a buildable and driveable LEGO BB-8 droid, badly enough that I have had a go at making one. So far, after 4 months and 200 hours of building, I have a half-size sphere that will roll and turn with about 0.5m radius under remote control. The sphere colour scheme is looking OK and many of the tools can be deployed manually - "thumbs up", panels, arms and tubes. The head has a new safe but versatile magnet encasement scheme as the first stage towards keeping the head on whilst rolling. I will work on the spinning function and keep pushing forward on the head attachment. So far the price point for a set would be £170-200, a bit like 42030 Volvo Loader with a similar number of motorized functions. I think the LEGO system is capable of this and we shouldn't let Sphero etc, take all the glory! If you would like to see a buildable and driveable LEGO BB-8 droid then . If you would like a new safe but versatile magnet piece for another project then please also support the project for that reason. Thanks, Mark Edit by Hinckley: Removed link. Please read the Site Guidelines. Thank you.

I have a solution to the magnet issue, a safe but versatile magnet encasement scheme, smaller than the cumbersome buffer beam, that would prevent swallowing. This is for my larger LEGO Technic BB-8 droid. In the project blog, the third picture down is the one showing the magnet encasement. The magnet scheme could be used in other models, perhaps singly for train couplings or for M-Tron as part of Classic Space revival. The encased magnets are stronger than a pair of the previous train coupling magnets by about 8mm distance for the same attraction strength. Whilst this could use the existing LEGO pieces (as per the rules of LEGO Ideas), it would be easy for TLG to make a 2-part case on similar lines, like a PF motor casing. Please support the project if you would like to see me take the magnet scheme forward. Thanks, Mark EDIT by Pandora: Please don't post links for self promotion. Whitefang literally just moderated in exactly the same way on the post before yours: If in doubt, check the Site Guidelines (tab, top right of each page) or PM me if you have concerns. Kind regards.

With Updates 10, 11 and 12 there have been quite a few changes to BB-8: - A new head design. - more details on the previous head self-levelling mechanism. - more details on counterweights and the working of the steering function. - further improvements to the colour scheme. - a new safe but versatile magnet encasement scheme for holding the head on. This can be applied to other models too. The LEGO system needs a more versatile magnet piece. BB-8's first outing to a show was OK, with some experience to feed back into the design. I'll be working more on the spinning function soon. Support is approaching 800 people. Please join in and help it reach 1000 soon. Thanks, Mark

The tightest curve is comfortable with centres at exactly 2x 48x48 plates, 96M diameter: One eighth of the tightest curve uses 19 2x2 tiles in a track piece. The next curve radii at 10M centres use 23, 27 and 31 2x2 tiles respectively. The curve uses these parts: 19 2x2 tile 36 2x2 plates 70 1x2 plates 4 1x1 plates 2 brick 1x2 w 2 holes 2 black pegs to connect to the next track piece Construction tips here and here. Add 4 2x2 tiles, 8 2x2 plates and 16 1x2 plates for the next size up, and so on. a half circle (4 pieces) should take about 20 minutes to make. A little over 3 years ago I took up Masao Hidaka's concept and added some engineering to recreate all the functions of the 1990s monorail in this system, including straights, curves, points, start/stop rail and hills. The purpose of this was to help with his LEGO Ideas project because that would need a kit with tracks fitting on a grid, like the original system. I also pushed out the limits: - Multiple curve radii; 4 tested, more and free-form possible. - Longer straights to span gaps, up to 64M tested. - Variable hills, 5-bricks-high and 10-bricks-high tested, easy to extend in multiples of 5-bricks-high. - Points are more reliable and also easy to fix. - 8-point junction possible with a single PF servo motor. Can make any subset of the 8-point pattern, such as reverse siding (3 points) or station throat. - Right-angle crossing possible - see pic above. - Faster trains. With PF M-motors in 6x14 chassises they are a bit faster than the original monorail but with PF train motors and smaller bogies 0.95m/s is possible, almost as fast as LEGO trains. - Helical tracks tested - ideal for world-record track length attempts, to beat the 562m of the original monorail. Double track helix improves density of tracks further. - Hidaka-san has also done a traverser and used an RCX to do train detection for automatic routing. The 2-PF-train-motor train did 3028m at full speed on a single LiPo charge, changing direction every 10 laps of this circuit. The first ten laps took a mean time of 11s per lap, the last 22s before the battery ran out. Further to the pictures of trains, I have found it best to have the motors at the front and rear of the train rather than in the middle. It doesn't like pushing an unpowered bogie. I'm doing a Space theme with trains of different Space factions. Hidaka-san bases his models on the Osaka monorail. The only limit on hills is the fact that it's a ride-on-top system, without a rack. This means the hills have to be a bit longer for the same height. a 5-brick hill is 48M long and a 10-brick hill is 80M long. Add 32M for every 5 bricks high (maybe an odd 1M or 2M to make up the diagonal to a multiple of 16M in some sizes). The original monorail did a 10-brick hill in 64M long so there is a 16M length overhead for each hill of at least 10-bricks high but the overhead does not increase. The best thing about this system is no track obsolescence as it is made of basic pieces. Picture folder here. Mark

If you want good torque at a high speed then the best solution from the current parts could be the PF train motor. Gearing up an XL motor would lose torque in the inefficiency of the gear friction. In the experiment to see if a Ninjago Airjitzu spinner (without pod or minifig) could be launched electrically, 2x M-motor geared up 15:1 did not manage it but 2x PF train motor geared up 5:1 did manage it, albeit an arm pull is a lot more powerful. This carries on from my previous use of 12V train motors with belt drive onto a Technic axle, which could provide 8 watts of electrical input per motor (not sure on conversion efficiency to mechanical power). I had 4 motors on a chassis (1 per wheel) and it climbed a pillow at 45 degrees before the electrical power fried a pole reverser switch! Mark