Brickthus

Nice. First non-studless Technic MOC I've seen for a while. Good that you've added more features of the engine compartment. Most models have just the engine block. I bet the chassis is much stiffer than a studless one! Did you base it one the tipper truck, Test Car or similar chassis technology? I hope your clear motors don't grind. More than half of mine have done so and I think one stopped altogether. I think it's the lack of proper location of one of the faster gears on its axle inside. Mark

Many thanks to Conchas for blogging my video, along with some kind comments I've tried the mechanism with a small pneumatic cylinder. It works OK, not quite as well as with a larger cylinder, but overshoots more when supplied from a hand pump. This is because the hand pump capacity is more than the capacity of the small cylinder and the air arrives all at once, which is at just one point in the dither cycle. To stop the overshoot, either a mini compressor should be used (with capacity smaller than the small cylinder) or the air supply should be spread over at least one dither cycle (about 0.25 second for 240-300Hz dither). A smoother air supply should be OK. The other thing to note is that the stiction of the small cylinder is greater in proportion to its size than that of the larger cylinder. The advantage with a small cylinder is that the turquoise lever of the first mechanism is not required because the blue feedback beam sliding range of 2M is the same as the small cylinder extension. The other advantage is that a small cylinder lends itself to making a pneumatic steam engine for L-gauge track (surely one of the holy grails of Technic Trains!). I've built a basic steam engine with two of the feedback systems, quartered onto the wheels like those on a normal 2-cylinder steam engine. My aim was to lay it out in a way that would facilitate using BBB wheels on the track. At the moment it's too tall for a railway layout because of the long valve lever extensions, and the size of the mechanism forward of the driving wheels doesn't leave room for the front bogie of the engine, but I wanted to get the mechanism working before making too many modifications for the space envelope. From the original mechanism I fed the output (from the blue beam with small cylinder) to the wheels. This is connected to a valve gear crank axle via a +/-20ish degree advance/retard gear mechanism, whose cranks move the red input beam. That's enough to make the steam engine go one way or the other. The advance/retard mechanism is a simple one with gears because Walschaert's valve gear or Stephenson Link Motion required sliders, which proved to add too much play to the mechanism, as well as causing size and torque difficulties for the mounting arrangement. An expansion link for a demo of steam engine valve gear is one thing, one for real feedback to drive pneumatic valve levers is quite another! The combination of a proportion of cylinder position and a proportion of advanced or retarded wheel crank position to produce the valve position is exactly what happens on a steam engine. I don't expect it to have the performance or speed of commercial LEGO pneumatic engines, but there might be an advantage in its slow speed control, such as moving a fraction of one wheel revolution. I have some more testing and tweaking to do before taking pictures but I'll keep you posted. I hope there might eventually be some proportion between speed/torque output and the amount of advance/retard angle if I can get rid of all unwanted effects and tune the proportions of each signal (as a real steam engine with Walschaert's valve gear uses about 1/4 cylinder position and 3/4 advanced/retarded wheel position) There are other, simpler, steam engine mechanisms that I've doodled as well. I'll see which ones are best for what. My first 2-cylinder uni-directional one was in 1995, soon after I had assimilated the 8868 truck mechanism, which was the original 1991 LEGO pneumatic steam engine, being the first pneumatic repeating mechanism. Mark

For your last point, some people have made the scenery change in height (which can happen steeply in a small space) rather than the track (which cannot). Trains are more reliable on flat track and the scenic effect can be just as good. I have a number of model railway books that suggest things for layouts. Scenery slopes (such as cutting sides) can be made steeper if the viewer is facing them and shallower if the viewer is at the top of them, to improve the view (since that's the ides of an exhibition layout) A maximum slope of 1 in 30 is recommended for a model railway. That's 1 plate per 12 studs length for L-gauge. Not even 2 plates per piece! 1 plate per 12 studs length fits well on a 48x48 plate = 4 plates per baseplate. Make the change in slope really gradual. For large train models, 0.5 plate change in slope per track piece ensures no motor speed fluctuation as the train passes the change in slope. Even 1 plate change per piece can induce a "wow" in the motor as the wheel-slip changes. This diagram illustrates slopes and changes in slope - how to extend a slope with extra modules inserted. Contrary to the above link, I have found points OK on slopes up to 1 in 30 as long as there is no change of slope adjacent to the points in any direction. It only really matters for the curved direction and the toe, not the straight road. Folder 1 and Folder 2 show sloping track modules. Mark

The LEDs in the existing brick are connected with a bridge rectifier, to turn on in either direction of C1/C2 speed. They have an 18 ohm resistive connection to 0V but I think that's a bias, capacitor discharge mechanism (to protect the IR Receiver) or for detection of the device type rather than anything else. It might mean the LEDs could turn on in single output modes with C2 high and C1 floating, since the resistor shorts C1 to 0V. Certainly the two LEDs operate together. I would really like a brick with one LED per C1/C2 polarity - it would make a great railway signal. I even drew the comparable circuit. Mark

You're right Jetro. The main benefit is if you have 2 LEDs connected between each C1/C2 output and the +9V rail. I did this in my IR Receiver Test Circuit. In Brake mode, the IR receiver output has both C1 and C2 at 0V. That would turn on both LEDs. Setting a speed (+ or -) would leave one LED fully on and would dim the other the faster the speed, to off at full speed. The opposite direction would dim the other LED. Use a 390 ohm resistor of 0.25W rating in series with each LED. Alternatively the LEDs could be connected between C1/C2 and the 0V rail. In brake mode, both would be off. With a speed set, one LED would light up dimly, reaching full brightness at full speed, the other LED staying off. With single pin commands, each LED may be turned on or off individually, without affecting the other. We have yet to test whether both C1 and C2 may be at +9V at the same time. I think it will be possible if the motor driver IC has no restriction. That would mean 2 LEDs connected from C1/C2 to 0V could both be on together, as well as both being off together. The same goes for 2 LEDs connected from C1/C2 to +9V. Mark

You should use the servo as an input to this system, connecting it to the red input beam to create a 2M movement. That way, the servo position would translate directly into the position of the pneumatic cylinder, with reasonable accuracy. Video here. Mark

I did mine in Visio. I made a template of parts. 90-degree curvy lines for rubbery hoses, thinner straight lines for flex tubing (following the principle of minimising the balloon effect by keeping rubbery hoses just for the corners). Components have attachment points that sit on grid intersections. It makes drawing a diagram pretty quick. I have a lot more template parts, such as offset valves. The important thing is clarity. In Paint you could make a bitmap of some of these jpegs, make sure the colours are solid for each piece (as jpegs blur the edges) and use those pieces as a parts template, saved as a bitmap file. Then just save as jpegs when you've finished a diagram. Sketch by hand first to get a good layout. Mark

Thanks Some people make the best-looking MOCs. Others make the stuff that's hidden inside to make them move. I'm really glad to have made something that others want to build. The simplest things that this mechanism should be good for driving are: - vehicle steering, either Hand of God or PF remote with motor worm drive to the red sliding beam. - one or more pneumatic valves, stopping in the middle. Either the red beam or the blue beam or both could be swinging rather than sliding, which might be useful for a steering mechanism. The cylinder precision needs the dither and compressor (if used) to be at sufficiently high frequencies so that their cycle times are shorter than the response time of the cylinder to a change in air pressure. Lower frequencies might make the cylinder oscillate, but this could be useful for particular applications, such as intentional chaotic motion, which might make an exhibition model appear more life-like. The closed loop system also needs excess air. I realise compressed air is harder to come by in LEGO models because we usually use tiny compressors, but this follows the industrial principles. Industrial closed loop oil and fuel systems follow this principle and have excess fluid pressure, feeding back the excess. Feeding back is less easy for us because we don't have sealed 4-port valves. I had thought of using 4 valve switches, but it would double the dither motor load, slowing it down. The extra switches might help in making a sealed system, feeding expelled air back to a reservoir, but it is the differential pressure that moves the piston, so that could work against the objective. A better use for 2 more valves would be for a separate air loop to a separate cylinder, with the feedback from both cylinders being added or averaged back to the blue sliding beam. I was asked whether mini compressors could be added to the dither motor crank. These would be fixed to the red beam, with the whole lot sliding by 2M. If so, a pair should be used like a flat-2 engine so that air pressure peaks equally on both halves of the dither cycle. Otherwise the uneven pressure would bias the cylinder one way or the other. An air supply that is uneven and synchronised to the dither cycle could be used as a fail-safe, as in industry, to bias a function to a safer state. Given the excess air required, a separate compressor of larger capacity is preferable. If more cylinders are driven from the same air output (2 pipes), the feedback cylinder should be at the far end of the pipes. Otherwise the performance of the extra cylinders is not so good. Movement of extra cylinders will creep to one end or the other unless their position is fed back to the valve positions. With 2 cylinders both fed back, equal pipe distances are recommended. Easier to control is a single larger set of cylinders using just the one feedback connection. The 8421 crane jib could work well with this system, as long as the feedback used the correct proportion of jib height (~2M full travel). A set of parallel cylinders could lift a heavier load with the same control system, giving greater amplification and showing how a simple low-power input moving the red beam by 2M could drive something more powerful. Mark

If you want to build a train vehicle to scale, make sure you have a scale drawing to build from. There are books with scale drawings of locos and wagons. I have quite a few books with 4mm:1ft scale drawings for UK trains. I know US loco drawings have been in the Model Railroader Cyclopedias (expensive but most valuable), not sure about European or other countries. Go to model railway shows and ask - there will be book stalls at the show. Choose a suitable scale. "Minifig Scale" covers a wide range of scales, depending on which minifig dimension you work to. For a fixed scale, 8mm (1 stud) to the foot is a good scale. Trains are 8+ wide for the UK, 10+wide for the US. Selective Compression In Minifig Scale, missing out a few carriage windows makes shorter carriages. They can be 28 studs long. For a fixed scale, running trains of 3 coaches rather than 8 is more usual, as happens on other model railways. Shortening vehicles is less usual on a fixed scale railway. A coach is typically 64 studs long (+ corridor) in 8mm scale. The whole layout design needs to centre on the train length, compared to the space available. 3 coaches + loco at 8mm scale = 16ft x 12ft, so fixed scale railways are for people with lots of space at home! Curve radii Design all vehicles to be pushed round a circle of curves at 40-stud radius. A fixed scale might need a larger minimum curve radius on the main line, especially if the vehicles are longer than standard 6-wide vehicles. Adding 1 or 2 straights between each curve can work well, but beware its interaction with the length between carriage bogie centres, or it could lead to oscillation of the train and derailment at speed! Slopes On a fixed scale model railway, the track slope should not exceed 1 in 30 (1 plate height per 12 studs length). 1 in 40 is easier = 1 plate per track piece. Minifig scale railways can get away with a steeper slope, as long as all locos have good traction and trains are not too heavy. The change in track slope matters. Too much and you'll hear motors spinning up with wheel slip. Engine Power For all but the smallest locos, 2 train motors (1 per bogie) is recommended. Connect a wire between them so that 9V trains give both motors the same power if one loses contact on the points. PF trains would have both motors connected to the IR receiver anyway. Cross connection prolongs motor life by avoiding them pushing and pulling each other. If the load per motor of a 9V train exceeds 200mA constantly, or 250mA regularly, add another motor to the train. Use a 0-9V bench power supply rather than the LEGO controller if you need more current than a controller will provide (if 2 motors are not enough to pull the train). I use a dual 0-30V 0-3A bench supply for the main lines, and LEGO controllers in the yards. Mark

Shame I didn't have time to enter any MOCs in the competitions! Then again, I didn't have time to build any new MOCs of the right types this summer. At least I got the outer circuit of my layout working though. I've found that too much time spent online reduces time spent building. After discussing the competition, maybe people are starting to miss building again! In a way, a competition synchronises people's building and posting times, so it concentrates effort into one time, rather than the steady trickle of MOCs we usually have. This week I've spent some time on a revolutionary new LEGO pneumatic technique: Continuously Variable Pneumatics Pictures and Info Demonstration Video It allows a pneumatic cylinder to stop at any desired position, not just at the ends or in the middle. It might take a little while to understand - it's not immediately obvious how it works. In time it could well have possibilities for a pneumatic steam loco! Mark

A lemonade bottle is an interesting suggestion. Is that how people power their pneumatic engines in cars (e.g. V8 Mustang), to make them untethered? Here's a 10-minute tutorial covering most aspects. Hope it's small enough Milan! Conrinuously Variable Pneumatics Video I used free Prism software to convert the .mov file to .mp4 - 58MB rather than 290MB! I think my voice is sufficiently audible over the compressor. I'm glad it works at 20psi because that's at the lower end of the range for pneumatics, prolonging their life. This system has a small risk of valve seal wear but it's a lot slower than the engines! I didn't write a script in the end, so maybe there's a few too many "um"s, "er"s or "so"s in there! (Not as many as some of my teachers at school though - 768 in a 40-minute lesson once!) An old computer power supply (12V at 6A) that I used to use for the compressor decided it was time to expire. Probably one of the power diodes on the mains side. The smell was awful. Glad I unplugged it before any overheating became serious! At least I also had the smaller 3 Amp supply that I use to charge my car battery. Enjoy! Mark

I'm glad you want to try it. I guess it's like that with a lot of innovations - it has to be repeatable by an independent person and it takes time to sink into the minds of those who can understand and think of ways to use it. Understanding quicker by doing is why I have made models of a few concepts I deal with at work. People at work really enjoyed seeing the models and I had 50 people turn up to see them one lunchtime! Unfortunately I can't share pictures of them on the web This system arose from my growing understanding of how hydro-mechanical units work in the control of fluids in real world engineering (my first degree was electronics, not mechanical engineering). The way it works is similar for cars, planes, trains, ships and many other applications. An electronic controller sends a variable signal to the hydro-mechanical unit. The signal moves a motor that controls a valve, which regulates the pressure on each side of a piston that moves a larger valve, which regulates the flow of fluid to an engine or other device. A feedback sensor detects the main valve position and feeds it back to the electronic unit. The electronic unit uses the error signal (the difference between the demand and the actual valve position) to determine what signal to send. The system would be inherently unstable (it would go to one end stop or the other) except for the closed loop feedback. A further, outer, loop is made by sensing the effect on the driven equipment, so that the electronic unit can work out what valve position to demand. I'll have to add that in another experiment. It was a pleasant surprise to find that it works well with pneumatics as well as with hydraulics. Hydraulic fluid (oil or fuel) is not very compressible (proper hydraulic fluid is even less compressible) but air is very compressible. The important thing here was to use the same air pressure for both sides of the cylinder. Then performance under load is proportional to the supplied pressure, within the limits of LEGO parts (30psi). I picked a cylinder with least air flow restriction and a good seal at the top (I sorted my cylinders into different bags depending on seal integrity and air flow resistance). A benefit of the system over the mid-stop pneumatics of a few years ago (apart from needing fewer components and being smaller) is that the system can compensate for the unequal piston areas at the two ends of the cylinder. The mid-stop system needed two opposing cylinders to equalise the areas, otherwise it would drift to the extended position. I knew this would be OK because a real system valve has to drift shut for safety reasons, so the shutting-side area of its piston is greater than the opening-side area. I have been thinking about making a video since Milan mentioned it. I was puzzling how to pump air, control the red input beam and hold the camera all at once! I'll see how it is with the car tyre air compressor - a bit noisy, but it frees up a hand! My camera needs a lot of light to do a movie, so lamps too, and I must think of a script that won't confuse everyone! Mark

I see Bluebrick doesn't yet support track heights and slopes. Could that code be rewritten to do heights and slopes better than Track Designer? I need multiples of 0.5 and 0.333 of a plate because the track on my layout changes at 1 plate for 3 straights or 1 plate for 2 straights in places. I need to specify the height of each track piece, with variable slopes and changes of slope, and for the height values to stay put, whatever changes are made to adjacent heights. Is this a feature that could be added at the next version? Could it toggle showing the numbers of plates height alongside the track? Mark

A big Thank You to TLG for listening to AFOL requests for pneumatics and producing set 8049 for 2010. To celebrate, here are the pictures of my latest pneumatic system. It allows the cylinder to be commanded to any position. - not just the two ends - not just two ends and one position in the middle - yes, any position within its travel! The variable input is at the red sliding beam. The variable output is the position of the turquoise lever. Techniques: dithering of valves (with the motor crank) overcomes their hysteresis and stiction offsetting the valve levers removes most of the deadband in the middle feeding back the output allows closed loop control of cylinder position This uses no more pneumatic or PF parts than the contents of sets 8049 and 8293, so the parts will soon be quite accessible. How it works might take a little more understanding! More detailed description / info here. Folder when moderated. PLMKWYT Mark

Does bluebrick allow different places on the track to be given different plate heights above the base in a better way than Track Designer? I had to abandon the slope function in Track Designer because it would not fix what I'd done when I changed something else and it would only set heights at the ends of track pieces and to a single slope and slope change. There are places on my layout where I need slopes of 1 plate per 12 studs (4 plates per 48x48 baseplate over 3 straights) or even 1 plate in 3 straights, changing to 1 plate per piece and 0.5 plate per piece in other places. The slope change is 0.5 plate per piece per piece, so some track piece ends are at heights of 0.5 plates or 0.333 plates! Unless bluebrick copes with this, there would be no point me redrawing my layout in bluebrick! Till I have a package that does these things, I'll not publish my latest layout online, but here are a couple I've exhibited before: This was the layout for my 4th exhibition, in April 2001. 12V with two main line trains running in opposite directions. Station at the top. Isolation at the crossovers between each main line and the yard. The yard has a loop to run round, and a gantry crane along one of the diagonal sidings. A level crossing at the top of the diagonal bit, just after the double bends. 8 8ft x 2ft sheets of chipboard, coloured tan. Soon after that I moved over to 9V (keeping the 12V for London Underground trains). I designed the 9V layout for easy swapping of trains, so that the public always see something moving: Size 16ft x 12ft with 8ft x 4ft operating well in the middle. Station on the right, 4 through platforms plus a bay platform for the Hogwarts Express (I had built just 1 coach then) Again 2 concentric main lines. 3 trains on the outer track, 5 on the inner track, plus more from the sidings in the yard. The bottom road at the top had a goods shed. The top road at the bottom was adjacent to a gas holder. Points electric (modded sliders and with micro motors) except for a few in the yard sidings. 20 tables 4ft x 2ft, painted green. A smaller layout 8ft x 6ft on some of the same tables. One running line with station passing loop and sidings. Fits in a large estate car (we couldn't do so many shows with the large layout, hiring a van, 8hr setup etc..., so this works well). I used the small layout to test the addition of trackbed to the track pieces. Each grey plates pack 10148 did 1 straight + 1 curve, or 1 switch point. I should publish this plan. Since then I have aspired to build a multi-layer fully scenic modular layout. For exhibiting at a model railway show, the standard for a LEGO railway should be as good as that for a railway of any other scale or gauge, especially since OO gauge has its own scale and gauge compromises. I doubt LEGO will ever compete with EM, P4 or O-gauge in scale accuracy, but elements of it will have no more compromises than OO. HO is probably more accurate than OO. To that end I have stuck to some standards in model railway books, such as a maximum slope of 1 in 30. So far I'm 4.5 years into the project. It was originally planned for 5 years but will probably be about 9 years! Plan not published yet but pictures are here. Mark

Maybe I will need a credit card after all. I have managed to avoid having one till now. With a debit card if you don't have the cash you can't spend it! It's probably worth having the extra credit card protection for online payments anyway. This may happen after I have tested the theory with the sensor from LS@H. I have to decide what part of the train to put the tag or disc on. In the train roof would be good because underneath tracks can have the sensor suspended above them. Visible tracks need a gantry as an excuse, with some black tape round the sensor, but that's OK. If the RFID will go thoguh 1-wide bricks than a train roof should be OK too, plus all the trains are the same height but not the same width. I have also considered that the bell codes for a loco will need to change depending on whether it is pulling a train or running light engine. I might put 1 tag in the loco and another in the train itself. The loco tag would trigger the single "ding" that wakes up the next signalman, to which he responds with a single "ding" (a different tone from the same NXT). The train tag could be in the last truck or carriage so that if it were not detected within a few seconds, due to a decoupling accident, the bell codes would sound for "light engine", telling me the train had broken before the loco ran round the layout and hit the back of the train! Otherwise the train tag would dictate the bell code, perhaps 3-1 for a passenger train or various ones for freight trains, depending on the type of load and whether the train has continuous brakes. The train code would be repeated by the next signalman (at the different tone) and the twin "ding-ding" for Train Entering Section would follow (maybe the NXT could also drive a gauge showing "Line Clear"). I have books on signalling that give a list of codes. I guess it will require some practice to get right but I've seen model railway enthusiasts running a layout with proper signal box operation before. I hear from HiTechnic that I2C is limited to 90cm, which (surprise surprise!) is the longest lead they make! Bluetooth is an option, yes. I had considered the need for 4 NXTs as there are 4 places on each circuit where I would want to change train speed, two of which would also need the signal boxes to communicate, making the bell code sounds before the train enters or leaves the viewable section. Each position on each line would need an RFID sensor and an IR Link sensor for PF trains, a total of 16 ports, just right for 4 NXTs given that a multiplexer can't fit between tracks. If the module gaps could be bridged detachably, I could try having an NXT per signal box, sending a Bluetooth code to the next NXT whenever it sounded a bell code. I have corner hills that could house NXTs on the outer circuit and space under the yards that could house the inner circuit ones. Not so many buildings though! Some layouts have lots of town with a train. My layout is the opposite - realistic trackbed up to 9 tracks wide with a small amount of hopefully-good scenery, more in keeping with model railway show layouts of other scales and gauges. It would have been nice to see the NXT screens, but the tables are 1200mm wide, which is more than the 90cm of I2C! Mark

A Technic train would be great, especially if it were pneumatic-powered. Making a purely pneumatic powered train go in both directions would require two reversers, to swap over the air connections for the other direction of travel. The way with just pneumatics would need 2 cylinders and 2 valves to go forwards, plus another 8 valves to reverse. Both methods still need compressed air. This leads to one of the pneumatic pieces I would really like TLG to make: a 4-port reverser valve, doing a proper 5/3 industrial valve function, in a piece not much bigger than the existing pneumatic valve switch, maybe 1 stud taller so that a complete 24mm circular piece with embedded seals is fully inside the casing. This would mean a reverser could use 1 cylinder to drive 1 valve instead of pairs of cylinders driving 4 valves. The other piece I would really like to see TLG make is a telescopic cylinder, which would increase the extension to original length ratio. The end-to-end cylinders in the 8421 crane were a reasonable compromise (cheaper to design a bracket than a longer cylinder) but the extension ratio is the key. An articulated lorry tipper truck trailer has a telescopic cylinder at the front of the tipping bucket, not underneath as many Technic models have had. I would suggest that the first prototype begin with a small cylinder (minus bottom end) inside a large cylinder body. Aim for an extension of 7M, double the 3.5M of the regular cylinder and more than a Linear Actuator. For a pneumatic train, a 3-cylinder design might work well because pneumatics are good at sharing the air pressure. Once 3 cylinders and 3 valves were connected up, the train could move in both directions, 60 degrees of shaft rotation at a time, as each valve switch changed state in sequence. Valve control could be done from a crankshaft driven by a PF medium motor inside the boiler, which allows for PF speed control (the air changes keeping up with crankshaft rotation speed) as long as there's enough compressed air to keep up! This is the simplest way to do a bidirectional pneumatic train. I welcome the return of pneumatics in the 8049 logging tractor for 2010 and look forward to more experiments and discussions! One thing we should campaign for is for TLG to put the pneumatic parts in the PaB Online parts selection. We used to be able to get the parts from the Spares Service back in 1991, so this would restore that purchasing capability. We need a minimum of 5 parts: pump, valve, cylinder, pipe and T-junction. Small cylinder and compressor would be aspirational. Now that a supply of parts is expected, have a look here for pneumatic ideas to whet your appetite! Mark

I decided to stick to LEGO scenery. Painting and sprinkling grass is both too messy and too permanent. I build to scale, to a limited extent. The limit is no modding of bricks for purely scale reasons. 8mm:1ft scale is my ideal. Most UK prototype vehicles begin 8-wide and gravitate towards the scale. I might end up with some being 9-wide as models are updated (that's the great thing about LEGO). I have a book with 4mm scale drawings of UK main line diesel locos. Knowing your prototype is essential. If I did US vehicles I would start 10-wide as a real boxcar is 10'5" wide. I'll have to see if Toy Story minifig legs make minifigs the right height for 8mm scale. The exception is steam engine rods, which are too wide for the scale. I prefer to make them work, even the valve gear ones, which is more than OO gauge engines have. For scenery I like to try to do it real, rather than cartoon-styled. I use RAIL magazine as my main UK track scenery inspiration because it shows the modern UK railway with all the track furniture. I got this far in doing scenery - still signs and more foliage to add, along with more variation in trackbed colouring, but a modular layout looking like the real UK railway as far as possible, and facilitating realistic operation of trains. Slopes are limited to 1 in 30 with 0.5 plate/track piece changes in the slope. The outer circuit of the 16ft x 12ft layout is now able to run trains! It's taken me 4.5 years to get this far! I'll have to take a video soon, but there are lots of boxes obscuring the view. The slopes are a challenge to trains but a 4-motor 9V train (double headed or equivalent) is OK with heavy coaches or trucks, taking 1.3 amps on an uphill curve at 7.5V (scale speed ~30mph up the hill, ~60mph down the hill). A PF hybrid loco is OK climbing the hill under moderate load but only with as much train load as a single 2-motor 9V loco could pull. Mark

That's good, though they're not so easily available to normal UK methods of payment. I saw them on the Codatex website. They're 22mm so yes they will fit in a 1x4x3 panel because a 3-wide brick fits there too. I'll also need an NXT sensor socket-to-socket extension so I can install sensors in a modular layout with wires trapped under tracks on each side of the detachable module boundary. I suggested it to HiTechnic as they make extra leads and multiplexers already. A 4-way sensor multiplexer wouldn't fit between 2 train tracks by the time the leads are plugged in. Mark

I'd better get on and order it I'd like to hide the tag in the train. I notice it's like a 4x4 dish with an extra bit for the peg attachment, so that's too big to fit behind a 1x4x3 panel. Therefore it'll have to be behind a layer of 1-wide bricks. Given the 30mm range, will it work through 2 or 3 layers of 1-wide bricks? Mark

These wheels represent 5'3" in 8mm scale, but would be only 2'4" in an 18-wide scale. They could run on L-gauge straights or curves but not points. I've not tried them with any serious weight, partly because the curves are too sharp for the size of the flanges. That's the reason BBB wheel do not come in a larger size than 30.4mm. I already want larger wheels than the 5'3" wheels, to do up to 6'8" drivers for steam locos. When scaling up, the curve radius must scale up too. A standard curve is 40ft radius in 8mm scale but would be less than 20ft in an 18-wide scale. On the real railway, 40ft radius is only for 0-4-0 engines in a yard. The minimum on the main line is 250ft. A wider curve radius would make a Big Boy a bit easier. It needs half the driving wheels to be BBB blind drivers on L-gauge track. Mark

I have some sections of track far enough off the table (up to 49 plates up, to pass over other tracks), but it's not in every place where I would want to change train speed. I need: Max power up the hill. This is at table height. Min power down the hill. This could be done at the top of the down hill. Medium power through the station. This could be done at the top of the up hill. Medium or low power through the fiddle yard, to minimise derailments. This is all at table height. I toyed with the idea of raising everything up some more, but it would be a lot more bricks for modules covering 450 x 600 studs! It would have to be about 6 bricks to accommodate the sensor facing upwards. Mark

I'd like to recognise trains as they pass (using an RFID sensor if I can hide a 4x4 round tag on each train to pass within 3cm of the sensor). The program would then play sounds corresponding to the signal box bell codes for accepting the train and would set the speed of the PF trains with an IR Link sensor. Several NXTs would do the same at different points in the layout (only one doing bell code sounds). One NXT would ensure that trains went slowly enough not to derail in the (unreachable) fiddle yard and another would set maximum power for going up a 1 in 30 slope. I plan to test it soon, as soon as I've bought all the remaining parts. I've had success with the NXT commanding a PF IR receiver before. The test will check feasibility, especially for the 3cm distance relative to train loading gauge clearances (i.e. I have to mount the sesor on a straight piece of track, not a curve, to get a consistent distance). Feasibility at train speed will also need testing - can the sensor detect the tag fast enough as the train passes? If it works, I'll buy more RFID sensors and tags. BTW my modular 16ft x 12ft railway layout can now run trains round the outer circuit (after 4.5 years of building!). This includes a PF hybrid loco, which makes a good test of the PF electric system capability for other high power applications too. Some of the points are pneumatically-actuated and others use PF medium motors. Mark

Yes, I can't see much need for a micromotor in Technic, especially now that the PF pole reverser switch cannot rotate freely (it would have cost more to do it). Universal Set 8882 used that technique with a micromotor. The satellite solar panels in the Space Shuttle were a great application. It was always good for space or ship radar antennas and I used them for train point motors and a station clock. I used four 2838 motors to turn an open rotor jet engine model. It needed the rpm of these motors (4000 at no load). With 2 motors each end, I used the motor bearings and had no axles through beams, to keep friction low. The main axle was extended from one motor to another at the other end whilst the other motor at each end used 40:40 cogs to add its power. It took 1.5 amps from the power supply at 9.15V (= fresh battery voltage). Most of the energy was absorbed by the friction in an epicyclic gearbox that allowed the two rotors to contra-rotate. In the end it achieved about half the speed of the real engine (one quarter or less of the power, given the square or cube relationship of energy to rpm in a fan) and the low friction techniques allowed the energy input to overcome the gearboc friction, such that contra-rotation was achieved. As an alternative in that rpm range, I attempted to interface a 135W 10000rpm drill to a Technic axle, and succeeded in gearing it down 3:1, but the vibration was too great and it melted the pieces Yes, the 5292 motor was designed for direct car drive and has a speed between the PF medium motor and the ungeared motors. Its spur gear train has lower friction than some motor gear trains and the actual metal motor is bigger than the one in the PF medium motor. The 47154 motors were a disaster for the gear train, as it ground and jammed in more than half the motors I have of that type. Mark

Great loco. I tried a 3-truck Shay loco in 8mm scale. The trick is to get it to go round corners with the transmission shafts extending and contracting. The engine is a lot easier to do in a larger scale. I have to make more length for the orthogonal shaft (where the pulley are) before I can try an engine crankshaft there. I must cut up some new 3M uni-joints 18-wide trains take a lot of bricks. No wonder most in the Legoland parks consist of a bought-in chassis with some LEGO bricks on top. Even 8mm scale trains take more than twice as many bricks as 6-wides. Given that a Shay is narrow gauge, I wonder about the scope for G-gauge or 3-inch track, which would enable corners to be made. Anything needing wider track than L-gauge track should examine track solutions from other large scale model railways. Mark