Brickthus

OK, I've uploaded the correct diagram. I've also added a circuit operation description to the info file. Correctness not guaranteed as I haven't done live tests on it with an oscilloscope! From my estimate of how it operates, adding more light bricks reduces the resistance between the darlington transistor's collector and the 12V supply. This shifts the voltage at the transistor's emitter during the cycle. With 2 light bricks, a resistance of about 18 ohms, it should be about 6V, but 4 light bricks with a total resistance of 9 ohms would make it about 7V. If you want to connect more light bricks, then I recommend you try keeping the total load resistance as near 18 ohms as possible. If 3 light bricks will work (an overall resistance of 12 ohms) then adding a 12 ohm resistor in series with each of the 4 light bricks would make a total resistance of 12 ohms with four in parallel. You would lose 1/4 of the voltage across each light brick though. Depends if visibility is critical under exhibition lighting or not. If' you're in simulated night time then it wouldn't matter. The 12 ohm resistors would need to be 0.75 Watt rating because power = VxV/R and 3x3/12 = 0.75. You might get away with 0.5 Watt rating because of the 50% duty cycle (spends only half the time on). The single-resistor alternative would be 3.3 ohms in series with the four parallel light bricks, a total of 12.3 ohms. Assuming 3V drop across the resistor, 3x3/3.3 = 3 Watt rating. This would generate serious heat and its leads might be too big to shove between the four prongs of a 12V plug pin! You might get away with 2 Watt rating for the 50% duty cycle. Make sure you don't melt any LEGO bricks or burn yourself! There is another way though, which might be simpler. You could try putting a 4.5V light brick in series with each 12V light brick, and using red filter bricks. This would save messing about with resistors but you would still need to re-wire a few leads. Make sure you don't over-volt the 4.5V light bricks though! Alternatively, adapting this for use with PF LED light bricks would be worth a try. Let's say a white LED drops 3V. According to the circuit of the light brick, if the bridge rectifier diodes drop 0.7V each and the LED drops 3V, the current is (9-4.4)V/4700 ohms, which is about 1mA per light brick. You want four light bricks, so say 4mA. Dropping (12-9)=3V at 4mA would need a 750 ohm resistor, so try first 820 ohms, then reduce to 680 ohms if the lights are too dim. If they are still too dim, you might be able to reduce the extra resistance in stages to as low as 100 ohms, but don't let the flasher unit output voltage go above 9V or you could blow the LEDs (for which I could not take any responsibility). Mark

That's what the 18 ohm resistor is for - raising some voltage in the event of an overcurrent condition, and so preventing thermal runaway of the darlington transistor. I expect this will limit the current, but I hadn't tried it with more and more light bricks to see where the limit was. It would make more sense for me to draw the diode and 18 ohm resistor nearer to the darlington transistor. The 4M7 resistor takes the feedback from above the diode and this makes the current limiting effective in the feedback loop, by varying the potential divider network for the BC547B transistor, raising its base voltage and biasing it a bit more ON, so cutting off some of the base current from the darlington. I have another small update to make to the diagram, as I got the other diode connection the wrong side of the 1Meg resistor. The diode makes the light bricks come on solidly if the power to the unit is connected in reverse polarity. This could easily happen in the 12V system if someone plugged it into the other output of the 12V level crossing panel as that reverses the polarity for the point (barrier) motors. Same goes for the light output of a signal switch. If the limit is 3 light bricks as you say, then the unit was designed for 2, as per set 7866. It was therefore an early decision to go with just 2 light bricks in the set. I wonder if it would drive PF LEDs just as well (~60mA per pair) if 9V were applied to the input, or whether it would be better to add a resistor to the output and run at 12V. That would allow 4 sets of lights at the 4 corners of the crossing. As we know, a proper UK level crossing needs 12 lights, 4 amber and 4 pairs of red. It would be possible to use the 9V alternate flashing units for the red lights, but it would need an RCX to reverse the current to do the "both reds on at once before alternate flashing begins". Just 2 RCX outputs would suffice though. It's not easy to get these diagrams correct at the first pass! Mark

I've not come across Abner or Conan. I did post on Lugnet.Trains but no reply yet. Is there another forum that they frequent? I think, given the circuit topology, that the 118 unit is more likely to have silicon NPN transistors (earlier posts had suggested PNP Germanium ones), so I posted a second diagram. I've also reverse-engineered the 7866 12V Level Crossing flasher unit. It's an R-L Relaxation Oscillator, quite an unusual circuit nowadays as most oscillators use capacitors. Circuit Diagram Mark

If the motor turns at least one way then it's not an open circuit on either of the control pins. Check for a short circuit on the plugs, that's the two receiver ports and both sides of the motor plug. Since the effect is the same on both ports, it's most likely to be on the motor plug. A short circuit from one control pin to either supply rail would let the motor turn one way and not the other whilst receiving commands correctly. Do you have another motor or light to test on the receiver outputs? Mark

Are you intending to just power an IR receiver with the regulated 12V input? It would not be as simple as using just a 9V voltage regulator! Yes, a 9V regulator would work with 12V input as most regulators need the input to be at least 2V above the output. The regulator would dissipate lots of heat (3 Volts headroom at 1 Amp = 3 Watts), requiring a heatsink as well as ventilation in its location in the train (to avoid melting bricks!). Heatsinks have a rating of degrees C per Watt and bricks would start to melt above about 45 degrees C (hence the 40-wash symbol on the box). If ambient temperature is up to 25 degrees C, you have only 20 degrees C to play with, so you would need a maximum of 6 degrees C per Watt for the heatsink, making it quite large! A Switch Mode Power Supply might be better (less heat dissipation and more efficient). The regulator in the 9V controller has a heatsink sufficient for it to provide about 300mA, but an XL motor needs a bit more current than that as the train gets longer. You could do direct drive (not IR) by turning the settings of the 12V controller into proportional voltage levels in the 0-9V scale, whilst providing enough current to the XL motor to get full mechanical power to pull the train. The required circuit for that would be a smoothed power divider, passing an average of 3/4 of the voltage to the 9V motor output whilst removing the unevenness of the input voltage (oscillating at 50-100Hz) and passing up to 1 Amp of current. 3V ant 1A = 3 Watts of heat to dissipate in the transistor, which would need heatsinking similar to the regulator. Alternatively you could use a PWM solution, providing power similar to the output of a Power Functions IR Receiver (higher frequency, which is why motors 'sing'). Either an LB1836 motor controller chip (the one used in the receiver and the NXT) or 4 transistors would be used for the motor driver. You would need a voltage-to-timing circuit to get the proportions right, perhaps using a 555 timer chip, which does that sort of thing well. The PWM solution should not dissipate so much heat in the transistors, but might heat the motor a bit more, but only within the PF system limits as if the motor were driven by a PF IR Receiver. Mark

Not quite. A PP3 battery (the square one) has a higher internal resistance than a set of 6 round ones. It is meant to drive a load of 100-150mA (even a Duracell one), which is not more than a light engine with a 9V motor. It was OK in the Technic set with the fibre optic unit because the load was small enough. Using the PP3 battery box for on-train lights is a good idea, especially now that PF LEDs use less current than the 9V bulbs, as the batteries will last a reasonable length of time. 6 AAAs would last longer still. A motor for a decent train, even in 6-wide, needs at least 6 AAAs. If this is what's in the new train sets, it would be a good compromise between the heavy weight and awkward fit of 6 AAs, the length of time before the power runs out (parents can have too much of a child's powered toy), and the expense of the rechargeable battery and charger. Mark

...because they only just made 1x4 tiles in dark brown, with 1x1s yet to appear! The track is kept simple for kids, so we AFOLs can enhance it with sleepers and ballast. Hopefully the track pieces consider out needs as well as those of kids, but the flexi-track was a bit of a mess! The Emerald Night was a big step forward for TLG. a GWR Pannier Tank might be possible with the EN wheels, as long as a motor would fit inside. A GM UK Class 66 would make a good diesel loco set for Europe, as they are in many European countries too. Given that we've had the Santa Fe and BNSF locos, I think it's Europe's turn for a diesel! Steam may be more exciting, but a diesel is more easily within kids' parents' price range! Mark

How about AAAs? If the battery box with the new train sets were a 6-AAA box instead of the rechargeable one? Mark

I've added a compilation of unit pictures in the folder, to go with the circuit diagram. Folder when moderated. Does anyone have a 138 unit that I could reverse engineer? Mark

1. The idea is that the train motor and the XL motor have to vary in speed together, so they should be powered from the same output of the IR receiver so that the user doesn't have to constantly fine tune two motor speeds relative to each other. The difference in gearing is the problem. 2. Where would the weight blocks fit? The XL motor fills the cab! Best to use the Emerald Night for what it is - TLG's best ever official steam engine set. If you want lots of coaches, add more engines as well! If the Emerald Night can be pushed without derailing, put two train motors in the first coach instead of powering the loco. Better to use 9V train motors than 8866 motors. I hope there will be a better motor in the new train sets than the 8866 motor! I have been petitioning TLG for a motor with regulation as good as the 9V train motor. You might like to try a 12V train motor at the front of the carriage as it has the equivalent power of two 9V train motors. When I converted from 12V to 9V I had to double up on motors for 8+ wide trains. The Emerald Night has a fundamental problem because it attempts to make 6 wheelsets touch the track and support the weight of the loco without using suspension. Try running without the front and rear bogies and see if it will balance. If so, reduce the heights of the bogies and let the loco weight rest on the driving wheels. If not, suspend the driving wheels off the rails, make each bogie into 4 wheels and put the motor in the carriage. My large steam engines, like Olton Hall, the 9F and this 0-4-0 chassis have suspended driving wheels (mostly wheels too big for the track, or not compatible) and use the front bogie and tender motors to carry the weight. That way the weight ends up on the motor wheels, which is where you need the traction, and the loco rides like a Bo-Bo diesel. Mark

Very nice MOCs as usual! I toyed with the 1-wide lever spacing for a signal lever frame but then tried to get them closer together because the real distance varies from 4.5 to 6 inches depending on the signalling company. Hence I used bad robot arms. Gallery here In the interests of working mechanisms, I might revert to the 1-wide spacing in order to make lever movement easier. With modified Technic axle pins (removing the peg and leaving 1M of axle with a bezel in the end) it should be easier to make the levers move with rods sliding through the base of the front wall, all in 1-wide per lever for those I choose to have moving (not all of them). I might use small pneumatic cylinders in the locking room, connected in parallel with the cylinders on other layout modules that do the points and signals. The disadvantage with 1ft lever spacing is that, after 2 months working out how to signal the layout, I will need 50-60 levers, and I wanted the length of the signal box to be right for the number of levers! I originally envisaged a 30ft signal box bit have already upped that to 44ft. Remember the lever colours: Yellow for distants, usually at each end of the frame, red for homes and starters, black for points and blue for facing point locks, which may be combined into blue/black on the same lever, white for spares, brown for ground frames and level crossing gates and locks. I had also considered using a ship's wheel as a level crossing gate wheel! I'd like to get hold of some red, yellow, blue, brown and white tubing for cutting up, but since I had no spare wheelset axles and could not do your bare lever end method I have used light sabre blades up to now. I like the instruments and bells too. Signal box window schemes have been in my mind for a while. None of them naturally fit the standard LEGO window range and I didn't fancy your trick of putting stickers on larger panes. Up to now I have 1x2 clear bricks as SNOT, stacked 6-wide and quite high with 1x6 white plates but I think 6ft might be too tall for the windows. Finalising that can wait till I know the final length of the box! Another half a pane below the opening panes would probably work for my box, but mounting them isn't easy. Perhaps you would like to extend your signal box for the size of your layout? Mark

I've reverse-engineered the 118 unit and drawn up the circuit diagram (when moderated). As I suspected, there is a high frequency tuned circuit on the front end, to detect the whistle frequency, followed by a flip-flop to switch the state, followed by a power transistor to connect the battery power to the motor. The RC tuned circuit has an input high-pass corner frequency of about 12.25kHz (a bit lower, subject to microphone impedance) and a feedback low-cut corner frequency of about 3030Hz. The bistable flip-flop relies on capacitors discharging together, the 4uF one balancing the 10uF ones for each transistor. The output stage has protective chokes and a large capacitor. Now that I have a circuit I could probably make an electronic version of the whistle, or replicate the circuit with modern components. In the 60s they used germanium diodes (0.2V drop rather than 0.7-1V) and the gain of transistors was a lot less at low frequencies, so some extra high-pass filtering might be necessary. You can't just replace a 60s transistor with a 21st century one! It was cutting-edge toy technology back then. I see from the Peeron photos of the 138 unit that the bidirectional unit contains a 14-pin chip and 4 transistors in an H-bridge. Still 5 electrolytic capacitors, so I suspect a similar technique is used for the bistable operation. No surprise about the H-bridge for bidirectional drive but I'd like to know which IC it is. I guess there are 2 flip-flops in there, one for on-off and the other for forwards-backwards. Probably either two tuned circuits or two different capacitor sizes too, to separate the long and short whistle bursts. Bricklink shows an adjustable whistle with a yellow bottom, so could it have used two frequencies? Steve, is the lid loose on your unit, so that you could see the chip markings? It could be a gate chip (7400, 7402), a pair of flip-flops (7473, 7476), op-amps (less likely as most need more voltage than 4.5V) or a transistor array. Take care with the wires to the unit's bottom terminals if you open the unit. I had to re-solder mine once! Of course I would reverse-engineer the circuit if I had a unit handy! Mark

What will you use for ballast? 1x1 round plates are the smallest round piece, since the various ball pieces are too big. Interesting to test a hopper to see how well it unloads your chosen ballast. Does it empty straight downwards, through the track, or to the sides? Mark

Is the railway policy to reverse the train into the factory siding, as it is in the UK? That would use 2 starting signals for the firard direction and a shunting signal for the siding shunt in the reverse direction. Will you be able to approach the factory from both directions? If so, a run-round loop would be required, adding interest to the operation of the layout, and more signals! I can recommend a book by Roger Amos, called "Complete Handbook of Model Railway Electronics" ISBN 1-85260-288-0, which has various signalling and control schemes and circuits. It shows a standard signal circuit that can drive 2 to 4 LEDs for signals of various numbers of aspects, and which can cascade the signals from one section to the next. Mark

In the UK there's a Kirow crane that keeps its jib horizontal when moving track panels, so as to avoid the overhead wires. The picture shows the crane using its beam to lift track panels. The crane has an extra hydraulic foot to support the jib. I built a working PF scale model of a Cowans Sheldon 76T rail crane. 76T would be 6 packs of 6 AA batteries, which I think it will lift at the scale distance from the centre of the turntable, given enough counterweights! I plan to have it in the yard on my layout, so that it can operate and be something other than a train that moves. Going round sharp curves is difficult, even with the turntable set to free run. Mark

1. No. More likely a box for 6 AAAs. This would actually be better for people who don't run their trains 24/7. The rechargeable battery is better for shows. 2. Yes. That would be the whole idea. Good to see such ideas in the engineering of the system, instead of too many bespoke shapes. 3. Could do. It should be easy enough for kids. I think the only time TLG encouraged people to have double track in a set was the 7866 12V level crossing. Kids have to start with a single track unless their parents have too much money (but of course LEGO is the best toy to spend it on!) The only other time was in the 7777 ideas book, when the bridge of the 7822 station was extended over 2 tracks and given light bricks. Ideas books are allowed to go beyond what we can afford today Of course 12V tracks were always adjacent, with no "six-foot" between them (it is known as that even if it is not exactly 6ft, and it compares to the "four-foot" between the rails). We need the 6ft to be 8 studs to give us more passing clearance on the corners, since the standard curves have a radius of 40ft rather than the 250ft minimum on main line railways. There are some bay platforms at British stations where a train could be entered from both sides (e.g. Derby Platform 5), but only one side is ever used. Some trains have lights on the open side, which has to be set to one or the other. Maybe it opens both sides in an emergency though. It would be useful to alight from both sides in double deck trains in cities, to get twice the passenger change rate. Mark

I put a motor in the coach but it just pushed the loco off the track. This is because the gear ratios are different and the 8866 motor is over-geared. I tried changing the XL motor for a medium motor (speed x3) but the loco didn't have enough power and was still pushed off by the motor in the coach. All this told me was that the 8866 motor is no good for trains that need real power. The 9V train motor is much better. I use them in pairs in diesels, wired together, with no problems at all on slopes or uneven track. A pair of 8866 motors would not pull a load up a hill that two 9V motors can manage, and would run away and derail down the same hill when the 9V motors would stay on easily. I limit the slopes on my layout to 1 plate per 16 studs on curves and 1 plate per 12 studs (4 plates per 48 stud baseplate) on the straights. This is the maximum 1 in 30 slope recommended for model railways. The change of slope is limited to half a plate per straight. I use Technic bricks on their sides or jumper plates to get the half plate height offsets for track heights. Mark

Great to see more people doing signals! Which country are you modelling and have you found any info on the signalling rules for that country? e.g. where to place the signals, what combinations are on a post? I've made some working semaphore signals for my UK-prototype railway. Single arm home signal (four on this layout, along with some simple colour light ones) Double arm starting signal with flexes, bell cranks, counterweights and lights Shunting signal Lever frame test for signal box I decided to work out what signals to build for my new layout but, on reading some books, it got complicated very quickly with all the rules of signalling! Fortunately I managed to trim back the number of signals but it still means the signal box will have to be bigger than I thought, in order to fit all the levers! It's taken me about 5 revisions over 2 months to get it right, but I learnt a lot! The PF light brick is great for lighting up signals. The individual LED wires are long enough that the 2x2 block reaches the ground on most posts and brackets. It's easier to hide the block on a gantry. A semaphore signal arm needs only one LED per arm, where a colour light signal would need one light per aspect. The latest searchlight signals in the UK would need tri-colour LEDs, which I have successfully driven from a PF IR receiver. The best yellow colour for signals is with the green LED fully on and the red LED at about speed setting 5. I preferred the semaphore signals, even though they're more difficult to control, because they are something that moves other than a train and they are purely LEGO with nothing added. Mark

Yes! I remember the old ideas book from the 70s, when these were all the cogs there were. How I used to wish! Now, many Technic sets later, we have turntables about the size of the yellow cog, but nothing bigger except tracks and the Hailfire Droid wheel. The objectives have changed, and I want to fit things into smaller spaces! Mark

Absolutely - fewer SKUs is why the straight and curves are both in one pack. A better solution would be: - improve the flexi-track - get rid of the check rails and make the curve smooth in all positions so there's no train juddering in any curvature position. The first prototype shared with AFOLs was much better than what got into the set. It's only derailment problems for kids that spawned check rails. - then supply the straights and the equivalent of the curves as flexi-track in the one set, and get rid of the separate flexi-track pack. That makes fewer SKUs again, helping TLG's profits, and satisfies both AFOLs and kids because of the option to go straight or curved with the improved flexi-track. There would be no need for 6-wide curves at all if the flexi-track were as good as it needs to be! The same could not be argued for the straights because straightness of flexi-track requires fixture, so straights are still required. In the parts and moulds range, replace the redundant 6-wide curves with 4-wide straights for the Indy Mine Chase track, or make Indy Mine Chase flexi-track! Mark

Hailfire Droid wheels anyone? I have attached things to them, so it's not difficult to make a hub in the middle. Mark

I have one of the 4x8x2 electronic units, which I was given when someone found it in their garage. Years ago I had a go at reverse-engineering the circuit but I'll have to look up where I got to with that. Basically it's a switch that turns the motor on and off every time the whistle sounds. To do this it must have a high frequency tuned circuit on the front end, to detect the whistle frequency, followed by a flip-flop to switch the state, followed by a power transistor to connect the battery power to the motor. I think there are 6 transistors in the unit. I think the unit I have is the unit from set 118, an earlier version of the whistle train. I think the 138 unit might have allowed the train to go backwards too. One reason TLG didn't continue with this is that all the plugs are the same. The user only knows which one to connect where by the coloured dot next to each port - blue for microphone, red for power, with the motor connections underneath. There was potential for the user to connect things up incorrectly, which might damage the unit. Something TLG considered in the development of Power Functions was trying to get rid of the ability for the user to easily mis-connect the electrical elements. That's one reason why the extension lead has a 9V plug on the bottom - if it were a PF plug it could short out two PF output ports. Mark

My favourite was the 8074 Flex Universal Set. It teaches the builder the difference between struts (axles) and ties (flexes). Bring back the flex system TLG! Mark

You're right that not all models should have motors. Perhaps any model with Linear Actuators should be capable of having a motor because 26 turns of a shaft is too many for hand operation, leading to user boredom! Perhaps any model with more than 2 pneumatic cylinders should be capable of having a motor for the compressor, especially if any automatic finite state machine operation is used (as in 8868 2nd model). Perhaps any crane with string longer than 1 metre should be capable of having a motor for the spool (as in 8421). Capable yes, compulsory no, but in some cases the set is too little without the motor. My favourite recent sets (8294 Excavator and 8049 Tractor with Log Loader) are enough without the motor but still benefit from having it. My beef with Technic sets as a teenager used to be that, with few exceptions, if a vehicle could be motorized it was never the wheel drive or steering that was motorized, primarily because the motors were never powerful enough (e.g. the 870 motor needed the 872 gear set to get any power at all and by then it was down to 200rpm at 3 Watts or less). PF sorted that out with a more powerful XL motor, used it for drive in the 8275 Bulldozer and will continue with the 8043 Excavator, though neither of these is especially quick. The PF medium motor is good for driving smaller vehicles and has more rpm and more power than the 870-872 combination. The PF IR system also gets rid of the long wires, which twist and tangle if there is more than one. We have to consider how long any non-AFOL Technic enthusiast will be interested in Technic. The age range is 7-16, so that's no more than 3 evolutions of the range of sets. By the time they've seen the same type of model twice, they're probably ready to move onto something else anyway (girls perhaps? :-) ), except the TFOLs who will become AFOLs and begin to care more about the parts in the box than about what the model is and are looking more at whether the new parts facilitate their own larger Technic MOCs. Innovation is for all. If TLG see us continuing to innovate, it encourages them. They need our encouragement. Some innovative models I have made, I am not allowed to share online because they relate to intellectual property owned by companies. Shame really 'cos there are many who would like to see the techniques I've used. Some techniques I have shared but don't expect to end up in a set model, such as a swash plate pump or proportional pneumatics. They might demonstrate real engineering quite well but they're too complicated for a set. Incidentally the swash plate pump with 6 compressors began to wear out the turntables under load! The PF LEDs are for a lot more than Technic. They have to fit with almost all set themes. When I helped design the part with the LED in it, I made sure the end would fit in a Technic hole or headlight brick, for maximum utility for as many types of model as possible. You can hide the LEDs inside a 4-wide car for headlights so that only the light from the front is showing. I had hoped that the parts might be able to fit right through a Technic hole but the need for cable strain relief put paid to that as toy safety means that the wires must not be able to be pulled out of the plastic part. The wires are long enough to reach round the engine of the 4x4 and I found that if I used the LEDs to light railway signals, most of whose parts are Technic, the wires will reach to ground level to hide the black brick part and the part with the LED is smaller than the 9V light bricks I used before. The PF medium motor has a stud base to fit with other themes too, but the XL motor was always going to be primarily for Technic so it didn't need a stud base. There will inevitably be some compromises with the PF parts because of fitting multiple themes and toy safety regulations but TLG has done well overall. I think more integration of the light brick into the town theme should be encouraged though, as it is at least as good there as it is for Technic. To pick up Andy's point about PF pneumatics, the most desirable element might be a PF electrically-driven pneumatic valve. This would need solenoids and springs, which would be large when you consider the amount of force needed to move a pneumatic valve. In fact the solenoids would be as large as a PF medium motor, which we already have, so why not use that instead? Other than that, more cases of pneumatic cylinders driving valve levers are to be encouraged, especially because TLG began it with 8868 in 1992 and we really want to see the technique in a set again (example). It would confirm the renaissance of pneumatics that 8049 has begun. By all means let's have PF-motorized compressors for it though! Mark

Before I started exhibiting LEGO trains I decided to try making a more realistic hopper but using a similar unloading technique. I had the 24-tooth contrate wheel turn an axle with a crank that withdrew a sliding axle from the centre of a fixed hopper, allowing balls to flow out of the chutes on both sides of the wagon. A similar axle withdrew from the other end at the same time, requiring two contrate wheels to run on two racks on opposite sides of the track. I used the 6x16 locomotive base (which has a 4x10 hole in the middle for a motor). This allowed the balls to glow through the hole and be diverted to either side by a ridge roof tile. The axles slide just below the base and could exit at either end, just above the coupling pivots. The hoppers approximated the BR "Trout" hopper that has doors on either side but not in the middle. One thing with the 7777 hoppers is that the couplings may detach because of the axial load of the cogs running on the racks. This was the same with my "Trout" hoppers. More recently I built some Railtrack hoppers whose doors are controllable to open one side at a time. This time it's a lever that opens them, so this requires a device to rise from the trackbed to lift the lever. The hoppers each hold 1400 1x1 round plates and will discharge reliably to either side, allowing me to build a loading and unloading facility. I have also toyed with the Merry-go-round system used by UK coal trains since the 1960s. This system has a lever for each set of hopper doors. A pin on the end of the lever runs in a track as the wagon is pulled forwards at 0.5mph. The track moves up and down along its length, raising and lowering the pin and turning the lever, causing the hopper doors to open and close at the right places. I didn't get very far with this because of fitting the mechanisms for 3 sets of hopper doors into a limited length of 4-wheel wagon. Trackside actuation of hopper doors is a great theme to work with but it all started with the 7777 hoppers! Mark