Brickthus

I'm all for people developing stuff to go with PF but this doesn't seem the right idea to me. It's bulky - the PF concept is a small modular system - so they should take the control out of the battery unit, perhaps use a standard battery box. The concept needs to check the current available from AA batteries. Alkalione probably won't do. I've tested AA NiMH to 4.5A before charging them but I'm not sure they would like that in use. Therefore a LiPo source might be best. An NXT could do the control bit. A phone app could send it Bluetooth messages. Therefore what's the difference from what is already available? What PF needs from 3rd parties is small modules of the input, processing and actuator kinds that fit with the PF concept. A processing one I would like to see is a slave motor driver with 2-Amp capability. Control would come from an IR receiver's motor drive output. Power would come from a higher current or 11.1V 3S1P LiPo battery (to get the full 9.0V rather than half power at 7.4V) 1 IR receiver would drive 4 slave motor drivers with a train motor each for a 2-loco double headed train. If a timer or phone receiver module were to be made it should be no bigger than a pair of 2x8 bricks as a separate item from the battery box. While we're at it, let's have a faster motor (1000-2000rpm) with the same torque as a PF XL motor. Can't increase the torque beyond about 35Ncm as it twists axles, so increase the speed instead. I wondered whether a geared down Mabuchi RS-380 motor is the most powerful PF could stand, derated to less than half its capacity by limiting the current, otherwise it might get too hot for the surrounding bricks! Bulky is definitely not the way to go for PF. I would like a smaller NXT with separate battery box! I see a touch screen for NXT is available from at least one educational supplier in the UK, so maybe it could make the phone obsolete! So there's a few ideas here that they should try but IMHO the one they have is not the right one. Mark

This is good Tony, especially for a mechanism with no modded parts. It gets closer to the amount of valve gear that OO models have. They tend to cheat and have a moving expansion link but static parts between there and the cylinders. I did the full Walschaert's valve gear for my BR Class 9F but had to mod a few parts. I did something similar for my LMS Garratt, thoguh that has to widen it so that the train motors can swivel inside the driving wheels. this meant I also had to raise the expansion links to clear the platforms! Some of the modded parts I could argue TLG should be making anyway, such as the 1x2 "+o" thin liftarm and 1x3 "ooo" thin liftarm. I have asked them for a 40mm pulley like the 24mm one before. The way I developed since then does it with fewer mods to liftarms to make the valve gear rods and cranks (no cut up 24mm pulleys!), but with modded model team wheels as an improvement from 40-tooth cogs ('cos TLG don't make train wheels big enough, even as blind drivers): Carl told me about the cut up baseplate technique. 1 baseplate can make quite a few 1x4s for liftarms! It shows the possibility of doing all 3 cylinders of a Gresley Pacific (such as Flying Scotsman) or maybe a 2-6-2 V2 like Green Arrow. I hope it will form the basis for my future large steam engine sometime. Mark

Mark

We've had Flexi-track for a good couple of years now. Has anyone found the best ways to use it? (I had a go) One of its design intents is that a child can always make a circuit, but what about AFOL aspirations? Do we use it for wide radius curves according to another of its design intents? (Photos anyone?) Do we like it, suffer it or ignore it? - Has anyone made the best of the check rails by making street trams? Who has deliberately bought more Flexi-track in addition to what comes with a train set? Mark

Using LEGO pneumatic parts with water is generally a bad idea. The pressure gets too high. Take a note from Robot Wars - the rules say 1000psi pneumatics or 3000 or 5000psi hydraulics - 3x or 5x the pressure. Years ago I tried it with an old style non-return valve and ruined it - it can't stand the pressure. I had hoped it would be an easy way to make a fire engine - input end in the tank, output to the hose, pump in the middle. I think some of the old studded 2-way valves also degraded because of overpressure. The pneumatic parts rely on grease inside them to keep the piston and valve seals. Water would wash it away. Also the seal at the top of a cylinder is not brilliant - some leak. The test is to extend the cylinder, put your thumb over the top nozzle, squeeze the cylinder fully with the other hand and let go. If it doesn't spring back at least half way, you have a leaky top seal. It usually leaks in the inlet direction more than the outlet direction, so it should be OK for light duty. I use the worst leaky cylinders for push applications only, such as rail points or the 8421 crane jib raise function. This is the way to get precise positioning of cylinders, nearer to what real-world hydraulics achieve. Mark

This is coming along nicely The 5M studless beams on the upper layer of 2 should be these upside down. There are 2 5M axles through the bush parts of them and the 3x5 L beams and the 5M beams on the lower layer, replacing the black pegs you have. The 7M beams on the upper layer might be 2x 3M means but only a top-down view if the set will tell us. On the pitch change mechanism there is one of these either side, with the pin hole connecting with a black peg interface towards the central shaft. Could be a peg or ball joint. If it is a large turntable holding the 4 3x5 L-beams then that's high-friction for rotating the rotor. A motor would need to do 1120rpm to get the rotor to a realistic 160rpm with the 8:56 ratio - faster than both PF motors. Mark

Here are some basic circuits: You didn't say what type of pneumatics. The current stuff has cylinders with 2 nozzles and pumps with a single nozzle. The old stuff has cylinders (with and without the spring) with just 1 nozzle and operates differently. Both types use 3-nozzle valves. With the current system the input is connected to the middle nozzle and the 2 ends of the output cylinder are connected to the other 2 nozzles. If the valve is moved to one side, the input connects to one output and the other output leaks to the atmosphere. The old system used a 2x4 brick with 3 nozzles - a non-return valve - on the input side and the switch valves were on the output side. You could not move 2 cylinders in opposite directions at the same time with the old system or it would lose pressure and collapse! You can take pneumatics quite far, to build automated machines: And yes, it is possible to make a pneumatic cylinder stop wherever you like, if you use a few tricks: There is stiction to overcome but once the cylinder is moving, more accuracy may be achieved than was suggested by earlier posts. Also load handling is a matter of how many cylinders you use in parallel. The above system may apply to many series or parallel cylinders to handle a greater extension or a greater load, without changing the servo mechanism. It is easier to parallel pneumatic cylinders than LAs, as TLG found with their 8043 excavator set! Pneumatics will load-share easily. Set 8455 uses them in pairs for the Backhoe's front implements. The classic crane truck set 8868 used a pair of cylinders in opposition to turn the crane's turntable. It doesn't matter that the movement of 2 cylinders is not the same throughout the range; it does matter with LAs. More pictures & info here Mark

I see how the NXT system is dedicated to robotics but my personal preference for Robot MOCs is Power Functions or pneumatics without the software. I would not be averse to including NXTs if they solved the problem best, perhaps controlling PF with the IR Link sensor. I find that position detection is tricky to get right and I would not want to spend too much time on it compared to the design time of the robot. A few PF robots: Walking Robot, 2x XL motors, can turn like a tank and buck like a horse. Rover robot that can travel over uneven surfaces, tip and pick up objects (2x XL for drive + 2x M for tilt and grab). I used this to test the prototype PF parts. Automated machines concept with PF devices talking to each other. Pneumatics: Pick and Place robot (6 Large + 2 Small cylinders, 9 valve switches providing 4 functions with 10 movements per cycle) that moves bricks from a chute to a bucket. Continuously Variable Pneumatics including Stewart Platform WIP Other building blocks and concepts If I were to build another robot, it would all depend on the problem I had to solve. So many have been done a few times. Mark

We Brits usually end up paying in £ what you pay in $ stateside. I would expect Tractor £40, Helicopter hope £80 (hope not £100 'cos I'd think twice) and Crawler £150 (i.e. Unimog money) The tractor does not replace the tractor with log loader because the pneumatics were the best bit of that set, but it is a bit more substantial than that tractor. Effective replacement for 851 all those years ago - my first Technic set on my 8th birthday! The Helicopter is better than 8856 so no real helicopter set to benchmark against. The lack of PF in the set helps the price but the bespoke blades hinder it. The Crawler effectively replaces the PF Bulldozer at a price point (2x M + 2x XL replaced by 2x L + 1x servo) so we could judge from that what the inflation rate has been between the two launch years. Mark

Yes, the XL motor has enough torque. What is needed is to provide that same torque at between 5 and 10 times the speed: 1000-2000rpm. That way, speeding up by x3 or x5 gives enough torque for a propeller. Reducing by /3 or /5 provides high torque with enough speed for a ground vehicle. I have used 4x 4000rpm motors with a dual contra-rotating rotor system. It takes a total of 1.3 Amps but needs more speed, maybe 3x as I was aiming for 7000rpm input. A 135W drill didn't work as it melted the axle! RC Car and plane motors use up to 40 Amps at 7.4V (car or 11.1V (aircraft) using the latest LiPo packs. One snag is that RC car and plane motors are usually rated in turns (maybe 17 to 27 turns), not in real units like torque. 40 Amps is a bit much for LEGO models but I'd go with 5 Amps at 9.0 Volts to get 40Ncm of torque at 2000rpm. That's 45 Watts, 1/3 of the drill power nut 3x the 1.7A stall current of an XL motor. It might need a bit more current than 5 Amps to get both torque and speed. Other electrical things I would like, to complement motor speed and torque: 11.1V LiPo and PWM controller to give 9.0V output instead of the measly 7.4V from the LiPo, which is half the power. Slave motor drivers (1-5 Amps output) to overcome the motor driver current limit of 400mA per channel. 128-channel control, either IR or radio. If it needs the voltage to rise in order to get the power without excessive current (and thicker wires), to 14.8V or 18.5V then that's OK, as long as kids would be safe to use it. The limit is usually 30 or 36VDC for safety but perhaps toy regulations are tighter than that. 11.1V is also a widely used standard for laptops and RC models so that might be cheaper - it has to be commercially sound. Unfortunately the AFOL market is comparatively small. Perhaps it should be like third party NXT sensors, where AFOLs could get TLG sponsorship with the cases and make the internals themselves? Mark

Please be careful with words like "faulty". There is no fault that I know of. I believe there will be some movement of blade pitch, so it fulfils its design intent and works as well as expected. It's definitely not in the category of the 8043 excavator with its original (now fixed) LA internal shredding problem! I think we agree that a designer would show off all the great features better than the people whose skills are primarily sales and diplomacy at shows. I think designers' pride in their work would make sure everyone knew how cool it is - same as many of us do with our MOCs! Let me explain why I doubt there's cyclic pitch change. It's easy to quote an opinion but it's important to understand the rationale. I built a replica rotor with the parts I could see or discern from the pictures and videos. I didn't go as far as dokludi in attempting to build the whole helicopter. I discerned that the blade mount is solid and should be good for the load spinning fast. Hence a new generation of MOC helicopters is possible. The parts at the bottom of the ball joint links are divided into 2 sets, one for each pair of adjacent blades. Theoretically this could still be used for both collective and cyclic pitch but it's complicated. For collective pitch, which I think there is, 1M of rise in the mechanism is required. Comparing the various pictures it is possible to discern that this is provided. For cyclic pitch though, there would have to be a tilt in the bit in the middle, on the rotor's rotational axis, and that tilt would have to remain static as the rotor rotated. This is not possible given the four vertical studless beams, which leave only a cross shape, not a circle, as the possible area for blade pitch controls to move in. I think it would be possible to design a rotor system with both collective and cyclic pitch on the basis of the one in the set but that will be left up to us to do in a MOC. After all, the 9396 system will be an improvement on all previous sets because it will have 4-blade individually-actuated collective pitch. To have cyclic as well would justify another £10 in cost of the set when I hope TLG are still consumer-cost-conscious! I still measure Technic sets in number of functions vs. cost. Anything more than £10 per function and it's no more than a parts pack. The Unimog has 15 for about £150 (steering, 2 suspension, 3 diffs (i.e. 2 driven axles), engine, cab tilt, compressor, 2 function changers and 4-function crane). I don't count simple door opening as a function; the crane outriggers, motor and pneumatics are a welcome bonus. The helicopter has wheel retract, ramp, collective pitch, main rotor rotation, tail rotor and then starts to struggle to justify any more cost. However, my Apache, which has collective and cyclic pitch as well as collective tail rotor pitch from the rudder pedals, would have cost a lot more - it would also need months of rebuilding to get up to modern studless set standard! See if there's a helicopter round in the next Technic competition (next year?). Then we'll see who can do cyclic pitch as well! To remove that "holy grail" from the community by putting it in a set would have been a bad thing. People would have thought "Now that TLG have done it, there's nothing left for us to do". What we have is a big hint as to how to do it without the 8856 swash plate parts, which is the encouragement the community needs. Therefore TLG have done what's best for us and I respect them for it. I just hope the set price is keen! Mark

I don't know about "simpler" but there is always the way I extended Parax's Chain Precession Drive I quoted his original model as prior art for a patent. The important difference, in moving from the prior art to my invention, is that the chain moves when it was static before. Of course I built a model first, but I can't release the photos yet. I used 3 new-type turntables with 24T and 16T cogs. This is another way of putting a single drive through a turntable without the drive being affected by the rotation of the turntable. It may be used in multiple, given enough concentric shafts, hence Claim 12. This would allow the pitches of two propellers to be controlled from a static position without the pitch motor shafts having to rotate with the propellers. It is equally applicable to turret vehicles. Mark

Looks like the L-motor has a flat DBG fascia like the other PF motors, but with just 2 round pin holes (above and below the output shaft in Crawler use) compared to the M-motor's 4-in-a-cross. A separate 2x3x3 bracket provides 4-in-a-cross holes axially and 2 pin holes to each side in Crawler use. It looks like the DBG bracket might wrap around the sides of the fascia at 45 degrees or so - I'm not definite on that but it might not be necessary and would reduce breadth of use if it were so. The bracket side holes align with the rear motor side holes. The Crawler uses 2 grey bush pegs to hold the lower side hole on each side. I don't think it's a gearbox but it might facilitate more gears inside the motor casing by reducing the motor case volume used up by holes. It is still not known whether the L-motor definitely has the recessed orange axle slot like the other motors, or whether a 1M protruding axle (with an axle extender in the middle of the bracket) would leave more room for internal gearing. The tan top axle might be just a tan 6M axle, or a new 6M with bezel. 4M, 6M and 8M axles tend to be picked out together from a mixed bag unless one looks carefully whilst picking. Despite the reverse engineered MOCs, it is still not known whether the pre-diff ratios are 12:20 and 12:20:28 or 12:20:12 and 20:28. A low front view should reveal this because if it the latter set, there should be a 12-tooth cog visible between the light bley angle beams, protruding below the 3x3 4-pin cross block. There is a factor of 2 difference in gear ratio between the 2 options. I checked Philo's motor page to look at the required speed and torque characteristics of previous motors to see where the L-motor has to fit into the range. Given the performance of the reverse-engineered MOCs with PF M-motors, I would say the requirement for the L-motor, assuming 3 12:20 ratios and 1 12:28 for the diff, is 300rpm at 7.5Ncm of torque at 9V, giving a power factor (speed/60 x torque/100) of 0.375Nmrps. This compares with previous motors at 9V: 275rpm at 3.63Ncm for the PF M-motor = 0.166Nmrps 146rpm at 14.5Ncm for the PF XL motor = 0.353Nmprs 117rpm at 16.7Ncm for the NXT motor = 0.326Nmrps (same motor as XL) 920rpm at 2.48Ncm for the RC 5292 motor = 0.380Nmrps Looks like a motor similar to 5292, geared down by another 3:1, fits the bill. It has the highest power factor at 9V. Gearing of 6:1 instead of 3:1 would work if the first 2 gear ratios after the motor are 12:20:12 and 20:28. Coincidentally the L-motor case size is similar the the motor part of 5292's case! However, Philo also reported that the 5292 motor needed a lot of current. This would not be surprising for a 4x4 vehicle! The limit for a PF IR receiver motor driver chip is 600mA total or 400mA per channel. Philo said the 5292 motor could draw 780mA. I hope the IR receiver will be up to the job, otherwise I see Truck Trial vehicles needing multiple motor drivers, just like my heavy trains do! Other pictures seem to show that the Servo motor is either 3M or 4M wide, 4M high including brackets at the bottom, and 6M long. The motor casing shows signs of inheritance from the XL motor in clip design but I don't think the servo motor is as wide as that. It is probable that there is a gear ratio from the servo motor shaft onto a shaft below it, aligned with the centres of the uni-joints to the steering racks, with the gears located in the plane between the front edges of the 5x7 brackets, to which the suspension links are attached. It is the 9M studless beams between these brackets that provide the benchmark for the gap to the motor and hence the motor case width. This is illustrated well by dokludi's MOC. Whilst a servo motor with lots of torque (like the XL motor) might be desirable, that would need a 5-wide case so it looks like the servo motor is based on a smaller motor unit, maybe the M-motor one. This should be compared with RC servos, whose power is not great. It would still be enough for its 4x4 steering role, and enough to move train points. Mark

The differences in knuckles can be used to advantage: Oval Square Double loop Running on any of these would be easiest with a sprocket and a pinch roller opposite each other. The first two, and single loop derivatives, would make some quite interesting tracked vehicles. Mark

Such a brick cannot arrive soon enough IMHO. I have suggested it before. The bar code truck unit had a "hiss", engine noise and 3 others. The WeDo software has wav files for "Snore", "Cheer" and "Boing", which would be fun. The NXT functionality allows for sound file download but the live recording is what kids need for playability. Of course this would mean only 1 sound brick would ever be produced, which might reduce sales of Merry Go Round (tune), House (doorbell) and others. If bar code reading capability were added (put the emitter-detector pair in the other end of the brick) then a set of pre-recorded sounds could be selected for each set that used the sound brick. Different codes for different sets, till some bright AFOL reverse-engineered the bar code scheme Mark

Comparing the two: The 9398 Crawler has 4WD. It's been done quite a few times. I even used the 8856 helicopter swash plates to do front wheel drive before the 8880 car was released! The crawler's 4WD is based on the Unimog, which shows over 500 AFOL sales on TechnicBricks! Therefore many of us already have 99% of the parts, as demonstrated by multiple attempts to build the Crawler from the pictures. Having enjoyed the Unimog (only 2!), the Crawler is missing the pneumatic parts! As others have said, it is primarily the motors that set it apart from previous sets. I suspect the L-motor might be a repackaged 5292 motor (a motor powerful enough for RC buggies but now in a more useful package for Technic models, perhaps following the precedent of the NXT motor transformation to the PF XL motor using the same motor unit). The servo motor is the real prize - new functionality for TLG and something we have asked for for ages. Truck trial enthusiasts will buy the Crawler but if TLG don't release the motors separately soon, they will have missed the rest of the AFOL market (hope for January 2013 if we lobby enough!). The 9396 Helicopter has precedents for rotor pitch control in 852 and 8856. Pitch control is rare in the Technic range. TLG are always looking for something original to put in a set and this is sufficiently rare to fit the bill. It's really about time a helicopter was done properly in a set. I don't bother buying any helicopter without pitch control because it's lame. I know Technic can do it - I've done it, so I've been waiting for TLG to catch up with AFOL achievements in pitch control in the range of sets they release. Flying is something LEGO models historically "can't do" (without helium assistance) so the closer we get, the more it whets our appetites. Inventors always want to solve the impossible problem. Who knows whether a few 9396s and L-motors could generate enough lift to cancel their weight? There has never been a better set of aerodynamic parts from TLG - rotor blades, tail blades and the latest range of body parts. Anyone who has 8856 cannot fail to buy 9396. Compare those blades with the previous lumps of plates! 9396 also has swooshability - ask Space and Star Wars fans about that! 9396 has a future among us. How fast can you make a set of blades rotate safely? How much lift can you generate? How easily can you make sets of more than 4 blades? Notice that the blades are reversible, so if we use 4 sets contra-rotating, like a larger version of a £300 Parrot AR Drone, how close would it be to lift-off? Ultimately it might be the tail rotor blades that are closer to actual flight for the same reason that most flying birds and insects are small. Lift = Area = L^2, Mass = Volume = L^3. If a LEGO model could ever fly unaided, I intend to do it! Both the motors and the aerodynamic parts are proof TLG is listening to AFOL input. That is the best prize for the community. TLG know we buy sets as parts packs. If the Helicopter had had the new motors, the Crawler's sales would be hit! Mark

Would also be good to get an underside view of the servo motor! Mark

I wouldn't be surprised if it uses the new 28-tooth turntable as the means of getting static pitch control into the rotating frame of reference. Then some of the parts at the rotor head transfer the vertical motion from the top turntable piece to the 4 ball joint links. See how many parts it takes to do it if you don't have the swash plate from 8856! Unfortunately that also means little or no cyclic pitch, unless they've done something clever with levers, to pivot the turntable about its centre in one or both directions. I did something like that with a large turntable in my swash plate pump. Mark

Yes, I think it might be a 4M axle with 1M, a bezel and 3M. Otherwise it would be too similar to the 5.5M one. I suspect there is an axle extender inside the dark bley bracket, and either a 2M axle into the motor (orange bit as M-motor) or the motor has a protruding 1M axle (leaving more space for internal gearing). This arrangement would carry high torque better than a twistable axle alone. Comparing motor case sizes, I wonder whether the L-motor is a repackaged 5292 motor. Chop off the extra gearing and fixing holes and the 5292 motor would look like a 3x4x5 package with plenty of torque. Remove the dark bley bracket from the L-motor and it is a 3x4x5 package with plenty of torque. (If it did not have plenty of torque then TLG would not have used it to drive the wheels of a large vehicle, or would have used the M-motor instead). Maybe the 4 12:20 gear stages and 20:28 differential ratio provide a similar overall ratio to the gearing inside the 5292 motor. The 5292 motor was good for RC buggies but awkward to use with large Technic wheels or standard drivetrain structures (now including the Unimog structure), so a new case on this sort of motor would make sense. Of course 2 separate engines on the 2 axles is not a road-legal configuration in the UK, but it might be OK for off-roading, depending on the event rules. Some RC cars have a toothed belt between the two axles, but they have more powerful motors where the heat from 40 Amps can be dissipated. The vehicle body of an RC car is also lighter. I would rather have the LEGO body parts (which are better this year, being more aerodynamic) than a single-piece moulded top. Mark

I agree, it would be nice and most models would work just as well with IR, because both IR and RF would have limited range. At the end of the range the Crawler would stop with the steering straight because a lack of signal would lead to return-to-centre from the servo motor. You're right about motor use - most AFOL mechanisms have stalled the motor at each end-stop. Those that don't, which use a white clutch cog, lose synchronisation between the motor and the mechanism, hindering return-to-centre as a command i.e. the centre position accuracy is at the mercy of the spring. I'm guessing the PF servo motor would include the whole servo circuit within the motor casing, hopefully enough to turn a PF PWM command into a servo position and compare that with the shaft rotary position (I do hope it's more than a 3-position servo). I wonder whether TLG have used a potentiometer, like in ordinary servos, a resolver (RVDT) or an encoder like the NXT motors. I think RF (other than the existing system for LEGO RC cars) would cost more in development cost than these two motors, as well as in unit cost, and anything more than the existing RC system would be more hassle to TLG to meet all the FCC rules etc... for a multi-channel pack of parts like PF rather than the fixed protocol that a LEGO RC car has. We know what we're like - as soon as we had RF we would be asking for a 6-channel system for a helicopter Speaking of which there is an ideal vehicle for that in the forthcoming range... The limited range of control for a 4x4 is the reason why the set has a bang-bang controller. If it had a train handset (and if the servo motor could do proportional control as we are hoping) then the IR Receiver would continue with each PWM speed command, changing only when it receives another. That would mean the Crawler could go out of range and still continue to operate - potentially dangerous if it is near a road with children or a dog chasing it. With bang-bang control the model will stop straight away if it goes out of IR range. Remember that (for the moment) the LEGO 4x4 is still a toy, not an RC scale model, so a short range is appropriate. We may wish for that boundary to be pushed back, but it would have to have too many metal parts to handle the stress and the 40-Amp motors and speed controller would get seriously hot (melting ABS), so it wouldn't be LEGO any more. I would like very much to make a LEGO model fly without helium assistance; if I ever succeed then a longer range with RF would be appropriate. For the time being I look forward to trying out the new servo motor. We couldn't use RF very well unless there were a servo to control! Mark

Good work Denny In this picture: you need to use a 5.5M axle on the level below the top 12T cog. The items on the axle are: 0.5M of nothing, 1M of 3x3 toggle joint with 4 pegs, 1M of 20T cog, 1M of a light bley T-beam, where the "T" is the right way up, 1M of #2 axle joint, (0M) the axle's bezel and finally 1M of 12T cog. The 12T cog on this axle meshes with the 20T cog below it, which in turn meshes directly with the differential gear. Once you have sorted the pegs as well this should work better. With a PF medium motor, the predicted maximum wheel rotation rate is 400rpm x 81/875 = 37rpm maximum, most likely 30rpm at no load, maybe 15-20rpm on a smooth flat floor. The increased torque (and hopefully speed too) of the L-motor will hopefully make a difference to performance over the M-motor. You will need some of the 2x3 cross blocks with 2 pairs of round holes for the ends of the steering rack arm, with a short rack in the middle, held together with 2M axles and 3M half beams. If TLG used the new cross blocks from the logging truck then you could try 4M half beams for now. Mark

From the pictures and video I discerned these things: Looks like the form of the drive motor is 3-wide x 4-high x 5-deep, with a 3x3x2 bracket on the front, in keeping with the ones for Linear Actuators. The bracket provides the side-facing holes for the front motor mount, to complement the rear holes in the motor casing. I suspect there might be an axle extender inside it, such that there might be a 2M axle into the motor (or a 1M axle protruding from the motor, leaving more space for gearing inside the motor casing), with a 4M axle carrying the 12-tooth cog, which might be the tan one. The reason the steering seems to be moving to the end stops and the middle only in the video is that the person demonstrating it is using a bang-bang controller. It remains to be seen what the response of the servo motor will be to a PWM input from a train controller - whether its position will be proportional (either linear or logarithmic) to the duty cycle of the input. At 23 seconds, evidence that the steering motor is the middle one. As the rear wheelset has tilted, its steering has been affected. This seems to be the only way to get different steering angles from the front and rear wheels as they are usually the same angle from straight. At 27 seconds, see the 8M axles with cross blocks onto possibly the new 5-hole cross block that I assume connected to a short rack piece in the middle under the motor. See also the 12:20 gear ratio in the chassis. There is another 12:20 in the reduction hub, seen later. It might be that the 12-tooth cog seen on the front has a 20-tooth cog beneath it. From there, there is either another 12-tooth cog under it, in which case no more reduction, or more likely a 12:20 ratio 2 studs further in, the other side of the light bley T-frame. It might be that the 20T cog of that interfaces directly with the 28T of the differential. Therefore there is one differential in each wheelset , and no centre one because the steering servo motor is there. The L-motor speed must be either 9/35 or 81/875 of the speed required for that size of wheel, with torque to match. Given that the name of the set is "Crawler" we should not expect it to break speed records, but the motor speed and torque combination looks encouraging nonetheless. Regarding the servo motor, it would not be difficult to make it return to centre actively. The PF system has a PWM signal on C1 and C2 and fresh power on 9V and 0V. Therefore the servo motor could include a servo chip to take the PWM signal and set the motor current such that its servo position was in proportion to the PWM: positions of +/-1 to 7 or centred for zero (float and/or brake). The 9V and 0V pins would be used to send power to the servo chip (they are unconnected in the M and XL motors). If it is done this way then the servo chip and motor would be all one piece as an off-the-shelf module. I would not expect an RTC module to be separate (like the hockey player block some of us have used with the M-motor). If it does include a servo chip on a PCB then that would be one reason why it is so large! 2 new motors and a new mounting block (on each L-motor). There is also a new type of cross block seen in the logging truck, which might also be used in the steering on the Crawler. Technic has a good run of new parts with the Unimog so I'm not surprised if there are fewer new parts this year. Every other Technic vehicle seems to have a fake engine. I've lost count of the number of cylinders of fake engine I could build now! Fair to say that a truck of this size would find the cylinder size in better scale proportion though, as does the Unimog. I hope TLG would release the L-motor and Servo motor as separate items in due course. I don't expect they would be available that way before January 2013. We should petition Jan Beyer and the other community support people for this. If the servo motor could do proportional position control then I would want quite a few. I would also like to see how the helicopter gets on with a couple of L-motors driving the rotor! Mark

I measured the track gauge at 37.8mm. If you're building to scale then the scale depends on which track gauge you are modelling. If it's a standard gauge of 4'8.5" (1435.1mm) then the scale is 1:38. Conveniently 1 stud to the foot is 1:38.1, so a 10ft wide train is 10 studs wide. Less conveniently the large wheel sizes of some steam engines would struggle on standard curves. The BBB XL size is the largest that will go round on standard curved track. The standard BBB or Emerald Night driving wheels model wheels of 3'9" but, with a tolerance, anything between 3'6" and 4'6". That's why, for wheels larger than 5ft, I suspend them off the rails and a large steam engine becomes like a bogie vehicle. It makes a minifig shorter to scale than the average person, but you can always add a plate or two to the legs, or use Woody legs from Toy Story sets. I've built quite a few trains to the scale. For large wheels I used to use 40-tooth cogs but more recently I have made my own from Model Team wheel hubs. Mark

I have used 5M liftarms and 1.5M dark bley Technic pegs for a while now and the scheme works well. It's a length-for-length replacement for a coupling set. One reason I used it was to avoid train detachment followed by the loco crashing into the portion that was left behind, next time it goes round the circuit! It's also good for keeping a rake of wagons together. This coach, of which I have 4, is heavy and the scheme works well with them. The train takes 1300mA and four motors to pull it, so it's well tested for a heavy load! Another advantage is that the parts are common, so if one breaks for any reason it is easy to replace. A lot cheaper than coupling sets too! Mark

The 2-wide brick shape is an interesting take on the problem, and makes for simple track. I decided to be a bit more adventurous in 2007, so I used the 5-wide tracks from Technic sets. My monorail train will do a 360-degree barrel roll and I have a loop-the-loop to test. I also have a scheme for points, hopefully more robust than those of the original monorail! Pictures here Video of flat circuit here Video of barrel roll here I might try it with the PF AAA or LiPo battery boxes. A lighter battery should help the climbing and loop exiting ability. Mark