Brickthus

Thanks With pneumatics it's easier than with hydraulics. Hydraulics definitely need a fluid return path to the tank but pneumatics can get away with selective leaks to the atmosphere. The way to do this in LEGO pneumatics is to always feed the air supply to the centre nozzle of a switch valve. Do the same for a cylinder nozzle that needs to be pressurised in more than one state of the logic gate. In the OR gate: The top nozzle is fed to the centre nozzles of the switch valves. I actually have the OR and NOR the wrong way round . This would soon be fixed in a model! For simple control in industry, especially where the driver controls the functions, a 5/2 valve can be easier but the LEGO 3/2 valves have an advantage in being able to make these circuits. Of course the logic is in PLCs and FPGAs in industry these days so the exercise of pneumatic logic is educational and done "because we can". Industrial logic would also be synchronous, where these pneumatic circuits would be asynchronous by default. To make them synchronous would require a lot more parts for the extra gates! It is quite possible to make an adder with these gates, or even an 8-bit computer with a room-full, but watch out for the logic races, which are the order of seconds in pneumatics, as opposed to microseconds in electronics! Mark

How did you guess? There are still a few problems to solve but it depends on the scope of machine and also the compression capability. My pick & place robot needs 6 pump cylinders and hands soon get tired! Until TLG produce a spray-paint compressor product, the compression and hence machine size may be limited. A professional one of those may cost up to £500 but at least it would be quieter and allow longer running than a car tyre compressor. I might have to visit an art shop to see if they have one running, to sense the noise level. I bought 10x 8455, most on 3 for 2 at Toys R Us at the time. Great parts pack Some cylinders have a better top seal than others. Some leak air in if you put your finger over the top nozzle, extend the cylinder by hand and let go. Where the cylinder ends up shows the seal quality. Pneumatic robots tend to need a good top seal. I used ones with a poor top seal for controlling rail points; 2 cylinders end to end using the bottom nozzles only. I think of the "L" cycle as a gray-code machine, since the basic loop follows a repeating cylinder state sequence of 00, 01, 11, 10. I have found that 2-state logic automata tend to need about 2 valves per cylinder on average. A single cylinder could push 4 valves in a 2x2 grid but otherwise a pair push up to 5 in a line. My 4-function Octopus arm, which picks up an item on the ground (sea bed!) and uncurls in reverse to let it go above ground, has a logic board to remove most of the weight from the arm. This fills a 32x32 plate and one logic stage on the board has 6 cylinders pushing 9 valves! The total is 17 cylinders and 24 valves. I dread to think how many it would need for a full Octopus! I think each arm would have to be cut down to 2 functions rather than 4, which can be done with some Technic geometry. Mark

I think the addition of fewer parts per stage makes Technic building less complex. It means smaller children can build larger models, though this goes against the modern trend that children want the model built in an instant, rather than patiently plodding through 3-6 books of instructions. The amount of paper could be halved if the stages were done in the traditional way, though TLG decided to halve it by having the second model online only. Take your pick but, other than downloading some pdfs, I haven't built any online second models; there isn't enough space round my computer for the bricks! Comparing gear trains and pneumatics, pneumatics can do a whole function with just a cylinder and a switch, compared to the complexity of a long gear train. Gear train issues include keeping torque low till the last stage and manoeuvring round other items, both of which are done easily with pneumatics. The real criterion is that pneumatics have been used simply in sets, with only one exception in 8868's alternative model. In no other case in a set has one pneumatic function initiated another. Unfortunately TLG has decided not to boost the education of the community in the complex pneumatics department in 42043 or its alternative model. I await developments for future pneumatic kits. If you would like to do complex pneumatics, try these problems: - Start with a few building blocks including logic gates. - How to make a pneumatic steam engine (this is where 8868 began). - How to do multiple automatic functions with pneumatics in a single cycle. - How to make a pneumatic stepper motor for remote rotation where an electric motor could not go in real life (e.g. hot environment) or for a LEGO steam loco. - How to have multiple pneumatic functions that are operated different numbers of times in a cycle. - How to stop a cylinder in the middle, and cascade that functionality. - How to stop a cylinder anywhere you want it to stop, and react to the load. - How to make a pneumatic flight simulator platform and servos for it (2 work, 4 more to make). - How to make a more realistic pneumatic steam engine with variable speed (potential to be developed). There will always be more to learn in solving the problems you pick for yourself! 8851 was a great introduction to the idea of multiple pneumatic functions. Unfortunately the Mk1 pneumatic system had a fundamental flaw; it was unable to drive two functions in opposite directions at once because the movement of switches in opposite directions caused all the air to leak out! The basic rule is: If you want to retain air pressure, feed the source only to the middle port of a pneumatic switch. The non-return valve was fun but when I tried emulating a non-LEGO fire engine by putting the input line in a cup, the middle to a cylinder and the output through a small tube I soon wrecked the valve - it wasn't up to the higher pressures of pumping water! 8043 also had a fundamental flaw in the misuse of LAs; it needed a rollover of LAs whose clutches were wrecked when the design flaw in the gear train caused uneven extensions of a pair of them. In the way of finding flaws in the technical aspects of the product, 8851 and 8043 have much in common! The Mk2 pneumatic system introduced a second nozzle at the top of the cylinder and this allowed complex automated machines to be made. Anyone for fitting the 1/11 cylinders of 42043 to a LEGO train? Mark

If you buy 3 of 42043, you can make a new micro pneumatic motor! Mark

Very nice! I would like to make a similar facility for some scale model wagons that have unloading levers that could be actuated from below. I would be using round 1x1 plates, presumably more difficult to hang on to in a conveyor system! Mark

According to a quick Google search, the Arocs is available with 12.8l or 15.6l engines. 6 cylinders means 2.133l or 2.6l per cylinder. The latter would have about 93.9mm diameter and stroke if the dimensions were the same. The diameter of the LEGO cylinder is about 9.5mm. The wheel scale, at 62.4mm compared to 385 65 R22.5 super-single truck tyres would be 1:13.169. The cylinder diameter should be 7.13mm at that scale. The LEGO cylinders would fit a scale of 1:10, which would use 81.6mm wheels. Effectively 42043 has a 37-litre engine! There was an earlier mention of the smoothness of the pneumatics. It may be down to a small compressor and restrictions in the circuit at the switch valves and the cylinder nozzles. Movement may be smooth in each single direction but there would be a lot of hysteresis one reversing the direction. Mark

The change in oil price has more effect on the transport cost of LEGO sets than on the price of the plastic. The price of the plastic is a very small proportion of the set price. If you take out all the Power Functions and other elements costing more than £2 each from Technic sets, then re-do the price per piece calculation, it should work out similar, with a factor for larger sets being a bit cheaper than smaller sets. The proportion of PF for the Volvo loader is high. For 42043 take out the motor, battery box and pneumatic parts. The proportion of panels is also significant these days - allow £1 each (less than the single-piece RRP). Mark

I also have a lot of these wheels but I bought an extra 50 hubs specially. If you have too many of them then you could always consider modifying some of the hubs to make train wheels, for a really Technic steam engine This one is a 3-cylinder engine, using the hole offsets to good effect to get the 120-degree rotation. More on hub modification methods here. I find turned parts come out better than other mods. I tend to do it mostly for mechanisms when standard parts do not provide what I need. Mark

I left a review of this set on the Shop at Home site. Overall not as good as some, mainly because the engineering lets it down and the price per piece is better in other sets, even accounting for the motor and pneumatics. It could do with the thinner wheels at the front, like the VW camper van. Those wheels have not appeared in enough sets, especially as a set of 4 or more. the trapping of the pneumatic hoses really lets it down. I tried the rear cylinder the other way round but then the hoses clashed with the stabilisers. It would be possible to open up a gap in the stabiliser system to accommodate the hoses. The rear wheel beams move about, which is sloppy. These could be easily fixed more rigidly. The chassis bends a bit in the middle. Needs extra beams for rigidity. If I remember (before I took mine apart) the uni joints liked to pop off the large LA under load. The motor lead was not easy to route to the battery box without it becoming trapped or fouling the steering mechanism. The steering rake is better than in some sets - this is a function at which Technic sets are usually poor. The best part of the build was the gearbox. There was a lot of adding panels towards the end - it even felt like a lot as I mirrored the green side panel builds as a single process. A good parts pack for green parts and pneumatics but not quite so good as the model it is meant to be. I would have preferred black or grey for the chassis beams, though I expect red was the colour of the prototype vehicle. I agree that the new 1/11 pneumatic cylinder would have improved the crane arm extension function. Perhaps a comparison is needed between the pneumatic cylinders in this set (1/5.5?) and the 1/5 cylinder in 42043. If the one in 42043 is shorter then this might explain the reason for doing this set, to use up the stock of older cylinders, especially if a newer design includes a reliability improvement. The 1/5.5 cylinder has been a bit awkward to build with as its extension is just less than 2M. I used it for some pneumatic stepper motors but they have to stretch a bit as the parts flex. Mark

From the video it looks like position 3/7 of the Servo motor is enough to turn the valve on at almost full air flow. I suggest you increase the gear ratio so that position 7/7 is where position 3/7 is at the moment, then control the Servo to position 6/7 to find a partial flow position. You have a gear ratio of 12:24 at the moment so 12:56 (turntable) would be OK, or 8:36. It might be easier to have a 1M lever on the servo e.g. 2x 2M "++" half beam, and a longer lever on the switch valve arm, extended with an axle extender. I used this method to hover around the partial flow sweet spot for my variable pneumatic servo. I used two valves in parallel so each cylinder nozzle can have a valve sitting on the sweet spot. Mark

I wouldn't recommend more powerful motors and batteries because the rest of the LEGO parts are not designed for it. You would need a conservative design, such as doubling up on gears at each stage and doubling the support of axles. Even an XL motor can twist an axle so an RS380 could do more. It's the torque that will do it though. Power = torque x speed. Might be OK with lower torque running faster but then axle friction may melt the axle and the beam. I once tried mounting a drill in a LEGO base to drive a static model but 135W was too powerful for the LEGO parts even with doubled-up gears - the axle melted. There is also the motor heating to consider - melting a LEGO mounting scheme. If I wanted a faster powerful LEGO motor e.g. for a propeller then I would consider removing one gear stage from an L-motor. I tried it with an M-motor with limited success. The LEGO LiPo battery has a current trip at 800mA. Some of my trains need up to 1300mA so it tripped and restarted on the uphill section. It is far better than AAs because it has a lower internal resistance. 600mA on one channel is recommended for the motor driver chip, or 400mA on each channel with both together. Mark

I don't think it's a 7M shaft; the shaft axles are a 3M and a 4M (see the teeth of the bevels are offset) so the central bush is on the 4M one and serves to keep the bevels on their shafts. I expect with 2x 3M ones they tended to pull out easily. The z24 crown wheel has been modified to cut down its hub so that it doesn't foul the diff cage. Mounting it without a beam behind it (see the half bush) means its engagement with the diff will be compromised under heavy loads. There is a single uni joint each side when 2 are needed to do parallelogram suspension, but it might be that the ball joint parts have been used instead e.g. the top shaft looks dark bley. See how it works and learn without criticizing Mark

I think the 5/3 panels would be useful as static vanes to go behind a fan in a large jet engine model. The fan would use the #17 or #18 panels that are more aerodynamic. No chance of a flat-pitch propeller causing significant forward thrust on this plane! Mark

Voila, from 2011: The lower crankshaft dithers the switch valves, which are connected in parallel and offset to sit on the edges of opening in each direction. The air supply is from the grey pipe. I used some non-return valves from BBB to isolate the servos from each other. The output to the cylinder is from the yellow and blue pipes. The feedback of cylinder position is via the flex that adjusts the lever and hence the valve position offset, with long levers to get a suitably small amount. The position of the valves is controlled by a large LA at the moment but I will soon try it with a PF Servo motor, which wold have a quicker and more precise response and can be controlled by IR handset or an NXT with IR link sensor. This servo drives 1 cylinder of my Stewart Platform. I have 2 cylinders on 1 axis working so far, with 4 to replicate. This will mean an NXT could set all 6 servos for the position of the platform, with the pneumatics allowing some position tolerance (compared to 6 LAs which would not). The lower crankshaft will drive all the servos with the same dither frequency so that they all have the same time response. I expect the longer pneumatic cylinders will work with this, perhaps by gearing down the flex return with a lever. The idea is that the length or power of pneumatic movement is far more than the flex return so the system can drive loads far greater than a flex. I contemplated a system where each load might have 4 pneumatic cylinders in parallel yet a single flex for the position feedback. Mark

Cylinders: consider the scale size of the piston and the spacing between them; what does 16mm represent at this scale, where 62mm wheels represent truck wheels? This would dictate whether the full 6 cylinders could be implemented in the model. Cars would have a bigger problem with this. It is quite possible that LEGO Technic cylinder blocks are suitable for ship engines. However, these parts are an improvement over the original 2x2 rounded blocks, which would represent even larger cylinders! I have robbed my built Unimog of its compressor. I originally added the parts from the 2nd Unimog to add the snow plough to the first. A single LA to raise the tipper seems weak if it were ever filled with LEGO plates. I expected 2x LAs but I'm all for reducing LAs in favour of pneumatics Whilst adding an air tank or boring out the holes on the cylinders (careful you don't damage the seal whilst doing the top one) might increase the performance of the pneumatic system, its application in this truck crane really needs it to move slowly like the real hydraulic crane. Higher pressure or reduced restriction would allow faster air flows but that would move the cylinders quickly to their full travel, with only the hand movement of the switch valves limiting the air flow rate. This would defeat the object of replicating a hydraulic crane. You have to pick your applications - modded cylinders for 2-state logic circuits and pneumatic engines; unmodded cylinders for replicating hydraulic systems. What does the small cylinder do on the crane of the B model? Does it release a pin so that the entire triangular arrangement of 6L yellow liftarms and grey curved panels can be set down? I dream, and make quite a few things "because I can" My swash plate pump can vary the swash angle and hence the piston stroke and volume. This is the nearest that can be done in LEGO because a real one (e.g. small 1960s jet engine fuel pump or early 1990s diesel car engine injector pump) often has the fluid surrounding the mechanism for cooling and lubrication. In the model the turntables have rather a lot of friction that loads the XL Motor. I might give it an update with the new 60T turntables from a couple of Arocs trucks in due course as these are likely to have lower friction. I also did a variable-stroke pump with 4x 5.5L compressors and 2x RC car motors that doesn't need turntables and does have a faster compression (no pics yet). This means the motors can have a rest at low stroke; they don't like to run under high load for long. All I have to do is add a feedback mechanism like this one to the swash plate or stroke lever - pressure vs elasticity to set the stroke. Mark

I would consider replacing LAs with the new 2/11 cylinders. I could increase the arm extension of my Pick and Place Robot with a 2/11 cylinder. 8421/42009 could do with two 2/11s end to end for jib raising and also several end to end for jib extension. The new 1/11 cylinders are more difficult to place since they might prefer slider applications where the sideways forces are managed separately. Certainly outriggers - 4 on 8421/42009 or mobile crane MKIII in a few years' time. the 1/11 cylinder would be placed easily enough in City buildings; they could make something stick out by a significant amount or retract as required. It would be great to see pneumatics used in ranges beyond Technic. The longer extension of these cylinders might help with train decoupling, though on-train compression and control would be difficult (L-motor for compression + servo motor for valve control = expensive!). Perhaps a scenery-mounted system with the cylinder(s) hidden under the track would work better. Real movements on a railway tamper would be great, and would justify the expense of the motors to control the pneumatics. There are great possibilities for Continuously Variable Pneumatics with the longer travel of both these new cylinders. It will be a challenge to improve accuracy and minimise play in the system, with more gearing down needed between the cylinder movement and the valve movements; for variable control hey operate on the edges of their nozzle opening. Whilst I might wish for an updated multi-pneumatic kit like 8455 Backhoe (of which I bought 10) as a parts pack as well as a great model, I'm not confident enough to get my hopes up. Rest assured TLG, if you do make another kit with many pneumatic parts like 8455 but including the new cylinders then I will be buying quite a few! Mark

If TLG were to make a new pneumatic valve compared to the "3/2" valve they have made for many years, then a "5/2" valve would be welcome. This is a standard part for industrial pneumatics, which currently takes 4x LEGO "3/2" valves to replicate its function: The valve would probably have a 3M diameter disc with an axle profile lever projecting from one side. There would be no leaks to the atmosphere with this valve, just as there are none with the circuit block above. Instead the 4 nozzles would be either blocked, connected 1-2 and 3-4 or connected 1-3 and 2-4. The existing elongated rubber-lined enclosures would be used, 2 with just a 1M-wide round enclosure for the "no flow" state with the lever in the centre-position and 2 with elongation around the circumference - longer than the current one used in the "3/2" valve, for connecting pairs of nozzles together in the positions either side of centre. The four nozzles would be arranged in a square pattern, with the square perhaps at 45 degrees to the orientation of the valve body, depending on what is convenient for the parts design. Such a valve would need the top of the lever to be further from the pivot centre, so a design criterion to consider is whether a #1 toggle joint on top would still allow a 2/7 cylinder to realise the full travel of the valve. The introduction of the new 2/11 cylinder would allow such a valve with a longer lever to be controlled without needing any gearing up in a lever ratio from a 2/7 cylinder, though a 1/5 cylinder would need it as its extension would be too short to cover the full travel. In that way the 2/11 cylinder now enables a larger valve to be designed, produced and used because we will have a simple, ready means of actuation for it in a loop system. The 5/2 valve would replace the above circuit block in a system like this one: This circuit is the one for my pick-and-place robot from 1996. The top 4 of the 5 "3/2" valves in the set driven by cylinders in parallel with the grab would be replaced with 1 "5/2" valve. Those 4 valves plus the right-hand one from the wrist pair for an Exclusive-OR gate; The grab, wrist and elbow are in a loop and the extension position dictates which half of the loop receives air to continue its movement. I would prefer this purely pneumatic valve to any amalgamation of a PF electrical part and a pneumatic part because the versatility of LEGO is one of its strengths - we can make our own choices which valve to drive with which motor or other pneumatic cylinder or whether to use a spring or belt to bias its position. A PF micro motor or solenoid with 3-position control (functionally like a 12V point motor) and capable of driving a "3/2" valve (i.e. a lot more powerful than a 12V point motor) would also be welcome; the PF servo motor can drive a few but £21 is a lot for some fans. I would also welcome an advance in air compression. I have tried various compressor configurations, some more for show as concepts e.g. Swash Plate Pump and some for raw air power, but a large power supply is needed to compress enough air for systems much bigger than the circuit above and I'm not sure the standard LEGO motors are up to it for long either; currently a single 1/6 compressor is driven from an L-motor so that's a lot of L-motors! I have previously used a car tyre air compressor but it was noisy, got hot, and it is difficult to find a suitable product in the shops for continuous use - most are designed to reach a pressure and switch off, or rest for half an hour after inflating a single car tyre, whilst others are capable of providing enough air for a whole house full of LEGO pneumatics (did anyone warn my wife about that )! Mark

Excellent! Nice to see a MOC with a pneumatic loop. Have you used loops in other MOCs too? You could consider doing variable pneumatic control for the steering! Mark

OK, I can stop being "mildly skeptical" now The turntable looks like it's 0.25M wider at each side by the overlap beyond the red beams. Compared to a 56M turntable at 7M wide this would be 7.5M wide i.e. 60 teeth, with bevelled lower edges like the 28T turntable as observed. That's good for ease of connecting gears to it - no more vertical axle with bevels underneath - simpler and less friction in the geartrain. I would have liked the new turntable to have a larger hole in the middle, perhaps room or a 28T turntable, but the fixing points still have 3M between them so this is most unlikely. The issue is the difficulty of getting 8 hoses through the turntable and still rotating it by a decent angle. I suspect only +/-135 degrees rotation either side of rearwards on the truck would be possible with 8 hoses. With the longer cylinders comes longer extension. This means more leverage to a mechanism, especially if feedback via a valve is used. The existing 2/7 cylinder would happily drive a valve with a #1 joint; the new 1/11 and 2/11 cylinders would need an extension to the valve lever to get the same angle range for full travel of the cylinder. This means a mechanism driven by the new cylinders would have gearing down for the valve feedback - not a bad thing, given that the valves have greater friction than the old light grey ones. Glad to see the 1/11 cylinder has a metal shaft as it would otherwise be prone to bending; the slider parts also take sideways forces away from the cylinder. The 1/5 cylinder looks like the existing one. From the box pic, the compressor looks like a 1/6, a blue version of the Unimog compressor. Perhaps the L-motor can drive it faster than the M-motor of the Unimog could. Slow actuation might not be a problem because real truck cranes are operated slowly for safety reasons. The test will be if a 1/11 or 2/11 cylinder is pointing downwards and extending to lower a load; if the top seal is no better than the existing 2/7 cylinders then the load may be lowered inadvertently (when held) or too quickly (when lowered) as air enters via the top seal. 42043 at least meets the £10/function criterion: 2x steering, 4x suspension, 2x diff, 1x engine, 4x pneumatics, 1x compressor, 1x tip, 1x tip flap catch, 1x motor, 2x stabilisers (extend, drop) = 19 without counting the cab tip and door opening, so £170 has good enough play value for money. New parts value is also good, with a turntable, 3 pneumatic parts and the clam shells; I'm sure someone will use many clam shells in a GBC! Nice to see that the grey curved panels are the old style ones. I prefer those to the new ones (like the blue ones in 42042) because they retain a more aerodynamic appearance. Good use of the hose brackets on the crane; still room to reduce the balloon effect by using more 3.2mm tubing as 8868 did, thoguht eh Unimog tubes were stiffer; I was worried that the nozzles might break when detaching hoses from cylinders and valves but it's OK using nails to prize them off slowly. The front wheels of 42043 bowing outwards looked a bit dodgy in one pic - hope the front suspension can take the weight OK. Might have needed the Unimog wheel support brackets. As for buying more than one, I would need to find a use for many wheels as they are a significant proportion of the cost; a truck like this with Unimog wheels wasn't going to happen as it would be twice the parts count and cost; maybe in 6 years when the community is clamoring for a 5000-piece Technic set? I already have so many 6-AA battery boxes too. The grey panels are useful for Space models too, and could accompany my monorail that has a Technic basis. Glad there are not too many red beams as I have plenty of those - I usually build a chassis in black, grey or dark grey. The L-motors will be useful and I could never have enough pneumatic parts; the latter is limited only by the ability to compress air. Since it seems to be 1 compressor per L-motor I would need 10 L-motors each driving a compressor to power larger automated pneumatic models. Since I have many 2/7 cylinders I will have to consider where 1/11s and 2/11s are most useful; rail points use 2/7s end to end quite well for the power requirement (a 1/5 does not have enough power) and they don't need a long travel. Signals would be happy with 1/5s. Perhaps I adapted my models to use what was available previously so it may take more time to expand to make full use of the 1/11s and 2/11s. A steam engine with larger wheels and a crank length of 2-3M might be a good start - using 4-6M of cylinder extension. That should have a fair amount of power. Maybe my Stewart Platform could be enlarged. In the meantime it will be fun experimenting, which is one of the most enjoyable aspects of the hobby. I would hope for a crane with the jib raised by end-to-end 2/11 cylinders in a couple of years - Mobile Crane MkIII! That is also an application for many wheels - I've seen a real one with 9 axles before. Someone mentioned a fork lift truck - the 2/11 cylinder is ripe for an 8843 update. I'm torn between saying "Well done TLG" and "It's about time" I expect the new cylinders will be put to use in the community a bit faster than Power Functions has been; apart from simple motor drives, PF is only just picking up after 7 years! I think Technic is ahead of Trains in this respect as Technic fans are not still in denial and trying to replicate 9V, nor is the need for a slave motor driver quite so pressing as it is with a multi-loco train! Mark

Perhaps the crawler crane will become a regular model type in the Technic range as the 8-wheel truck is becoming. I will remain mildly skeptical about the pneumatics in 42043 until I see the pictures for two reasons: 1) Preliminary boxes can have images different from the final product. This has been borne out previously with the PF motors for the Crawler 9398. Toy fairs often have preliminary toys, not polished final versions. 2) The cost of pneumatic components and their development is high. We have had 1/5.5 and 2/7 cylinders for many years and whilst it would be logical to add a 2/11 to the range, it would cost more to change the original cylinders. a 1/5.5 is OK for a grab and 2/7s have done OK as lift and elbow actuators for cranes. The only missing function is an arm extension. Contrary to that is the idea that a new subcontract partner for this type of actuator could make a comprehensive range offering. Certainly there are some improvements that could be made to the existing cylinders: - Better seals, especially at the top; I sorted my 2/7 cylinders into groups with good, mediocre and bad seals. - A metal shaft for the 1-wide cylinder. This would facilitate longer 1-wides such as 1/11. - Integerisation of the length and extension measurements. The 1/5.5 extends by just less than 2M - awkward if you want to make a steam engine because cranks don't come in "just less than 1M" sizes, so 1/6 with a full 2M extension like the Unimog compressor is overdue. I had wondered whether TLG were testing the 1/6 concept on the Unimog compressor (lower production volume than cylinders) with a view to extending the 1/5.5 to 1/6. - Improvement in the L/d ratio - more extension per unit length. The 1/5.5 extends by just less than 2M so its L/d is about 0.35. The 2/7 extends by 3.5M, so its L/d is about 0.5. A pair of 2/7s with brackets begins as a 2/15 and extends by 7M so the L/d is about 0.47. It would be good to have L/d of 0.67 or more in a single (non-telescopic) cylinder design without having to go wider to mount the cylinder in brackets with pivots half way up. - A wider pump cylinder or compressor, to get more per revolution of the compressor shaft; previous compressors in set models have been slow with just one 1/5.5 (8868) or 1/6 (Unimog) compressor. - Some people would like a colour other than yellow. I'm happy enough with yellow as it's good for construction equipment and models of other colours get to show off their actuators by the contrast. I'm not sure silver would be a good idea because it's a more difficult plastic colour to work with; grey would be OK. There is another possibility though - silver could indicate actual metal, which would mean stronger actuators, though that would make it more expensive so it is unlikely. Of course if there are really 3 types of new cylinder and a new compressor then it will be well into 2016 before they would be available separately - if they come up on the TLG parts site in January 2016 then they will sell out instantly to people who would not want to build with them but would sell them on (knowing when to buy is a key skill in the marketplace). Therefore I hope TLG make a good stock before selling them individually. I guess the prices may be higher. The 1/5.5 was quite good value at £2.23 compared to its price as a spare part in the 1999 leaflet. I got a few for doing railway signals as I use the 2/7s for points but the signals would not need a lot of muscle. Pneumatics have a clear advantage on truck loading cranes because of the difficulty of getting several shafts into a rotating domain for the LA alternative. Two shafts have been taken through a turntable for an excavator but it is a large mechanism. This difficulty and insistence on using LAs has resulted in some poor cranes with some LA functions and other manual hand controls on the cranes in some recent sets, including 9397's grab and 42038's arm extension. A 1/5.5 or 1/6 pneumatic cylinder is also easier to mount in a small space or on the end of something because the mini-LA is wider. Since there are pneumatics in 42043, I wonder what would be a worthy B-model. A refuse truck with looping pneumatic system is 24 years overdue since 8868; this time, with Technic panels from the tipper body of 42043 allowing a body structure to be made, the loop could be employed for the refuse crushing function. Of course pneumatics are not just end-to-end devices if you are prepared to add some complexity. This makes even more sense for adding versatility to the operation of a longer cylinder. In some cases this makes their operation capable of emulating LAs. I still suspect LAs would be used for tipping 42043's body though. The same LAs could tip an alternative body for a refuse truck. It might also be tyhat the B-model uses the same chassis with another industrial function. Whilst a skip hook has been suggested earlier, any body shape could be done, even a cement mixer with a pneumatic loop driving the rotation! Mark

Very nice. Would be good to see it on the LEGO Ideas site. TLG may be looking for new pneumatic model ideas now that 42043 will have new pneumatic parts later this year. Generally Technic sets are £10 per function but also have their price in proportion to size. As you have packed the functions into a smaller space that might improve functional value for money! Mark

The old one did 25rpm with 0.3W of mechanical power in a 2x2x2 volume. The key challenges are getting a motor in the small space and getting a high gear ratio in the small space with the whole thing having enough starting torque. In theory a harmonic drive can give 300:1 ratio in a single stage in a small space but it would need a flexible element, probably nylon. Possible with 3D printing but needs reliability. I would be happy to see a short version of the PF M-motor, but a micro motor needs a higher gear ratio, not a lower one. Therefore a shorter M-motor would be better as a faster motor e.g. for propellers. Mark

I take the point you are making about lower voltage but an explanation might be beneficial A dc motor like the PF motors would have a speed proportional to voltage. This is not quite the case with Pulse Width Modulation from the IR Receiver or LiPo battery as the PWM gives better torque at low speeds. The electrical power is proportional to the square of the voltage, so 7.2V (6x1.2V) gives 5184, 7.4V (2x3.7V in the LiPo) gives 5476 and 9.0V (6x1.5V) gives 8100, each divided by the nominal resistance of the motor windings. On that basis I had wondered whether the LiPo at 5476 would be able to power models as well as a battery box of 6x1.5V cells at 8100, given that the LiPo's number is 0.676 times that of the 1.5V AA cells. For a train it may be that the maximum speed is lower with the LiPo but trains that operate at the full 9.0V (or 9.15V with very fresh cells) could easily derail on the sharp corners. This means full speed is not used so often (by AFOLs; I accept that children would like to see how fast the train goes!) and hence the LiPo's current advantage (800mA cut-out limit) shines through (till you overload it with a big heavy train!). The internal resistance of ordinary Alkaline cells means they struggle to deliver the current whereas the LiPo has a low internal resistance. I have previously given reluctant Duracell cells a brief short circuit through a multimeter on the 10-amp setting to see the current rise but this is not generally recommended. Then there is the cell/battery discharge characteristic to consider. Alkaline cells reduce in voltage fairly linearly till they reach about 1.35V (typical for Duracell), at which point they are "flat" and can no longer deliver the current. 6x1.35V = 8.1V, which is considerably higher than 7.2V or 7.4V! This means a lower voltage does not necessarily mean the train goes slower. The motor characteristics also matter. Typical LEGO motor operation for 9V motors starts at 3-4V because they hardly move below that; the product is designed to power models, so the outputs from 9V train controller and the PWM from PF parts tend to fit that range. The most efficient use of most dc motors is where the current is equal to twice the no-load current. This may be easier to measure using a steady 9.0V than with a variable voltage or PWM scheme! If the model goes slow, change or charge the batteries or add more motors in parallel. Mark

The 4-wide battery boxes are only a restriction to the narrow parts of the bodies of diesel locos. In some cases a bespoke casing colour was used, similar to the practice for the 9V PP3 battery box. The latter i still an option for low power items - probably excluding trains except fora shunter on its own. There have been ideas such as a different shape of LiPo battery box but it consists of 2x standard 3.7V cells and a control part. At £43 the cost is quite high enough and could not justify being rearranged into a longer, flatter format, even thoguh that would be easier to hide in a train coach. PF has a good range of motors but there are still gaps in the range. We have slow high power (XL), torque for big vehicles and locos (L), general purpose (M), generator/high speed & low power (E), steering/point control (servo) and the best train motor we could have wished for. A replacement for the 9V micro motor (click on translate) would be nice but its original inventor died :( A shorter version of the L-motor with just a single gear stage might be nice, for faster medium power for a propeller. At least there is scope to add to the range. Adding to the range of PF-compatible lighting options is a bit easier than making our own motors, hence my railway loco and signal light experiments. Mark

I have asked whether here could be a light brick with alternate lights, rather than both being on at the same time. This would be great for red/green signals. In the meantime I'm using red/green 3-lead LEDs with a circuit conforming to standard PF protocols that can also make a good yellow, the 3 colours R, Y, G corresponding to up, middle (off) and down of a pole reverser switch in a panel. Will any further lights with a conventional PF plug input be forthcoming? Mark