Saturday, March 13, 2021

CaDA C61042W Italian Supercar: A Serious Exercise of Power




It shouldn't surprise you to hear that, as a standard-issue LEGO fan, I felt quite an inherent resistance against LEGO clones for most of my bricky journey. And I can comfortably say that for the large share of that period, that resistance was well founded: those clones were regularly of such low quality in so many respects, that trying to deal with them felt like a grotesque travesty of what proper LEGO experience should actually be like. 

But, throughout several recent years, the scene began subtly changing its shape. Yes, there were (and still are) truckloads of inexcusably horrible cloning companies, blatantly ripping off official sets and community's ideas, and those are the manufacturers I truly don't want to have anything with. But interestingly, a small but significant number of exceptions floated to the surface. Exceptions which did not hide their LEGO-based fundamentals, but at least showed intention to do things properly, to the extent possible. And one of them is CaDA, whose Italian Supercar set received some unexpectedly strong praise. Sufficient praise, mind you, for me to cautiously order it.

Honestly, I had no idea whatsoever what to expect: this is among the most expensive clone sets one can find in the market overall, yet compared to its functions and a staggering number of parts, nevertheless it seemed almost suspiciously cheap. The only sensible reference for comparison I could put it against were LEGO's own contenders in the field, id est, its Technic supercars of comparable shape, complexity and role (among which I was lucky enough to have the heavyweight Bugatti Chiron at hand).

And so on one rainy Friday, CaDA's package arrived at my door, inconspicuously hidden in a plain beige box. 


No love at no first sight

Now, it would be easy to just dive into a "classic" overly comprehensive review which enumerates every single trifle in great length, but that would be completely pointless. Not only because there already are many such police-investigation-type reviews out there, but because I want to focus on the actual experience of this set, and not on the finnicky technicalities that tend to get in the way of essentials.

Sparkling with curiosity, I opened the huge box with a nice motif of the car's bodywork, and was facing a group of smaller boxes, all densely packed with parts, fitting snugly together in a neat slanted pattern. This is nothing new; indeed, LEGO does the same for its flagship sets. But for a clone, this is unusually sophisticated, thought I to myself.

So, we've got enumerated boxes with, furthermore, enumerated transparent bags inside, all refering to major building steps in the six instruction booklets. Sounds familiar? It should, because this is how LEGO's sets have been organized for quite some time already ― and to that extent, I should say, following its successful recipe makes sense. All that remains is finding a suitably large building surface, and we're ready to rock. 

History, branding... and the lack of them

Before proceeding to the set itself, a word about its background story. For this project, CaDA hired Bruno Jenson as the designer. If you have been diligently navigating the Technic community waters, the name should have a familiar ring: that's no other than Brunojj1, the well-known member of the Technic community, renowned for his widely acclaimed Technic cars. Consulting one of such distinguished masters totally makes sense if you're trying to make an impact in the Technic (or rather, Technic-like) world right away. And Bruno's authorship is not hidden in any way ― to the contrary, his signature is prominently featured in many places around the box, instructions and elsewhere.


As for the model, in everything but its name, this is clearly a Ferrari 488 Pista, the 720-horsepower Italian beast. Official Ferrari branding, even if it were possible (because LEGO may have an exclusive contract in place), would probably add a hefty chunk of price to it, just in return for some prancing horse stickers ― therefore, the decision to omit it was to be expected here. On the other hand, LEGO would probably have no choice but to cringe and license their model, since selling an obvious replica would cause them more harm than good.

The choice of a 488 among several Ferraris is an interesting one. It may be down to someone's personal taste, to its difficult bodywork to model (though ― any Ferrari would be difficult), or to something else entirely. 

Tremble to assemble

So, back to our crafting table ― and several striking facts surface immediately, before even finishing the first instruction booklet.


As you are probably aware, LEGO follow certain coloring rules when dealing with the elementary Technic parts, especially connectors. For example, 3L friction pins are usually blue, frictionless axle pins beige, and 2L axles red. The rationale is well known: it makes the parts easier to find while building, and also easier to sort during disassembly.

To put it short, CaDA does not, and it relies on black almost exclusively, thus producing two effects: disassembly of this set will definitely be a headache, but on the other hand, the set looks neat and consistent ― both its bodywork and interior.

Furthermore, while the "sea of black" may look tidy, it makes the assembly a bit harder, as it may be occasionally difficult to get proper bearings when trying to find your way around a million black items in the instruction booklet. But more on that later.

Another observation is that CaDA does not shy away from very advanced building techniques, which occasionally cross slightly into the territory considered illegal by LEGO. Mind you, they are still perfectly valid and widely used among the community builders, but as we know, LEGO go to great lengths to avoid even the slightest part tensions. Here, concepts like a half-inserted friction pin or a stud inserted into the axle hole are simply and pragmatically accepted ― whether you like it or not. 

Mastery of design

And soon it becomes apparent, and stays so for the remainder of this building adventure, how insanely designed this car actually is. I'd go as far to claim that ― in the world of LEGO and its clones' sets ― this is easily the most sophisticated mechanical design I've ever encountered. The designer went to great lengths, accepting the cost of increased construction complexity, to follow many subtle curvy shapes of the original Ferrari perfectly. The amount of modeling skill and care invested into the bodywork is just tremendous.


And for the most part, its underlying chassis is strong and clever enough to keep the bodywork sufficiently rigid to easily survive usual driving and play. As many of us have learned, reaching either the trueness of design or the construction strength alone may be difficult at times, but is within the grasp of most builders. However, hitting both targets in the same model, like this Ferrari does, requires some impressive skills, as well as lots of patience. The cabin interior is just as impressively detailed, and contrary to many examples, it is free from various cables, axles, or other stuff.

Yet once we dive from the bodywork deep into the mechanics, the story just keeps becoming crazier.

So just to give you a clear picture, what the designer managed to squeeze underneath that glorious 1:8 bodywork are: remote steering, a remotely controlled sequential 4-speed gearbox with working steering wheel paddles, full independent double wishbone suspension, a V8 engine, remotely controlled lights, openable engine cover and doors with working handles, and a remotely powered drive.

Yes, while you're getting up from the floor, you did read that correctly: there are two servo motors (for steering and the gearbox), and two PF L-equivalent motors for drive, along with the lithium battery powering all of that, somehow crammed in the car along with all those other mechanical features. Density of features is breathtaking, even if you are otherwise familiar with top community builders' work.

The gearbox alone deserves an article on its own, such is the engineering prowess displayed here. It takes quite some concentration to grasp its functioning, featuring two synchronized gear changers that produce four trasmission ratios. We have already seen gearboxes with more speeds, but a sequential one powered by a servo motor, of this compact size yet considerable strength, breaks the ice.

It's not only about the gearbox, however. At many steps in the entire building process, I found myself wondering at how something is supposed to work, only to have an "aha! How clever" moment few minutes later. 



Blood, Sweat and Tears

As one would expect, all this sophisty doesn't come for free. Quite simply said, this is a difficult set to build. Very difficult. And not only because of the experience and fine motor skills required to assemble some complex mechanisms and bodywork pieces, but even more because of the concentration which should be kept at maximum all the time.


Namely, it is very easy to make a mistake ― and at this level and density of features, any oversight will probably cost dearly later. This leads to what is, in my view, the main weakness of this set: its instructions. Only once you've tried the subperfect instructions can you appreciate the level at which LEGO do it.

I wouldn't say the instructions are downright bad ― after all, I managed to build the car ― but there is plenty of room for improvement to get to the standard LEGO level. For example, not all steps are perfectly clear; as a remedy, you can easily turn a few pages to find a suitable view later where some questionable part appears in its correct position, but the very need for that is an unnecessary complication, especially if there already is some space in the page for a proper "before & after" for the detail in question. And having so many black parts squeezed together does require an eagle eye to spot them. (Or preferably, two eagle eyes.)

Also, sometimes they progress from left to right, sometimes from top to bottom; due to monotonic colors of the innards, the chassis rotations are not always emphasized enough; and then there are pure errors: parts which are displayed in wrong positions, or are not there, or are there prematurely. Interestingly, instructions show the parts as they are manufactured by LEGO, rather than by CaDA: as if the instructions were created first, and the actual part design only completed later. This is in no way a problem for the actual building, but is a curiosity. 

Bold mold

As for the quality of the parts themselves, the story is somewhat similar. These are, hands down, the best quality parts I've worked with among the LEGO clones, but at the same time, still trailing LEGO a bit.


It boils down to minute, precise tolerances. While the average force that holds e.g. an axle in an axle connector is roughly the similar to the one in LEGO, the variance here is noticeably wider. None was so bad that it would fall apart by itself, or that it could not be connected in the first place. But once you got used to LEGO's strict tolerances and consistent clutch forces, CaDA's seem a bit rough. Again, it does not affect the look or the end functionalities in any way, but affects the overall building impression.

Several reviewers have reported occasional miscast parts ― I haven't, but I cannot be sure whether I was just very lucky not to get any, or has CaDA improved its manufacturing process in the meantime.

There are also slight inconsistencies between the parts' color shades themselves: when observed next to each other, as they anyway end up in the finished model, even I with my poor color sight could discern slight differences between the various shades of reds. It's not a showstopper, especially if not under a strong light, but you may want to be aware of it should you plan to use this set as a showpiece. But then, to be fair, this is a criterion where LEGO's own performance has occasionally slipped in the recent years as well, so we can say CaDA is "down to standard".

Other than that, the parts are 100% compatible with LEGO, but are themselves not always identical. Especially not the beams, which have a distinct asymmetric pattern next to their axle holes: I'm not sure whether it makes the parts easier to manufacture, or cheaper, or more homogenous in the long run. It does not affect building in the least, but if you would like to mix these parts with your LEGO, you may want to be aware of that.

Sometimes, however, CaDA's different parts actually work better: its 36-tooth gear is actually superior to the LEGO's version, for example. Also, its trans-turquise brick separator with a special tool to extract friction pins is quite nice and useful. It's a matter of taste, but I'd also say the custom chrome rims work well ― as long as you pay attention not to leave any greasy fingerprints, that is. And the metal U-joints, fully optional because the familiar plastic ones are included as well, are the stars of the show. 

The much-anticipated track day

So we've established that the design and engineering at work here is second to none, while the part and instruction quality is second only to LEGO. But how does this F488 fare as a remotely controlled toy car?


It's pretty agile despite its considerable size and weight, steers well, and the twin L-equivalent motors provide sufficient torque and power to move it around. The gear ratios don't differ as much as they do in real cars, so the actual chosen gear doesn't change the top speed that much, but you can still go for a thrill of starting in the 1st gear and then shifting upward as the car accelerates.

As the floor is rather low, you will have to make sure the driving surface is flat, and of course, any climbing is completely out of the question. If you're into that, better check out the current assortment of LEGO crawlers.

All the functions work very reliably, and the car itself is not prone to leaving a trail of lost parts behind itself. It's surprisingly sturdy ― in fact, once you learn where and how to hold it (that is, not by the roof), you will never worry about it disintegrating.

The wireless remote control controlling four independent channels on the lithium battery (drive, steering, gearbox and lights) in the car does its job fine and is not difficult to use. It is just an on-off control, without any progression in between, but you won't really miss that while driving. 

To disassemble or not to disassemble

It's a hard choice, and I guess it will boil down to each individual, whether they consider themself primarily a collector or a builder. The set looks great on the shelf and being driven on the floor, but it is also a fantastic and abundant source of building material which would be very welcome to any Technic (or technicky) builder's assortment of raw materials. If your pockets are painfully deep, just go ahead and buy two: have one on display and the other as a source of parts. 


Certain parts, like roof panels, are otherwise well known in the official LEGO Technic world, but are available here in a color (typically red) which is not widely offered by LEGO. So, especially if you're into building vehicles and don't mind them being dressed in a red-black palette, consider purchasing this.

While still on that topic, personally I find the white-black-white stripe along the centerline of the car incredibly ugly, stupid, infantile and wholly unnecessary. I don't blame CaDA nor Bruno Jenson as they just modelled the stripe from the original car the best they could, but Ferrari who painted it there in the first place. My hands simply resisted building something that ghastly, so I converted the entire scheme to fully red by replacing two dozen or so key parts with their red LEGO equivalents, and some help from the side. And am so much happier with its full-red outfit.

But that's the point I'm trying to make: perhaps including also the parts for a fully red F488 would have been nice ― it wouldn't mean adding that many extra parts, yet the builders would have an option to avoid the awful stripe if they wanted. 

The alternate future of power

Having covered all the necessary formalities about this car, it's time to consider some less obvious points as well as this set's role in the big picture.


As you are probably aware, Power Functions have been LEGO's main line of electric Technic components for quite a while, and recently they began being progressively replaced by the TC+ which introduces some completely new concepts, but doesn't keep all of the previous ones.

Technic builders' community has generally been pretty much satisfied with the Power Functions, but with a small permanent complaint about wireless remote control never making it to the portfolio (which SBrick took the advantage of, but that's another story). Here, in this set, we've got a Power Functions clone, fully interchangeable and compatible, featuring its wireless radio controller indeed.

To my mind, this shows the alternate future of what Power Functions could and should have evolved towards: a USB-chargeable lithium battery with a built-in radio receiver featuring four channels, and its corresponding controller. And to that extent I'm excited to see whether CaDA will keep on developing its own vision of PF-compatible devices, perhaps introducing other components the community asked for (like small motors, passive potentiometers, etc.), from the point where LEGO stopped and froze it. And yes, if you were wondering, it is possible to order just these electric components in a separate pack.

Should they decide to pursue that path, it may rejuvenate the Power Functions, even though not by its original creators ― and thus establish one interesting foothold for CaDA in the technicky market. Let's not forget that many devoted Technic builders still have buckets of Power Functions components lying around, which they will certainly not want to abandon yet. And with various new problems introduced by the TC+, that abandonment may actually not come for a long time.

Therefore, this here aren't just Power Functions clones, but in a certain sense, possibly a lifeline to its continued and evolving future. Let's see how the things will work out. 

Face to face with the rival(s)

The harder questions to answer are ― how does this set compare to its LEGO rivals, and what does its particular sets of pros and cons mean for both sides? Let's try to dissolve the argument to several areas.


First of all, it has to be reiterated clearly that, in several respects, CaDA C61042W beats LEGO at their own game. Despite all the undoubtedly impressive features of LEGO's recent 1:8 supercars, this F488 outplays them in functionality, design, engineering and overall sophisty. Although some come close, they are not quite at that mad level. And even if they were, they would probably not feature four motors nor a lithium battery pack with radio remote control.

But just as importantly, the underlying cause is not CaDA's personnel doing a better job than LEGO. I think noone in their right mind would ever think LEGO, with all their resources, is incapable of designing such a car as this F488. And nothing would stop them from consulting Bruno Jenson or some other renowned builder anyway, had they decided to do so.

Rather, LEGO would be reluctant to launch such a car due to several limitations they simply could not avoid from their position. They would have to license it with Ferrari, because being ridiculed for launching such an obvious copy would be damaging. Then, it would have to be simplified because LEGO would need to adhere to its policy that a large majority of the target audience should be able to build it without frustration. And with all the electric components, at LEGO's rates, it would cost the amount noone would ever be ready to pay for, except for a miniscule share of fanatics LEGO cannot make a living on. It could perhaps serve as a promotional pseudo-set, but that's a dangerously slippery slope.

So, not constrained by these inherent limitations, it is only logical CaDA had "free hands" to develop a car that could turn out to outmatch LEGO's contenders. Why didn't it happen before? Mostly because the quality of LEGO clones' parts was abysmal, and they were too concerned creating $3 garbage sets rather than having a proper attempt at something much more serious.

Then these folks from CaDA came along, improved the part quality to a level not quite as high as LEGO's but good enough, and hired one of the best guns in town to come up with a stellar model. Which resulted in a great, and surprisingly affordable, set we're dealing with here. At the time of writing this, it still costs 200 € (shipping to Europe included) on Amazon

David and Goliath, regardless

However hard LEGO may fight against these clones, let's be frank: even if CaDA quadruples the sales of this set overnight, LEGO would not even feel it ― such is the difference in their revenues, distribution and brand power. But the point is that the clones have now shown that they do have what it takes to play a competitive game, should they put adequate efforts in it indeed.


It doesn't alleviate the fact that clones are exactly what their name says ― clones, who reap some of their success from the work done previously at LEGO, and from the fact they can get away with an "Italian Supercar" set which is quite clearly an unlicensed Ferrari. In my view, if they want to improve their standing in that department, they should not just copy, but improve the parts: they already did so with the 36-toothed gear, and have thus shown that they can do it. That would, in the fullness of time, establish them as a more of a standalone parts company trying to complement LEGO, not just outright copying it. But that's a tall order, no doubt.

But come to think of it, why would we consider it so improbable? There is a striking analogy in the automotive world: Jeremy Clarkson once remarked wisely (for a change) that China has ascended from producing bizarre quasi-tuktuks to nearly perfectly conventional cars in just a matter of a few years. If they continued at such a pace, they would be a major player in the top league very soon. Perhaps we may witness something similar in our LEGO and LEGO clone world: the Danish giant will certainly not be dethroned soon, but it may get some quite decent competition from those dismissed as underdogs not so long ago.


And in the overall picture, that's a good thing for all the involved parties. The companies that will step up from the cloning business into good standalone set and part design should get rewarded ― there is nothing saying LEGO has a prerogative monopoly on any plastic parts sticking together. That would benefit the audience as well, with more choice: regardless of the market leaders always whining in such situations about the new players hitting below the belt and dismantling the market, in the long run it always worked out fine.

Finally, it could actually benefit LEGO to get some respectable competition. Lots of its competition so far has consisted of copycats blatantly stealing anything they could, violating all possible copyrights and intellectual property laws in existence ― and these should still be fought against. But the aforementioned serious, proper competition implementing its own ideas and skills would be a different thing entirely. LEGO themselves have become a bit overconfident with their decisions recently, and having someone directly capitalize on the wrong ones is a useful tool to keep oneself sober. As is the case in nearly all mature, established markets ― but is still an exception in this one.

It's a long shot, but maybe this is one of the first tiles on such a path. Time will tell. And in the meantime, grab those 200 Euros you've been saving for that fancy polar jacket you will ― let's be honest ― never wear anyway, and get yourself one of these instead.

+ Top-notch design
+ Insane mechanical engineering
+ Full PF-compatible wireless lithium battery, motors and remote
+ Alternate metal joints
+ Cool packaging
+ Excellent source of building material
+ Price

• Fine tolerances of parts still slightly below LEGO's standard
• Difficult to build, requires plenty of mental focus

- Occasional errors and unclarities in the instructions
- Some inconsistencies among part colors (primarily red)

Part count: 3187
Dimensions: 54 x 25 x 13 cm

Gallery











































Sunday, May 10, 2020

WiPy and how to take advantage of it


For a some weeks I had been looking for something that could in theory replace the Mindstorms EV3/NXT smart bricks, while adhering to some extra criteria:
  • more "directly" programmable than with blocks, preferably using Python
  • low power demand
  • many output ports (at least a dozen)
  • not too expensive
  • connects to WiFi
  • does not strictly need to read values - control is enough
Among several candidates I chose Pycom WiPy, ordered it together with a breadboard, and gave it a try. It runs MicroPython, a subset of Python designed for such SoC devices, and featuring some extra stuff controlling the pins, sensors etc. conveniently. But if you're familiar with Python, you will certainly not need any extra effort to get along with MicroPython.

To sum up: yes, it is possible to conveniently control LEGO motors using its output pins. WiPy's outputs are actually just logic pins (0 or 5V), and not power outputs, meaning there has to be some kind of power control connected to these pins. I took a relay module because I had one handy and wanted to control high-powered stuff too, though it is clear that transistors could be used as well, bringing better response time, reliability and longevity too.

So I simply let one of the relays pass the 5V supply, the same one powering the WiPy, towards a PF motor. As already known in the experimenting circles, the motors work just fine powered by 5V, but at slightly lower speed. Still, they are perfectly usable as long as you don't need plenty of power.

One of my first experiments was building a LEGO device that would physically turn my laptop on by pressing its power button. It works fine, and on the other end, it is actuated when the WiPy device receives a network command for it (which, in turn, comes from the Internet, via a bit of port forwarding and Synology dynamic DNS service). So WiPy simply allows controlling its pins via network, and in this case lets me turn my laptop on while away, and use other means to see the stream from its webcam.

Reading values is impossible with PF and an extra interface could probably be built to handle NXT/EV3 motors, but that is something I will probably do sometime later. For now, I wanted to tell about the "first success" in using WiPy, and confirm that it can indeed be used for the original purpose. It has got 24 outputs (some of which can be used as inputs if that is what you need), so a lot can be automated with it.

Tuesday, April 9, 2019

Some thoughts on Trial Truck construction & design

Having constructed a small parking lot (2-digit number) of Trial Trucks so far, and compared their live performance with their features, there are definitely some lessons to be learned―which I shall briefly cover here, hoping to help other builders possibly struggling with similar problems. And especially to help beginners avoid some tempting ideas which turn out to be cul-de-sacs.

If I had to condense all my Truck Trial experience into one single, most important lesson, it would be to avoid complexity. Features are great for bragging on the forums and in front of the less tech-savvy friends, but in practice they rarely actually help the truck's performance. Among the heaviest mistakes I have done in my early days was to cram tons of advanced mechanisms into a truck, believing they would help just they way they do in the real-life off road cars. But quite simply, they almost never did.

For example, one of my first entries (and worst failures) was an off road SUV which had no less than three independently lockable differentials, independently adjustable ride height at the front and the back axle, five-speed gearbox, four-wheel steering and, well, working headlights―and all these controlled remotely. While it was mechanically sound, in the competition it got easily beaten by the trucks half its power and quarter its features.

Not only was it too heavy, but also all these features unnecessarily raised the truck's center of mass, making it less stable. None of the advanced features actually turned out as advantages―all these differentials and ride heights only improved the performance very slightly, and altogether it was still miles behind competition

So―I can't emphasize this enough―just avoid being seduced by the fancy mechanical contraptions and go simple. A simple but responsive steering, fixed four wheel drive with plenty of power, long-traveling suspension and good ground clearance is all you will ever need. In fact, sometime later I have built a Trial Truck I got almost ashamed of due to its straightforward simplicity―yet it turned out to trample the earlier one in all the criteria important in competition.

Simplicity leads to structural strength which is another point to consider. Simple, multiply reinforced beams with several trusses in critical places will keep your truck strong and resistant to hits and vibrations, will make it easy to repair even if something goes awry, and keeps all the gears and moving parts firmly in place. This is something TLG has done finely with their official Crawler which, despite all its mechanical intricacies, is essentially simply and strongly built.

Don't be afraid to reinforce―indeed, you cannot get away without it―but simplicity makes the truck lighter and stronger, therefore requiring less reinforcements in the first place.

But what about those cases where the rules mandate the minimum truck mass? Sometimes as much as 2 kg? Does it warrant including some more advanced features if there is anyway ample capacity to build them?

Well, it seems that the general answer is no. Again, these features are hardly ever necessary, whereas they introduce issues with strength and reliability that always seem to surface exactly during trials. In such cases with minimum mass requirements, at least for me, it turned much better to devote the extra weight to very strong and reliable suspension and steering, robust transmission, and designing the chassis so that all the heaviest components are as low as possible, while retaining sufficient ground clearance. Of course, in a typical Trial Truck, it's the battery packs and the motors that have a large share in the overall mass.

Ideally, even according to the standard car construction books, the weight distribution should be approximately equal on all wheels for maximum performance. In Trial Trucks, as well as in the real world cars, it is not always possible, or at least it would require too many concessions in other areas. In such cases I've found it helpful to move the center of mass a bit forward rather than back, while it should obviously never move aside. The reason is that the forward center of mass allows for better distribution of weight over all wheels while climbing, which is the most critical operation. It makes descent slightly more unstable, but usually that drawback is largely overshadowed by better climbing performance. Of course, for this to work as intended, the truck should have all-wheel drive, but this is something one should consider as granted for a Trial Truck anyway.

And if there is some mass to be added even after all the necessary components have been completed, it helps to distribute them evenly around the chassis (or again, slightly forward).

Onward, then, to some construction details. Try always to have as few gears in the entire transmission chain as possible. If you opt for a gearbox, make it as straightforward and robust as possible, with two gears usually being enough. And try to rely on the newer, bevel-style gears as they seem to be stronger than the older, spur types. If you absolutely have to go with spurs, the new 24T and 12T seem to be all right, but avoid the small 8T, even in their newer versions. And make sure all torque-bearing gears are properly braced from both sides and mesh cleanly.

Mentioning torque: it is generally wiser to trade a bit torque for speed. What I mean to say is that, instead of letting most of the axles turn slowly but carry large forces, it makes sense to use a couple of gears to make the axles turn faster yet transmit less force (i.e. torque). Of course, one should not go into extremes as friction comes into play at very high rotation speeds. But a transmission which mostly relies on higher speed/lower torque configuration, only to reduce the speed for more torque at the final stage at the wheels (preferably portal axles!) is often a good idea. Less torque also means less reinforcement, and more reliability.

Tuesday, February 19, 2019

LEGO Technic Seismometer Prototype



Well, in its barest concept, a seismometer (device for measuring Earth's surface movement, namely quakes) is a rather simple thing. A large weight, suspended to move freely, attached to something that can note its movements - and that's it. And while assembling a rough one from parts scavenged from an old rusty car at the dump may seem just as simple, building one from LEGO Technic brings its own set of challenges along.

Although a group of dedicated weight bricks (such as 73090a) would have been a "purer" solution and possibly serve its purpose better, I went for the more pragmatic approach and connected a set of several battery packs together - thus indirectly using batteries as weights. Altogether, something the size of a large coffee mug ended up weighing well over one kilogram - more than enough to have the concept proven.

This weight is suspended, hanging on two rubber bands. I agree that, in theory, going for a more complex setup featuring six or eight bands (one for each corner) would have helped with force distribution and rotation, but would have been a nightmare to tune and set up. Two rubber bands allow for sufficient freedom of motion, however, and have a better chance of retaining the seismographer's sanity.

Furthermore, the weight is attached to two independent pieces of thread, one in vertical and another in horizonal plane, both forming a closed loop, which is led through a system of pulleys to translate the weight's movements into proportional movements of the threads. Finally, attaching pencils to the threads and letting a writeable surface slide perpendicularly underneath the pencils, completes the essential seismometer.

Hence, each pencil notes the weight movement in its own direction over time - one vertical, and other horizontal. This is a common approach with real life seismometers as well, because horizontal and vertical movements of the Earth surface lead to different effects and are measured separately. True, this weight could have included a third thread loop measuring the weight's movements in the third direction (sort of like X-Y movement) but that would add lots of complexity while bringing only marginal functionality, and likely go beyond the 48x48 baseplate I decided to use as a caliber for this project.

The writing surface, which is in this case a cascade of white Technic panels, is driven across the direction of the pencils via a rack and pinion system with variable speeds, controlled with a gearbox, allowing for 10 possible settings, balancing between precision in timing and total duration of a single "plate". With only minor adjustments, one could convert it so that a standard roll of paper (e.g. like those used in shop counter printers) can be used.

Finally, apart from dozens of accurately perpendicularly tuned pulleys (perpendicular pulleys retain linearity better), both threads pass through a simple, manually controlled tension mechanism employing a large 40T gear and a worm gear, allowing the seismographer to adjust the correct tension level finely once the seismometer has been set up. The aim is to keep it tense enough so that all weight movements are precisely translated to the movements of pencils running across the "writable" white panels, yet loose enough to avoid friction that would overly hamper the movement of the weights. With some experimentation, I've settled for the tension of about 1.5 Newton, roughly the tension you would get if you suspended your mobile phone off a thread.

The only active component in the entire mechanism is the motor driving the writable panels, mounted on springs to further isolate it along with its vibrations from the rest of the mechanism (though I admit it's not as bad with fixed mounting either). But this could even be done manually if one wanted - the motor at least assures that the graph movement speed is at least somewhat constant. The range of possible speeds, of course, increases if one uses regulated voltage at the input, like I did with the classic 9V train controller.

So, an obvious next question is - how does it behave? The good news is that it proves the concept indeed: having been set upon the table, it is sensitive enough to measure, at least with some consistency, average table bumps, and would surely have no problem measuring a mild earthquake. But it is, on the other hand, far too insensitive to actually measure micro-movements of the Earth surface which are typically imperceivable for people, and which real life seismometers are built to measure. Let alone even finer things like steps of people in the building, opening and closing faucets, bass kicks from someone listening to music somewhere, etc.

I don't say that the latter group could not theoretically be measured with LEGO, but that would require a different design, utilizing more specialized parts, and would likely be far too sensitive for actual quakes. It would probably revolve around a multi-reflection design, featuring a mirror on the weight, with a laser pointer pointed at it, and furthermore with the ray being reflected multiple times to artificially create distance and thus increase resolution. Finally, the laser spot's movements would be monitored by a camera, or by a cascade of Mindstorms' light detectors, or something similar. Such a setup's sensitivity can in theory be increased virtually infinitely, but at the cost of maximum measurable extents' window reducing. Id est, it may well measure the vibrations of the roof being shaken by the wind five stories above, and someone sorting out cutlery in the kitchen in the other wing of the house, yet go completely off scale if breathed into.

Finally, a bit of historical trivia: although commonly thought to be of modern origin, the first seismometers actually got built as early as 2nd century AD, in China. It was a rather simpler approach, with a precariously ballanced ball at the top of an inverted bowl. The fact that the ball dropped from the bowl indicated that there was an earthquake, while the direction of the ball's fall gave a rough idea about where did the earthquake originate from - letting the owners know in which direction should the help and rescue teams be sent.

Built for BrickStory 2019.

Tuesday, August 8, 2017

LEGO Aquatics lecture

A lecture I gave at Paredes de Coura Fan Weekend 2016 in Portugal. Enjoy!
I've always been attracted to various aquatic sets, but the game gets serious when we start dealing with real water.


Tuesday, January 3, 2017

Tolerances, accuracies and their sensible limits

Now with the Engraver v4 completed, as shown in the previous post, it's no secret that the large amount of its development had actually focused on improving its accuracy which, in turn, meant reducing tolerances wherever possible. So perhaps some of the experience and the lessons learned throughout may be of help for builders who are attempting build similar constructions. Keep in mind that these are merely my notes and thoughts, rather than something I believe should be set in stone.

First of all, it is important to be able to judge where should the tolerances be reduced for the maximum effect. In most situations, I've found that only about a dozen parts or so are responsible for well over three quarters of all inaccuracies in the system, and several carefully observed test runs should easily pinpoint them. There are a few standard culprits:
  • backlash between the gears
  • unwanted flexibility of the beams and axles (or similar supporting parts)
  • slack of the frictionless pins
  • friction (and therefore hystereses) all around
  • axles which do not fit into Technic pin holes snugly
  • overcomplexity of kinematics or control bars
All of these can be dealt with one way or another. Overflexible supporting structures can usually be easily reinforced, or the studless beams replaced with the studded. Axles can be replaced with the pins, and frictionless pins with the friction pins. Backlash cannot be fixed, but it can at least be controlled by keeping the entire system under a mild tension or resetting (recalibrating system with a run-up) ahead of each change of direction. The same is applicable for hystereses as well.

Combining all these techniques, in this case on the Engraver but basically applicable anywhere, I've managed to reach the final resolution down to about 60 µm, based on the linear actuators. The reasoning is simple: the Mindstorms motors are directly connected to the large linear actuators, and 15º is about the smallest angle they can be turned reliably. With the actuators' ratio of 240º/mm, it is clear that the smallest reliable linear movement amounts down to 0.0625 mm, i.e. a sixteenth of a millimeter.

The engraving head itself was levered at about 1:3, so its theoretical accuracy would be closer to 20 µm, were it not for the hystereses that bring it to about 100 µm, and which is perfectly enough for its purpose anyway.

The obvious question related to the point I'm trying to make: can the precision be improved further using downgearing to an arbitrary level, at the expense of speed? Theoretically yes, obviously. But I fear practicality gets in the way. The amount of reinforcements, pullbacks and tensioning mechanisms already required to take advantage of the 60 µm resolution is staggering, and pushing it further asks for even more such mechanisms which are heavy by design, in turn requiring even more compensations. Altogether I wouldn't exclude it as impossible if lots of resources and effort is invested, but for practical matters, I don't think going past, say, 10 µm makes any sense.

Even 60 µm sounds good (after all, many standard papers are thicker than that) as long as you keep in mind that we're talking about the resolution in the relative sense. Id est, we can know that our engraver, needle or something similar has moved specific 60 µm from its previous position. But the absolute accuracy is a different ball game entirely. At the best of times, having its resolution at 200 µm is very good, under controlled and stable conditions.

We are basically asking for plenty of accuracy from the parts that were not designed for it in the first place. Sort of like making a miniature engraving on a single bean using an old kitchen knife: possible, no doubt, but requiring a huge amount of "special techniques".

Thursday, December 22, 2016

Engraver Mark 4

P1180074

Yes, it has already gone as far as the fourth iteration; I'm still convinced that the "proper" 3D CNC cannot be done using this design for a variety of reasons I've touched on in the earlier posts, but at least the engraving part seems to be more or less satisfactory now. I've experimented with about a dozen possible improvements, among which some made it to the final version (that is, final for this fourth one at least).

P1170650The one that made the largest difference was possibly the conversion from pure "vertical" rack mechanism that raises and lowers the drill into one based on parallelogram linkage. This means that the vertical movement of the drilling head is circular rather than perfectly vertical and linear, but if adjusted properly, it does not cause any problems, while it simplifies the mechanism a lot and also makes it more precise and sturdy.

The first design's parallelogram was controlled by a linear actuator fixed to a point on the drilling head itself. Although very practical and elegant, this had an unwanted side effect: the entire construction was strained which led to bending and therefore, inaccurate drilling points. Therefore the linear actuator was moved onto the "sledge", reducing the effect to a negligible level. This also offset the center of gravity of the sledge, but this turned out not to cause any problems, it seems even to have actually made the entire structure more stable.

P1170723Drill bit is now held by a structure that forces two rubber 2L connectors one next to the other, as the drill itself is squeezed right in between. Though I've got to admit being somewhat skeptical about this solution at first, it actually turned out to keep the drill bit nicely and reliably centered. And has the advantage of allowing nearly any sort of drill bit being used, at least as long as they are not overly thick, that is, over a few millimeters.

In order to conserve power and the classic 9V motor used to rotate the drill bit (without any gearing), it is run only when needed, i.e. during drilling, and turned off while moving the drilling head about. This is done indirectly, using a fourth EV3 motor which flicks the Power Functions switch directly through an axle. This also, after a few adjustments, turned out very reliable, and was dismantled only when I accidentally manually rotated the motor too much.

P1180078The results have altogether been pretty good - the engraved images are noticeably clearer and more precise than its predecessors. The critical parameter is the force by which the drill point pushes against the engraved surface, and this is something that needs to be accurately set manually. I suppose this could in theory be done by using contact sensors, but it would - because of very little distances involved - ask for a complete overhaul of the design. That might perhaps turn out to be the mission objective for the future Mark 5, should it ever turn out to see the light of day.

Finally, instead of controlling everything from the computer, now the engraver itself offers a green and a red button, the red one to terminate the process at any time, and the green one to firstly confirm the engraving area (the drill point makes a run around the to-be-drilled area to ensure there are no obstructions), and later to pause the drilling process.

P1180105The maximum engraved image size is still limited by the extents of the linear actuators, which turns out to about 5-6 centimeters in practice. This is more than enough for cutlery or bricks, of course. Shifting the approach over to the new linear racks could increase it to at least 8-9 centimeters square, but would require more compensating movements because of the backlash between racks and pinions and therefore perhaps reduce the accuracy a bit. I've written already elsewhere, and I still think, that a longer linear actuator or actually a linear rack as envisioned by MinuteBot would help a lot - if only it was viable for its makers.

Another thing this design has improved upon is reproducibility. Since it relies mostly on common parts and is a very barebone approach, it is rather easy to rebuild it on demand quickly and painlessly. Anyway, nearly all X-Y-Z control mechanisms tend to converge to similar designs after all, given the same premises. As a pure challenge I've thought of building it as an alpha-beta-Z mechanism, i.e. involving two rotors rather than X-Y axes, but I don't see any particular advantage over the existing design, except for perhaps looking better.