gyenesvi

Indeed great looking and functioning alternate model, could really be an official one. I like the part used for the side mirror and the build of the basket (including the hose), and the way it lowers right to the ground as the first stage of the boom elevates :)

It's great to see some nice old stuff from the time when I was sleeping :) I like that you made some proper live axles back then with built portals, I was always curious of a good solution for that part, and now I could check it in this video. Even the weight distribution of that chassis looks good with the battery and motors in the middle. Cool stuff!

I'm afraid such things don't exist, otherwise they would be pretty popular :) But let me know if I am wrong, I'd be also interested. I think radio controlled components (ESC, receiver, battery) only exist in forms that are not quite possible to build into your lego model, people rather just place them in the trunk/cockpit somewhere.. The best I've seen so far are the Geekservo motors, they have a lego compatible housing at least.

What does this exactly do? The wheels won't turn this way, no?

That would be awesome. Maybe it could be a more maintainable solution for TLG than PaB, because they would be pre-packed and not needed to part things out separately (and probably manually) for each order..

Oh, okay, I get it now, thanks!

Oh sure, indeed, what I wanted to say is they cannot be built permanently locked, without using that locker of the case, in a way that fits into the 5x7 frame. With the old grey diff that could be done by replacing the gears with a locking bridge, and in case of the red gear on the heavy-duty diff, the whole can be replaced with a 28T double bevel gear to get a permanently locked axle with the same gearing ratio. Neither of these tricks can be done with the new yellow diff I guess.

I think the heavy-duty diffs with the read 28T gear may have bigger friction that the Ferrari's because the read gear and the 12T spur gear are double bevel and are both rubbing against the 5x7 frame (how it's typically built), while the yellow gears are not double bevel, so less friction is expected at the frame. Also, the teeth are more like a crown gear which has less friction I guess. Similarly, the old diff is not double bevel, so it's not rubbing against the frame. I'm sure it is, it's probably designed for that use case. The only thing is that the gearing is slightly slower than that of the old grey diff (14:22 = 0.63 vs 20:28 = 0.71), which may or may not be good depending on the use case. However, the above slight difference gave me a really interesting idea to try once I get some of those diffs. In rock crawlers, the front wheels often has about 10% overdrive compared to the rear ones to pull the car onto the rock more than to push it (and flip it backwards). With using the old grey diff at the front and the new yellow diff at the back, it would be possible to achieve that ~10% speed difference in an elegant way (otherwise it would be really complicated and would not worth it). The only downside is that the yellow diff cannot be locked..

Great model, I already liked it very much in the BuWizz camp! Especially love the V-shaped engine build. Apart from the space advantages you mention, it has another important benefit: the output axles get close to each other, 2 studs away from the central axle, which allows the use of medium sized gears, and so opens the possibility for the gearbox (which could not be done if the motors were just vertical). I also like how the double gearbox distributes the load, and how simple the whole drivetrain is, despite it has the gearboxes. Great idea with the fast output driving the fake engine as well! I love that it can run in neutral when the car is not moving! Also, it's nice that the interior is structurally so simple and clean, yet the whole model is rigid. And nice clean overall shape! Finally, this answers some of my questions in the other thread about performance; using the slow output to drive the simple hubs is also an effective solution that seems to work, and you even addressed the front wheel drive case! I can see that you are using the new CV joints there with an offset towards the center, but not sure how that works exactly. I am guessing that as the wheel steers, the pivot point of the hub and the CV joint are not aligned right? So that way the joint pivot moves a bit forward/backwards and pull the drive axle with itself, but the reason why that does not matter too much is because the inner brick-built joint compensates for that? Do I understand that right? Furthermore, is it that movement that allows the axles to slowly move out of the differential? Did I get that right? Nevertheless, it's an ingenious solution! Let's hope that the yellow differentials were actually designed for some future RC car with a plain hub that uses the new CV joints! :)

I agree with you on these points about portal hubs, for me the most annoying is the terrible steering geometry, not only because they need a stronger mechanism, but also because they require a lot of space. Furthermore, you can't put a locked differential into a portal axle, because of the large scrub radius the wheels are actually travelling a noticeable distance in the opposite direction, and that would twist the drive axle. In my model steered with an M motor and linear actuator, the motor was not strong enough to steer it in place when locked (even without weight on it), because of this twist effect (in this case the drive axle was strong enough to stop the steering). Anyway, when I said I'd like to compare non-planetary hubs with the above mentioned gearing, I actually meant plain hubs, not even portals, that would have a similar gearing to your setup as @Daniel-99 has calculated. However, I have forgotten about steered front wheel drive, which would be only possible with old CV joints in case of a plain hub, which I think is not a great idea for high-power fast applications (apart from maybe popping apart, I think they would wear pretty fast, and also steering radius is bad). So the plain hubs should be reserved for RWD off-roaders only, nonetheless a trophy truck could still fit that design. @FriedlS, currently I am using BrickController2 app with a gamepad for Buwizz, that's my current best option. It already gives much better controllability than the touch-screen.

I like the design process so far, dealing with all the problems I typically also encounter with driven and steered live axles. Curious how this double drivetrain turns out, how you place the diff in the chassis. I guess it will be easier in a longer truck than in a small Jeep. BTW, can you summarize what problems you found with suspensions using the rubber piece?

Nice little buggy, I like the overall shape and the compact engine build in the back. About the rear suspension, is that a proper live axle that can actually articulate (one side move up-down without the other), or is it a trailing arm that moves the rear wheels together? To me it seems the latter.

Interesting topic! I have been thinking about whether I should experiment with more than 2 motors after the race in the Buwizz camp, because my model seemed quite slow compared to some others using 4 motros (and yours of course), and figured it's exactly as you say, because of the number of motors to weight ratio. I will probably try it once I get my two extrs motors posted, though I am still a bit afraid of smoking away too many components.. But your approach of distributing motor torque seems like a safer way to go. Unfortunately, that does not really favour realistic drivetrains and suspensions, like live axles with a central driveshaft, or there will be a lot of torque directed to those central axles. Even with two motors, the new CV joints in my drivetrain are starting to deform a bit, and also my planetary hubs are running too smooth compared to when they were brand new. I have done the swap of the diffs to older ones to gain more speed, and now the drivetrain is like this: slow output of two motors coupled onto the central driveshaft, going to 20:28T diffs and then to the planetaries). This setup has quite good playable speed with a 1.4 kg model, but still plenty of torque, it can climb a 50 degree slope crazy fast (and it still does not cut out the Buwizz 3 if I avoid running it back and forth quickly). So for me, this is kind of a sweet spot I find both playable, reasonably priced and component friendly (I lubricated every moving component, including the suspension towballs, as even those had visible wear, i.e. plastic poweder). Anyway, about your model. As far as I can see, the high gear up-gears 3:1, right? With using the fast output, which is about 1.5x faster than the slow output, that's altogether 4.5x faster than the slow output itself which causes a lot of friction and probably loss of efficiency, and then the whole thing is slowed down at the planetary hub 5.4x, which about cancels that speed-up. So maybe, it would be interesting to see how this setup compares to instead using the slow output of the motor in combination with non-planetary hubs. Maybe it has a better efficiency and is less prone to damaging parts. Probably it would be good to lubricate the hubs themselves. Have you ever tried such a setup? I want to test that in a trophy build in the future (with 2 buwizz motors). Finally, I'd like to make a note about controllability. I think these speeds achievable with more than 2 motors quickly become uncontrollable and hence somewhat impractical, which is also an important playability factor, especially for offroaders. In the camp we have seen a nice build with 3 motors, it was very light, so good power to weight ratio, crazy speed, but it could not perform well in the race because it was soo uncontrollable (also partly due to the app-control). Also, I had the feeling that you were a bit careful with your model in the race, not making use of all the power it has, because you wanted to keep it controllable. That's also one thing to consider when designing a fast model.

I was lucky enough to see this in real-life, amazing technical design and brutal power. I like how it showcases the way of leveraging that amount of power without overloading the plastic parts. It's cool that each wheel has its independent drive with a gearbox, and that allows it to turn in place by spinning left/right row of wheels in the opposite direction. Also the double suspension is a nice trick, and the way the pendular front axle and bogie rear axles complement each other is clever. Great design overall! This encourages me to go beyond 2 buggy motors per model, to leverage more speed, because really as you say, it's a whole new level of performance that opens up new play possibilities :D

Nice truck and presentation, I like how clean and simple the whole frame is! Though the drivetrain and suspension is also somewhat simpler, and it helps keep things clean :) I also like the use of the Geekservo for steering, I'd like to try that myself. Is that strong enough to steer a model of this size on rough terrain as well? Does it also steer comfortably when the model is stationary? That Wixy controller also sounds interesting, as I'm looking for something simple that could be used to control that Geekservo. Is that little white connector thing and RC to PF plug converter? Can that be used to connect the Geekservo to a PF reveiver or a Buwizz for example? Or does that need some other control unit to actually work? It will be helpful if you present the whole system with Wixy controllers next week! Thanks!

Interesting! Nice that it can differentiate 3-4 classes by using only ~50 examples, though that's probably because the objects are really simple (practically no within-class diversity) and sufficiently different (large inter-class diversity), they can be recognized basically by a few colors. Will you try to scale it up? I wonder how that could be trained, for example to recognize cats and dogs. I expect that it would need much more images (thousands), because of large within-class diversity and more similarity between classes. Also, to avoid false positives, you'd need a lot of and diverse background images. Curious to see future developments!

I think the turntable design needs greasing to work well, that's how it should be compared, but I have never tried it (yet) so cannot really tell, just read it in its thread I guess. Sure the power may be boosted by the cada motors and buwizz 3, though the build does not really change by putting in more powerful ones, so that's not a real upgrade for me. Maybe the PU servo needs a bit of rebuilding because of the different form factor, but still probably not too substantial, and you haven't mentioned if that resulted in better steering or not (in your video you don't seem to be able to control it too well).

Wow, that's a pretty smooth walking robot! I always wanted to do computer vision with Lego (since I work with CV in my job). I guess the image processing is running on an Arduino, right? And that controls the geekservos as well.

Why did you need to redesign the wheel hubs? The original design based on the turntables seemed very sturdy, and you replaced it with something that relies on a single axle that does not hold by itself so you had to use superglue (this is how I understand it). How is that an upgrade? What was wrong with the original design?

Yeah, that might be true, but the red gear for the diff is also useful actually. With 2 buggy motors 1:1 to the diff, and planetary hubs, it just gives good speed and a lot of torque at the same time :) Also, a setup without planetaries but with slower diff gearing is also an interesting one. So I don't think it would just replace everything. Furthermore, the old 3-gear diff could be easily locked, while the new one cannot be. The 12:28T diff could be replaced with solid gears in the same ratio, but the 14:22 ratio cannot even be built with fixed gears, so you cannot build a car with a locked rear diff but open front for example.

The axles themselves contain a 20:12 upgearing before the diff, which I find nice and may alleviate the speed problem. However, the way the 20T gear is fixed with those corner connectors does not seem stable for a heavy duty application. It doesn't really seem form-locked (close enough) agains separation from the 12T gear upwards, I think when pushed really hard it could crack. Though not sure how hard this will be pushed.

I also speculated that the new yellow diff gears were actually introduced for something RC, maybe something that has planetary hubs but is faster than the Zetros was. Or simply something that does not have planetaries but needs a more solid diff. Unfortunately, I don't really believe that Lego will go for something really fast like a Trophy truck, I think that's kind of against their policies. But maybe they realized that the Zetros came out not very playable. I mean its speed is actually good for crawling on real rocks, but that's again something that stock Lego sets are not designed for, so that slow gearing is kind of useless in stock models. Anyways, I hold my fingers crossed for something (off-road) RC for next year.

@aFrInaTi0n, I am totally with you on this, I have just returned from the BuWizz camp, where I tried to discussed these issues with the team members in person (along with others having the same problems), so let's hope there will be steps forward soon. I agree that the highest priority is to make it failsafe so that it does not shut down for fast models (the biggest problem is that it's unpredictable when it shuts down, because it depends on many circumstances), but as I was told it must be done in firmware, whose development is outsourced, so it is particularly cumbersome. On a positive note, we were shown a demo version of a gearbox shifter module, so let's hope that that one gets incorporated into an update soon. BTW, the camp was a nice event, a great way to meet in person, compare our models, approaches to the same problem, and interesting lessons learned from racing a bit :) So I can encourage you all to participate next time! :)

This is a beautiful build @Timorzelorzworz, great choice of model, and very nicely implemented. Really nice curves, I especially like the rear section, and how everything is openable, and great that it has a gearbox that's different from a usual supercar builds! Congrats!

Okay, thanks for the info!