Hrafn

I like that you're using commonly available parts, but ultimately I think the axles are going to slip out of the connectors. To withstand tension, the assembly really needs liftarms, which can't pull apart.

Except for this one which came in blue, lime, and red as well.

Those are the hinge parts I was thinking of. They're not common in Technic, no, but they're cheap on BrickLink. As long as you're working in black or gray, they're useful. The ball joints I was thinking of are the Mixels ones, as mentioned by others; those are only 1 stud wide.

Yes, I don't know that you'll be able to make it any better with Technic parts - A-pillars are tough. You might be able to use some of the Mixels / Chima Legend Beast ball joint sockets, or maybe System hinges (especially click hinges that connect to Technic axles), but yours is probably about as good as is possible.

True, it does stress the pins, though my experience was that it wasn't too bad. Alternately you could keep the 3 outer pulley wheels, swap 7L liftarms for the 9L ones, and use 2x in the middle. And you could keep one of the spokes at 9L for additional rigidity. EDIT: yes, there's still some stress in the parts because the geometry isn't quite perfect - the spokes should be 0.08 studs longer than they are for perfect geometry. In practice it works reasonably well.

Glad to help! What are you building that has such a large triangular structure?

Interesting. I paid for an order today as well and was given the same time estimate for parts shipping from Denmark - but I'm in the US, and would have expected it to take longer to ship here than to other parts of Europe. Maybe the 10-12 days is a conservative estimate, though - an "under-promise and over-deliver" sort of thing to keep customers happy.

Would this work? I'd add at least 3 more friction pins (connecting the short liftarms to the pulley wheels / wedge belt wheels) but it's surprisingly strong in-plane as is.

Your photos aren't coming through, but yes, blue pins are friction pins and tan ones are non-friction ones. It makes a big difference.

Hmm. I'm in the US, can't buy sets on shop.lego.com, but (per this thread) I've started the process of buying some parts through Lego. I was quoted the following prices (exclusive of tax and shipping): 6055519 (dogbone) 0.49 6055628 (5L A-arm) 0.63 6055629 (snowmobile kingpin) 0.28 4268659 (CV male) 0.97 4610374 (CV female) 0.53 4525184 (3L differential) 1.00 42000's hubs weren't too bad (0.94 and 0.80 front/back) but the tires were absurd (4.52 each!) Part 4610378 was 1.64 each, as well. So some parts are cheaper than expected (the dogbone), some are reasonable (the differential), and some are much higher than I'd expected.

What if instead of the T-liftarms, you used wedge belt wheels to connect the outer three liftarms? That might move the inner spokes towards the center just enough that they too could be joined by wedge belt wheels or other parts.

Beautiful! As for the ball joints, my experience (outlined here) is that the ball joints will not pop out very easily and certainly will not have friction. While they have friction in the sockets that are designed for them, they don't have friction when sandwiched between two pin holes 1L apart. That said, I would change the outer axle hole to a pin hole and get rid of the countersink - it's not needed for smooth and secure ball joint operation, and it would weaken the hub and make the ball joints slightly less secure than if they were simply sandwiched between pin holes.

How do the new hubs compare to the old () hubs, when used with the new 5L snowmobile arms? Are they more wobbly than the old hubs in this setup, or are they about the same? Thanks!

Beautiful! The red stripe on the sides is a nice touch. Do you have any higher-resolution photos?

Good point, though I've found that if you increase the axles to 5L and add a 1x3 thin liftarm on the top and bottom, the ball joints will pop out of their sockets under pressure way before the lime connectors start to slide on the axles. This was just a spur of the moment sketch to convey the basic idea in case someone found it useful; since I'm not currently building anything that uses wheels of this size I haven't tried to develop it further. Certainly it needs more work if it is to be practical. EDIT: A more robust setup with the CV joint is possible (though the steering connection is a bit wobbly), but the wheel still doesn't want to stay on with only 1L of CV axle connection: If we switch to a u-joint, this next hub gives a 1.5 stud offset from the edge of the wheel, but it's a bit less robust and I'm not sure how to attach the steering link: The u-joint design is theoretically subject to the problem of the lime connectors sliding off the axle, but again in practice the ball joints pop out first - and it takes quite a bit of force to pop them out. I wouldn't use it for a truck but it might serve for a car.

The closest I've ever gotten to the rim was 2.5 studs using 62.4x20 wheels; with larger hubs it might be possible to get closer. Note that this construction uses 10.2mm ball joint parts (not the smaller tow balls used in dedicated suspension parts) which work but are more fragile. The suspension has a number of issues (friction, awkward placement of drive axles, no differential, very limited steering lock, etc.) but at least it's narrow. EDIT: yes, it can be used with the 44772 (Technic Racing Medium) wheel to get a pivot point 1 stud away from the edge of the wheel. However, the CV axle then only protrudes 1L into the wheel, which isn't enough to keep the wheel on reliably. Adding 2 wedge belt wheels and some pins would give you a pivot point 2L away from the edge, but make things sturdier.

If you attach the wheel to both the axle (going through the clutch gear) AND the turntable, what's the point of the clutch? In the setup you showed, I don't think you really have a LSD - if the wheels get stuck, the axles going through the clutch gears will just spin fruitlessly, with the torque going to each wheel limited to the slip torque of the clutch gear. That won't happen with Thirdwigg's design - most of the motor's torque will go to at least one wheel in his design no matter what happens, minus the slip torque of the clutch gear which will be transferred to the stuck wheel. If both wheels are stuck, they should both get half the motor's torque.

But then he'd no longer have a portal axle, if I'm understanding your idea correctly.

Between 42000 and 42025, white panels should be getting more common (and hopefully cheaper on BL). To those two sets, add a non-technic one - 70810 "Sea Cow" from the Lego Movie: http://jaysbrickblog...-ahead-to-2014/ I remember seeing someone post a picture of another set with 8 (?) male CV joints in red, but can't find the post anymore. What other sets in non-Technic theme have useful parts?

Looks like 44135. Along with 44136 it's one of the few Bionicle pieces that's reasonably useful for Technic.

Wow! You should open a new business: Zblj's Lego Towing Service.

It helps me understand what you did and why, yes, thank you. I had misunderstood what you meant by a 'floating' diff.

Ah, now I see what you did, that makes sense. The diff there isn't really floating, it has support on each side. I was trying to have a diff supported only by the u-joints on each side, which didn't work.

When I tried a floating 3L diff (leading to a u-joint and then a CV, and using 6L wishbones pivoting so they act like 5L ones) I had two problems. First, the diff would move away from the driving gear when the wheels were stuck. Second, that situation led to the diff teeth getting mashed. How did you avoid those issues, or are they not a problem with the 4L diff?

+1 for parts for narrow driven suspensions. The hangup here, besides there not being steering links shorter than 5L, is the length of the CV and u-joints. I'd like to see a 3L or 4L CV part with two 'male' ends. Then you could use one of those and two of per axle, have independent suspension, steering, AND fit in a differential in a narrow vehicle. Right now if you want to do an AWD car with diffs, the width of the components pushes you to 1:10 or larger.