I'm working on a 4WD chassis that uses the new power-up motors and battery unit, the new CV-joints from 41099, the new differential from 42109 and the new rims from 42110. I found a way to use the new CV-joints with the old wheel hubs, at the cost of one stud offset. The setup also incorporates a substantial caster angle. I will be using the new defender rims form the 42110 Defender to compensate for the bad offset. I intend to use this chassis for a RC rally car. The model will have 2 stud suspension travel all around.

I will use this topic to show progress.

Edited May 16, 20204 yr by Didumos69

Looking good as always

Did you test the front suspension setup IRL? I have used similar setups in my builds, but because of the caster the drive axle might sway when steering. It could end up squeezing against those vertical half beams...

That’s a really nice looking chassis. Lots of freedom to put a variety of shells on it. 

22 minutes ago, Rudivdk said:

Looking good as always

Did you test the front suspension setup IRL? I have used similar setups in my builds, but because of the caster the drive axle might sway when steering. It could end up squeezing against those vertical half beams...

I tested it manually, not with motors yet. All pivots are in one plane and the pivot point of the CV-joint inserted in the wheel hub lies exactly in the middle of the pivots for the suspension arms. So it's all symmetrical. The only concern I have is that the CV-joints can insert slightly too deep into the hubs (0.2 studs), causing them to de-center slightly and maybe pull the other CV-joint out of the differential.

1 hour ago, Didumos69 said:

So it's all symmetrical.

Well of course, it's a Didumos build, why did I even doubt that...

Have you considered using your small turntable based hubs? You've done the testing yourself to see they have less friction, but maybe the wheels would be forced even further outboard?

I spotted this on @Charbel's 720s. Seems like a brilliant way to dodge the inherent slack in Lego's small Balljoints. Any chance to use it?

Ps, I'm definitely in for this thread :)

Of course, that anti-balljoint won't work with the caster angle.

Turntable hubs because they might let you flip the CV joint around, giving it a hard stop against the wheel centre. I'm not sure about that though, don't have the new wheels.

1 hour ago, amorti said:

Have you considered using your small turntable based hubs? You've done the testing yourself to see they have less friction, but maybe the wheels would be forced even further outboard?

I have that as a backup and it indeed gives half a stud extra offset. To tie the turntables to the Defender rims was a challenge though and required quite a few parts. I'm willing to accept the slack in the wheel hubs against the extra parts for the turntables.

1 hour ago, amorti said:

I spotted this on @Charbel's 720s. Seems like a brilliant way to dodge the inherent slack in Lego's small Balljoints. Any chance to use it?

That chassis is a marvel btw. Using a different pivot point horizontally and vertically won't work with the CV joints, which have one pivot point for vertical and horizontal rotation.

Edited May 9, 20204 yr by Didumos69

@amorti, turning the CV-joints around will make the wide cup of the CV-joint collide with the axle holes of the turntable. However, by squeezing a 8t gear (without doing damage) over the axle side of the CV-joint inside the wheel side of the turntable, the CV-joint gets a firm grip on that side of the turntable.

Edited May 9, 20204 yr by Didumos69

I completely missed that charbel's connection pivots on two different points. I guess it doesn't matter on an undriven wheel but I see now that it can't work for this project.

Looks like you're talking yourself round to the turntables... just be sure the complexity is worth it before you do it. How fast do you think the model will drive? Will better hubs make a difference?

Edited May 9, 20204 yr by amorti

55 minutes ago, amorti said:

Looks like you're talking yourself round to the turntables... just be sure the complexity is worth it before you do it. How fast do you think the model will drive? Will better hubs make a difference?

I will most likely try with both options and test whether there is a difference. I will have to find out how fast it goes. Weight is a factor too of course, hence the preference for using less parts. Although you need some weight to get some inertia, I mean to get a tendency to respond slow to changing forces. So it won't respond to every single bump and will tilt in turns. To an acceptable extend of course. With too little weight you get what we see so often with LEGO, a silly bouncing vehicle.

Edited May 9, 20204 yr by Didumos69

As a side note: There is something weird going on with wheel weight. In a car, having heavy wheels has all kinds of negative effects. For an offroad car for instance, there would be lots of unsuspended weight bumping up and down if the wheels would be heavy. However, when you ride a bike up a hill on tarmac, you obviously want the bike to be as light as possible. But if you have a minimum weight requirement, like for instance for road races, you better put most of the weight in the wheels. For riding uphill that is. The reason is that you can put more kinetic energy in your bike with less speed. And more kinetic energy means your legs can put in more pulse like power shots instead of putting in an almost constant power level. More pulse like power curves are more natural. We all know how it feels to ride up a steep hill, when you stop pedalling, you lose your speed immediately. If your wheels would carry lots of kinetic energy, just like a flywheel, you could stop pedalling without slowing down that quickly.

But that's of course all not important to this thread.

Edited May 9, 20204 yr by Didumos69

Okay, time for an update. I'v build the front axles and made some test drives with it using a single rear wheel. Conclusion: The hubs with caster are too sloppy, having no Ackermann geometry causes too much tire scrub in turns and the performance is disappointing. So I decided to go all the way and use 2 Buwizz and 4 Xl motors: I changed the wheel hubs so nothing relies on pinhole-axle connections, which should reduce slack substantially. I still have a caster angle, but slightly more modest than the setup I started with. The setup also has Ackermann geometry and steering rods that are heavily secured.

And last but not least, I decided that this will the model to incorporate the dual diagonal drive idea that I had waiting. (The dual diagonal drive idea originates from this project, which combines dual diagonal drive with a 4-speed sequential gearbox. At the end of that project, I decided to no longer combine the two ideas, but to split them over future models, which is happening now.)

Background

I have been playing with this idea for a while already, especially after seeing @KevinMoo's dual drive models (Mitsubishi Pajero and Dual-Driveshaft Pickup). @KevinMoo rightfully addressed the vulnerability of LEGO parts in RC models and the fact that using independent drive trains for the left and right sides, loses the benefit of differentials while cornering. This got me thinking. Using independent drive trains for left and right in a 4WD model does indeed drop the benefit of differentials while cornering, but what if we would pair the wheels diagonally, so pair the left front (LF) wheel with the right rear (RR) wheel, and pair the right front (RF) wheel with the left rear (LR) wheel? The resulting 'dual diagonal drive' (I borrowed the term from the electric skateboard scene) would serve two major benefits:

• While cornering, the LF and RR wheels will average to a speed that is very close to the average speed of the RF and LR wheels. So not having an open distribution by means of a differential between the two drive trains is much less of a problem as with separate drive trains for the left and right side wheels.
• On a very uneven surface, where one or two wheels may lose contact with the ground, the wheels that do have contact are typically lined up diagonally, see image. With dual diagonal drive, the vehicle would still have traction, even without locking any differentials. Only on slippery surfaces, there are chances of spinning wheels.

So this model will have two separate drive trains, one for the LF (left front) and RR (right rear) wheels and one for the RF (right front) and LR (left rear) wheels. The drive trains cross using 16t gears and and two auxiliary 24t gears. So no clutch gears are involved in this crossing. Each drive train has 2 XL-motors. The outputs are geared-up 3:1 and then down 5:7 to engage with a new differential, one for each drivetrain. The overall wheel-motor ratio is 15:7.

Currently I'm in the process of wrapping it all in a firm structure. I intend to incorporate a V8 mid-engine driven by a M-motor.

Edited May 16, 20204 yr by Didumos69

I fear those bevel gears above each CV joint will be under tremendous stress.

I've been busy reinforcing the back of greyhounds swingarms (you can easily get a 5*3 I beam piece on the back) but no matter what I do, in a swift back/forwards move those gears will click. I think it's now to the point that they're meshing perfectly but still the material isn't strong enough.

Any chance to gear up further and use planetary hubs?

The planetary hubs have (I think) less slack in them than normal hubs, even if they have more than turntable hubs.

Edited May 16, 20204 yr by amorti

8 minutes ago, amorti said:

in a swift back/forwards move those gears will click. I think it's now to the point that they're meshing perfectly but still the material isn't strong enough.

Those swift back/forwards moves are indeed too much for LEGO material, whatever you try.

9 minutes ago, amorti said:

What if on the way out of each XL you went not 24:8 but 8:24, and used the planetary hubs? Overall vehicle speed would be very similar but with increased speed and decreased torque all the way through the drivetrain.

For those bevel gears this would mean less speed and more torque at the same vehicle speed.

Yeh, I need to slow down and read my replies before sending, not to mention "read" the renders. My mind read the gears the other way round.

Any way to use planetary hubs to gear up before the differentials, before gearing back down at each wheel hub?

6 minutes ago, Didumos69 said:

Those swift back/forwards moves are indeed too much for LEGO material, whatever you try.

Today's idle Google search was wondering if any RC cars have 16mm bevel gears in their differentials...

You are right that the bevel gears are the bottle neck in the gearing. If I can find a simple way to gear up more and gear down in the wheel hubs, I will do so. But the 41099 planetary hubs have a lot of friction, so I fear I will never get any speed comparable to greyhound with them. If my setup is optimal for forward drive, I am not going to bother too much about the swift back/forwards moves. Moreover, the whole gearing as designed now, is based on what I know the bevel gears can handle in greyhound. In this case I use XL-motors, but I also use a 15:7 ratio, which I reckon should give me more or less the same torque as with L-motors at the same vehicle speed. It's all theoretical of course. I will have to see in real life. 

Is there any stronger 90° transition? Are the black bevel gears stronger?

Knob gears maybe? But probably more friction.

34 minutes ago, pleegwat said:

Knob gears maybe? But probably more friction.

At least that won't slip. I will keep it in the back of my mind. Thanks!

59 minutes ago, amorti said:

Is there any stronger 90° transition? Are the black bevel gears stronger?

That might be an option?

https://www.bricklink.com/v2/catalog/catalogitem.page?P=g9#T=S&O={"iconly":0}

Are you sure you want to use those 8 tooth gears in such a powerfull drivetrain?
I have a feeling that they are relatively inefficient.

 

39 minutes ago, pleegwat said:

Knob gears maybe?

Those have been used in several crawlers. But with high power transfer the inefficient tooth design heats up and they die easily.

1 hour ago, Didumos69 said:

In this case I use XL-motors, but I also use a 15:7 ratio, which I reckon should give me more or less the same torque as with L-motors at the same vehicle speed. 

I'm no expert in Lego PF. Wondering why to use the notably bigger XL motors if they'll be geared faster than an L motor and you're looking for the same torque from them as an L motor?

1 hour ago, amorti said:

That might be an option?

Who knows, but they are massively big. I have a few of them laying around though. I'm that old 

59 minutes ago, schraubedrin said:

Are you sure you want to use those 8 tooth gears in such a powerfull drivetrain?
I have a feeling that they are relatively inefficient.

My experience is they are especially inefficient under high torque, so for instance when used directly on the output of the motor. But I use a 24t gear to drive the 8t gear. I do need to gear up towards the differentials. I could use a 36t-12mesh, but things would grow very big.

1 hour ago, schraubedrin said:

Those have been used in several crawlers. But with high power transfer the inefficient tooth design heats up and they die easily.

Too bad.

Edited May 16, 20204 yr by Didumos69

53 minutes ago, amorti said:

I'm no expert in Lego PF. Wondering why to use the notably bigger XL motors if they'll be geared faster than an L motor and you're looking for the same torque from them as an L motor?

Valid point. I was not being accurate enough. With 4 XL-motors the whole model should be able to go faster than with 4 L-motors, so these bevel gears will suffer more too. If I don't want more torque on the bevel gears than in greyhound I should indeed use L-motors. With L-motors I could actually use a 12:20 mesh in the 90 degree transitions, because they spin faster. I think I'll need to test drive with a few layouts to find out what the best option is.

Edited May 16, 20204 yr by Didumos69

L motors might be a better bet then?

I have some scheme in my mind where they're opposite each other into a differential, they can then share gear-down paths into their diagonal diffs? Idk... You're the savant here :)

In testing, be sure to replicate the total model weight. I was certainly surprised how many more grams came to greyhound after the technical part was done.