Mud

I don't think this would be very good for the battery though - I think occasional charge/discharge cycles are required for Li-ion batteries to keep them healthy in the medium-term. Since the battery wouldn't be doing much in this setup, I think it'd be better to bypass it.

That looks neat although I'd have to see it going to intuitively understand it. Patent-speak isn't very intelligible at the best of times...

I used the 28 tooth differential because if you go from 28 teeth to a 12 toother (in my case via a 20 tooth) that's a ratio of 2.333:1. You can match that ratio with a 24 tooth gear to the 56 teeth around the outside of the turntable. I may have missed a trick though, I'd certainly be interested to see if there's an easy way to use the 24 tooth differential instead since there's the possibility of a much tidier package. I posted a video earlier in this thread. Passing a single function through a turntable is probably pretty easy with the 24/16 tooth differential since you can use an 8 tooth gear on the inside teeth of the turntable for an easy 1:3 ratio.

Yeah, I'm not sure if I had rose-tinted glasses regarding the speed or if my batteries have passed their best. It should be about the same speed as any other 1:1 geared M-motor tracked vehicle I guess (it's quite skeletal and therefore not very heavy). It runs smoothly on carpet, but shakes on a hard surface. The third motor and fake engine help to reduce it from rearing up. Panasonic HDC-TM900 - it's good as long as you have lots of light.

Now with video in the OP. I'll try to up production values on my next build and my next build will be more challenging ;)

Sorry for any misunderstanding - I'm swamped at the moment and haven't read your tutorial properly. I will though.

The wide apertures were mostly to compensate somewhat for my wrinkly muslin...and it makes the model look smaller (and therefore cuter) than it is ;) I used an off-camera flash through a small (~20") softbox, but obviously that has the potential to give a lot of specular reflection. You won't see that in the EXIF as it's all manually set. I must say Lego is really tricky to photograph, people and places are so much easier. In your tutorial it looks like you have a lot of room to play with, wish I had that luxury. I'm not into macro photography really, so not sure I'll invest in new gear just for lego shots (unless I see something cheap) - I'll see if I can make something approximating your setup though :)

Aha, I said I had all the gear but I don't have a light tent - that would probably make all the difference. I'll try to do a video soon.

Connectab introduced me to the Yanmar T80, and it was too fun-looking not to make. This model is mechanically pretty boring with no auxiliary functions, but I think it's cute and goes like a stabbed rat. I'm not too hung up on the details, but tried to get the proportions about right. It's a simple skid-steer with 1xM-motor/side. The third M-motor just spins a fake inline 4 pot hidden in the front (no idea what a T80 really runs). Not that you can see it, but the crank is a 'proper' inline 4 configuration, I die a little every time I see the lego inline 4 and V8 crankshafts It obviously takes some cues from 8063 (it's where most of the parts came from), but it's comprehensively reworked under the skin and a different scale (these T80s appear to be relatively small). Video to follow if I get a chance, but playability is limited to skidding it around on laminate flooring OT, does anyone have a tutorial for photographing lego? I've got all the kit, but am making a complete hash of it...blown highlights and even complete colour channels a-gogo

Wow, there're 6 motors hidden in there!?

Those guys are great

That is really tidy...might be able to modify that into a McPherson strut arrangement too

I have done it, . I would be interested to see if anyone has a simpler solution.

Burf, there's really not much to it - my struggling in MATLAB was probably just a quirk of the serial communication or me incorrectly formatting the data. I may revisit making it work in MATLAB...I want to do as much as possible myself though to maximise portability, so I won't be using other people's toolboxes. In C# the code is relatively simple, here for example - you simply connect over bluetooth, acquire the com port, then send and receive the communications in accordance with the manuals. The first two bytes give the packet length, so it's best to write a subroutine to sort those bytes, then the calls become simple, like for the message I mentioned in the OP: byte[] NxtMessage = {0x01, 0x88}; NXTSendCommandAndGetReply(NxtMessage);

I've got something working now, albeit in C#