rgbrown

No, you need the gearing - this one is a nontrivial exercise. If v1 and v2 are the speeds that the two wheels are travelling along the ground, then the corresponding rotational speeds of each wheel are w1 = v1 / r w2 = v2 / r where r is the radius of each wheel. Doing the maths with the differentials, the rotational speed of the crown gear that links the differentials is then given by wc = 1/4 (w1 - w2) = 1/(4r) (v1 - v2). The angular speed of the overall chariot is given by wt = 1/L (v1 - v2) where L is the distance between the two wheels. So L would have to be four times the radius of one of the wheels for the system to be balanced without additional gearing. For the depicted chariot, I estimate L to be about 17, and r to be about 3. So the compass wheel needs to be geared down by a ratio 4r/L ~ 12/17 ~ 0.7. The 24:40 gives a gearing down of 0.6 (i.e. the right ballpark). I'm sure if you were to actually measure the distances, you'd find those numbers would be very close.

Just as long as it's not one of these!:

lol - I had no idea anyone anywhere said tek-neek until this video came up! In New Zealand it's always been teck-nick. But we aren't very good at vowel sounds (tick-nuck) over here so maybe we say it that way because it's easier to say.

Or maybe one of these:

I've noticed that they've been quite inconsistent in the past with how tightly the grip the axles. Some of the old ones gripped very tightly while some just slipped loosely on and off. Perhaps it is difficult to manufacture the old one to tight enough tolerances? At any rate, I don't think it's a strength problem - unlike the old z24s which were useless

Ouch - bring on the repetitive strain injuries

I guess I must sound like a bit of a wet blanket to you. I'm not denying that yours works, nor am I denying that running motors on a higher voltage will give you more speed and torque. And I agree that your excavator runs very nicely (and that you're a very good operator), and I wish they all worked like that. I just don't want people to get the idea that exceeding maximum voltage ratings on integrated circuits is a good idea, or at least people should be well aware of the risks. I'm trained as an electrical engineer, and doing that in a design is an absolute no-no. Of course these are only toys, and if the product breaks it doesn't matter, so normal engineering rules don't apply - but there is a definite risk of this happening. Chip manufacturers don't just pull those numbers out of the air. And once again, I'm talking specifically about the receiver, not motors. Anyway, no more from me on this! I think I've made my point quite often enough now

The component in question here is the LB1836 motor driver IC that is in the IR receiver. The manufacturer states on the data sheet that the allowable operating range is 2.5 to 9V, and the absolute maximum input voltage range is -0.3 to 10.5V. The IC manufacturer may have incorporated a safety factor into their specification of the absolute maximum input voltage (10.5V), but as far as I'm aware, there is no standard industry figure for this. All it is saying is that the manufacture is guaranteeing that if you take a sample of these chips and apply 10.5V to them, they won't break down. Some chips may be able to tolerate much higher, some (even from the same batch) may be able to tolerate barely more than this In brief: PF is using this component at the upper end of how it is intended to be used, I would personally advise against using higher voltages on the IR receiver, but, as shown by this thread, doing so won't automatically destroy the device. Again, it should be noted that the breakdown may occur over time. If I felt I had one spare (which I don't), I would destroy one this way just to make the point - any volunteers?

More complex, yes. More difficult, not really. This model is fundamentally just a large-scale exercise in gear combinations (i.e. elementary arithmetic). The difficulty in building it is one of getting all the gears to fit in a frame that is not too big, with the right axles protruding from the right places. And yes, figuring that out and building it must have been hard work. (I'm ignoring the work that has gone into calculating the dial patterns, etc, which is a different kettle of fish) The difficulty in building an efficient mechanical clock is to design a movement that minimises frictional losses, and an escapement that efficiently transfers just the right amount of energy to the pendulum per cycle without losing much. Much more design scope and a much more difficult and open-ended problem, in my opinion. Don't get me wrong though, this is an incredible creation, and a nice complement to the recent research on the device.

From: http://us.service.lego.com/en-US/BuildingInstructions/default.aspx (e.g. search for 8043) "If you plan to print the building instruction, please be sure to download the correct version: Building instructions labeled “NA” or "V39" may be printed on US standard letter size paper (8½ in × 11 in, 215.9 mm × 279.4 mm). Building instructions labeled “IN” or "V29" may be printed on EU standard A4 paper (210 mm × 297mm, 8.3 in × 11.7 in.)" Sheds some light

Two independent PF circuits with receivers on the same channel (2 battery boxes) is one option EDIT: Sorry, I didn't read the thread properly - many have suggested this already

Really? The 8878 has an overcurrent dropout too ... well at least I'm pretty sure it does, but all the product specs seem to be gone from powerfunctions.lego.com. Where are those pages that have detailed descriptions of each PF element? Am I just incompetent at browsing TLG website, or are they actually gone?

Can we just all agree that using 11.1V 1. works for Rien (at least for now), and will probably work similarly for anyone else that tries it. 2. is done at your own risk, as you're exceeding manufacturer's absolute maximum voltage ratings. i.e. if it blows up, it's your own fault, and you've had adequate warning in this forum. And then have a cup of tea?

You are fast becoming my favourite builder! I really like this model. I think I'm going to try to reproduce it for my kids - my daughter would absolutely love it. How well do you think it would work with an m-motor?

Very awesome! Parts scanned this way will be a great boon for the LDraw parts library. It should be said though, all of the hard stuff here is in the computer vision components of the system - not the LEGO components. The NXT is just programmed to move the laser slowly. So you can't really call it a LEGO 3D scanner (and neither does he). As an (almost) pure LEGO project, the probe scanner he built is pretty incredible: http://philohome.com/scan3d/scan3d.htm.

Loving the rear suspension!

I shouldn't, but I find this unreasonably hilarious. It is one of the more bizarre technic videos I've seen. Anybody else got any good ones?

Hmmm ... I would have said the bottom line from that (excellent) thread was that it is not a good idea to go above 10.5V for the PF Receiver (outside of the motor controller manufacturer's stated limits). Mortymore did an awesome job of reverse engineering the PF Receiver for us. So it may well work, at least for a while, but I certainly wouldn't want to recommend anyone try it! 10.5V is the stated absolute maximum rating, i.e. that means the chip-maker is saying there is a reasonable chance that the motor controller will start to deteriorate if you exceed those limits. Even if it does, you probably won't see it straight away. Personally, I would keep well clear of the 10.5V limit -- I say stick with at maximum 6 alkalines (9.6V) for the PF receiver. The motors are a different story - I'd be willing to experiment with higher voltages on those, so long as you're very careful not to let them stall.

Yay! Mine just arrived in NZ after 3 months of radio silence

What you're describing sounds more like a solenoid to me ... I think there is actually scope here for two devices A PF S-motor that has reasonably high speed, low torque, and maybe a built-in, very light, clutch for steering regular (not offroad) RC cars. This may be what the racers base does? (I don't have one) I know that's the kind of thing you often see in toy RC cars. I can't think of that many other uses for it though. A PF solenoid - basically an electromechanical linear actuator, that provides very strong linear switching. It could be designed to go in and out and self-center using a spring. Such a device could suit TrTr steering, pneumatic valve switching, etc. We can dream

That is impressive, especially with the plate, as the slightest bit of inaccuracy will make it not fit into the hole Makes sense. Does the building robot lower the part down until it meets resistance, at which point it lets go, or does it let go at a predefined point in space? Again, really awesome model! So many possibilities!

That's fantastic - what a labour of love! Very nicely done video too. I can't wait to see the full videos with some big models being built! Don't suppose you're touring it to New Zealand any time soon? I have a tonne of questions, feel free to answer as many or few as you like Looks like this model has got to the point where it is almost entirely programming now. Does the building robot run in open loop, or does it use feedback to sense when to release the bricks? What happens when something goes wrong - does it realise and tell you? How pipelined is the production process - do the fetching, rotating, and building robots work independently, or does the entire fetch-to-place procedure happen before the next one starts? What are the limitations on the models? Do they have to be formed in separate layers, or can you mix and match heights? (e.g. could you put a plate beside a brick, and then build on the plate? Any plans to try to increase the speed? Anyway, I look forward to seeing more of this project!

Hard job to pick one of those! Great to see so many from here represented. I only wish that they could allow videos and bigger pictures, and maybe a wee profile of each model, to really do the competition justice. Right, I've made my choice Just bad web design. Forcing each image to be the same size ... at least they're not squished when you click through. p.s. @mahjqa - v.odd, makes you wonder exactly how they picked the shortlist.

Nice review, but in particular, great job on the photos! It makes all the difference.