Davidz90

Very impressive! I like Your earlier compound bow as well, very clean construction. Surprisingly poweful for the springs used. I wonder how much more these limbs could handle. Reinforcing them with arched panel pieces is a nice idea. A few years ago I built a reverse draw crossbow as well, reached 50 m/s velocity. You can see it here: https://www.youtube.com/watch?v=zRED3K14D84&ab_channel=DavidZiemkiewicz I used shorter, faster limbs (reducing limb inertia is crucial) and way bigger springs. Worked nicely for few years, but eventually the limbs gave up and broke.

Wow, the suspension in impressively compact. And if I see correctly, there is also a four wheel steering?

I think he meant piston engine, not turbine. 2800 rpm is pretty impressive!

Happened to me several times in the past. Very annoying in applications where low friction is crucial.

Cool! Very nice way of securing the ring pieces in place.

I guess the point is to have ring rotating independently from the center axle. Unfortunately, as others pointed out, it cannot work with 28T gears. The small ring meshes nicely with 12T gears, but needs at least 3 (and better 4) to really stay in place: 20230303_131251 by David_Z1, on Flickr

Seems so. Unfortunately I bought it already; right now I'm trying to convince myself that I need a second copy

Absolutely beautiful!

Maybe with approach used on actual (older) aircraft, i.e. control cables. There should be little space for few strings and pulleys.

I've put the grandfather clock in its final place in the living room and recorded a short hyperlapse video: 1 hour of running in one minute. The speed is basically dead on (without seconds hand, I cannot tell the difference after 24 hours, and any longer measurement is subject to long-term factors such as thermal expansion of the pendulum). The video also highlights the drop rate - in the span of 1 hour, the main weight drops by approx. 9 cm and the counterweight goes up by 9 cm. With 145 cm of space, this results in 16 hours working time.

Where is the version with anti-aircraft cannon? Jokes aside, really nice, thought out modular design.

The operation time is directly related to the amount of energy the driving weight stores - how heavy it is and how far it goes down. It is only indirectly dependent on the pendulum - longer, slower pendulum loses less energy due to air resistance, so it needs a little less power. Real clocks are a good deal more efficient than Lego ones, but even with Lego a one week running time is not out of reach; this particular one would need a driving weight of approx. 3 kilograms to run for one week. My previous clock uses four 1.5 kg weights and runs for 10 days. Overall, pendulum clocks are fantastically efficient machines - this one has a power of 70 microwatts (0.00007 W). Done: https://bricksafe.com/pages/David_Ziemkiewicz/functional-grandfather-clock Rebrickable requires a full brick inventory and.. I'm not doing that

Sure, I'll look into it tomorrow. Understandable. Frankly, I wasn't sure if there is a point in including the tower instructions.

I made full instructions for mechanism, tower and face :D Link: https://drive.google.com/file/d/1EwCfTyq6BigSetNr13E3AM7ZKMQYWOEe/view?usp=share_link

Thanks! Yes, I plan to make instructions. Maybe even for the whole thing - tower build is quite repetitive. Thanks! Yes, the mechanism is back drivable and time can be set by rotating the minute hand. In addition, each hand has its own slip clutch so they can be adjusted independently - quite handy for small adjustments of their relative position.

Thanks! Good thing they come in red color. At first, I used a black one but that turned out to be completely unreadable.

Here's a helpful video that explains addition/subtraction with differential and how it can be used to get prime number gear ratios (like 1:17) that are impossible otherwise:

Thanks! The lack of ticking sound is sometimes troublesome, my standard way of measuring accuracy is to record ticks and run the recording through computer program to get average periods, standard deviation etc. Here I had to use more rudimentary methods - sitting in front of the pendulum with a stopwatch.

I present my new build: over1 by David_Z1, on Flickr It is a 170 cm high grandfather clock, built with about 9000 bricks. The structure is meant to be bio-inspired, with a double helix (dna-like) pattern. The main load bearing element are four technic brick reinforced pillars at the edges, which are connected by the said helix-like grider structure. The power is provided by a pretty light (by my standards anyway :P) 0.3 kg weight. It is sufficient for 16 hours of operation. mainw by David_Z1, on Flickr The clock face is constructed mostly with 1x2 plates, the whole structure is under slight tension. over2 by David_Z1, on Flickr The hands assembly uses my new solution with worm gear. The minute hand (axle) is geared up 2:1 and then geared down 1:24 to differential casing, which holds hours hand. hands by David_Z1, on Flickr The escapement used here is a so-called Grasshopper escapement, invented by John Harrison. My particular implementation uses two wheels to make the whole mechanism completely symmetric. The system is highly efficient (needs 70 microwatts to run) and accurate (clock error is under 1 minute per day) escapement by David_Z1, on Flickr Video:

I'm not sure how relevant that is, but recently I found this mechanical calculator: and it seems very doable in Lego form. It even uses 40 tooth gears!

Great to hear! Good luck!

If I can add my 2 cents, base-10 approach might actually be easier with Lego. Binary mechanical calculators are a rarity, and compared to base 10 ones there is usually much more parts. Binary numbers are simply bigger (more digits), so there are more carries etc. Less precision is needed in the sense that all is cool as long as 0's are not mistaken as 1's and vice versa, but in electrical system it means that the signal in the circuit is amplified to proper level every few logic gates. How about simply gearing things up so we have a rotary dial with 10 numbers, but inside the calculator 1=1 rotation? That could be easily processed in a differential for addition/subtraction. It is hard to get a backlash so bad that output is full rotation behind input.

Now that is something unique! Amazing!

The key here is that they are connected in parallel, so the total current is quite high. That high current passes through AA batteries, which have quite big internal resistance. Regarding question 2, the supply voltage must be greater than total led voltage. In principle, the 1 ohm resistor could sort of work, although that setup works a bit differently than the one with two leds. Instead of reducing the current, we rely on the fact that the voltage drop over the resistor reduces the voltage of leds so they don't exceed their specs. But as I said, trying to control leds by voltage only is a bad idea.

Yes, Ohm's law indicates that 0 resistance is needed. However, by the same Ohm's law the current I=U/R, where U=9V and R would be resistance of 3 leds. Since that is very small, the current would be enormous (like, several amps). Sorry, I have no knowledge of Lego part specs so I cannot fully answer questions 1 and 3. Regarding 1 I'd say that yes, 1:1 replacement without changes should be fine.