Definitely fake. For one thing, with rings of magnets like that, they always cancel each other out and move to some static equilibrium (ie they stop.) This is the same problem that those shifting-weight things have. You need either a moving magnetic field or a moving electrical field for this style of generator, and the energy is taken out of the moving side and turned into electricity.

I think it's important to point out that he says that the lightbulb is not normal.

The easy way to fake this would be to have a battery behind the fan, powering both the fan AND the light when he presses some switch that's not visible, or that's powered by, say, a hall effect sensor.

For some more fun fake circuits (this one is lame and easy) check out this guy's work.


He assembles the circuits in front of you and proceeds to break physics. Best part: They're puzzles, and they all really work (but there are lots of disguised components)

A recommendation for making the actual video smaller and faster: Interlacing

This technique is frequently used in TV transmissions. It allows you to cut the signal data in half without reducing the resolution. You update every other line in each frame, alternating. So, you would have effectively 15 fps for video, but by updating each other line like 30 fps you can still get a pretty good picture. This is particularly effective on blurry displays like the 84+ LCD, since the slight jagged edges caused by movement lag are softened. That's what I get for only reading the first page of posts.

I really like it though! It looks really nice! <3 the touhou music

IMO the only thing that really needs improving for TI calcs is TI's attitude toward programming.

Oh, and also the price. $100 is too much for a z80-based system that doesn't even give you all the RAM you could have.

Tilemapping isn't a mode, in that you'd have to write the routine yourself. In this regard, the lack of documentation for it is because it's not part of the language. You're not going to find official documentation on how to make gravity physics either, because they're not a built-in feature.

I would agree with you more on the bitmap command if there was actually any justifiable reason to use it. I regard it as a fairly worthless instruction, since stacking Pt-Ons is faster and the same number of bytes.

More documentation, with better details, would be better, though that can be said for many programming languages, not the least of which is x86 assembler, or, for another example, low-level openGL.

The thing about small community projects like this is that the list of people who can do the things you want is very small, and the list of people who are willing to is even smaller. If you want something changed, you'll generally have to do it yourself, since the fact that it hasn't yet been done is evidence that it's not going to happen.

The chips in a TI-83+ are different from those in a TI-84+.

You would need to find the replacement chips, figure out how to solder them into place (They aren't the same size, or in the same places, so good luck) and then wire a new USB port in, along with its controls.

The architecture (the style of stuff used inside) is similar, but the designs are different enough to make this idea impractical.

Why do you want to do this? For the extra archive space? For the extra speed? For (please not this) MathPrint? There really aren't any other reasons than these, and all of them entail more complications than would appear at the surface. Let's go through them

1. For extra archive space, you would need to make 2 changes. First, you would have to physically replace the Flash ROM in your calculator with a larger chip, ideally already programmed for this purpose. This requires that your current TI-83+ has one of the hardware revisions that has a separate Flash ROM. After you got the new chip into place, you would need a modified TI-83+ OS that would recognize the extra space and use it. That's probably not a very complex modification for an experienced OS modder, but that doesn't make it easy.

2. For speed, you are pretty much out of luck. If you have a TI-83+ SE, you already have the speed of the 84+, because it also included the more advanced timing architecture. However, the TI-83+ has no extra timing hardware and nowhere to put it. You would also have to make extensive OS modifications to support any modifications you did make, since they would probably not be close enough to the real 83+ SE to use the same OS.

3. MathPrint is not good. It slows everything down for a little bit of pretty text, and it's extremely unstable. It crashes a lot, and it makes lots of BASIC programs incompatible across calculators. I don't know why you would want it.

That depends on how you phrase the input. When using solve() you need to put it in terms of a variable.

solve(log(x,2) = 2,x) will solve for x, saying x = 4


TI-89 log supports the log(entry,base) syntax btw

The biggest reason to have a decent on-calc text editor, I think, is so we can move away from the idea of using the built-in BASIC editor for programming. It's a bad editor for non-tokenized languages. If we had a real text editor then we might even be able to get named functions in an on-calc language.