![](/i/fill.gif) |
![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Someone with far too much time on their hands:
http://goldfish.ikaruga.co.uk/logic.html
Enter mechanical computing. :)
--
~Mike
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> Someone with far too much time on their hands:
>
> http://goldfish.ikaruga.co.uk/logic.html
>
> Enter mechanical computing. :)
Imagine how hard you'd have to push if you connected together a handful of
these blocks!
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>> Enter mechanical computing. :)
There's nothing wrong with that. Some of the iconic first computers were
mechanical. ;-) (And I don't just mean Babbage's Analytical Engine; go
look up the Zuse Z1 computer.)
> Imagine how hard you'd have to push if you connected together a handful
> of these blocks!
Well, that's just it. The entire assembly is passive, whereas normally
logic gates have their output power derived from a source independent to
the input. (Especially buffers - that's essentially their entire point!)
It's not particularly easy to see how you'd do this with reciprocating
mechanical signals; if you were using continuous rotational motion, it
would be clearer that amplification is possible. (Although it's still
not obvious what the best way to achieve it is...)
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>> Imagine how hard you'd have to push if you connected together a handful
>> of these blocks!
>
> Well, that's just it. The entire assembly is passive, whereas normally
> logic gates have their output power derived from a source independent to
> the input. (Especially buffers - that's essentially their entire point!)
>
> It's not particularly easy to see how you'd do this with reciprocating
> mechanical signals; if you were using continuous rotational motion, it
> would be clearer that amplification is possible. (Although it's still not
> obvious what the best way to achieve it is...)
You'd have an big drive shaft going down the middle powered by a huge motor
rotating at a constant speed. Whenever a logic gate needs to output a logic
1, it would just connect its output shaft to the big shaft. Would make for
interesting circuit layouts :-)
With linear motion you could have a big "activate" rod that you push with a
big force to "run" the circuit. You set up your inputs by either connecting
them to this rod or not, and each logic gate could be arranged to either
connect or disconnect its output to the core activate rod (they could spring
back to "0" if not connected). I haven't thought this through yet if it
would actually work or not with more complex circuits!
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
scott wrote:
> You'd have an big drive shaft going down the middle powered by a huge
> motor rotating at a constant speed. Whenever a logic gate needs to
> output a logic 1, it would just connect its output shaft to the big
> shaft. Would make for interesting circuit layouts :-)
The difficulty is comming up with a mechanism which will connect one
shaft to another if a third shalf is rotating, and disconnect them when
it stops rotating.
> With linear motion you could have a big "activate" rod that you push
> with a big force to "run" the circuit. You set up your inputs by either
> connecting them to this rod or not, and each logic gate could be
> arranged to either connect or disconnect its output to the core activate
> rod (they could spring back to "0" if not connected). I haven't thought
> this through yet if it would actually work or not with more complex
> circuits!
I'm not sure that would work. A rod that's just staying still doesn't
impart power.
I think whatever system you choose, whether it's rotating axials or
water pressure or flowing electrons, the fundamental device that you
seem to need is a thing to connect or disconnect two halves of a signal
pathway based on an input. (And I think it might be neccessary to
configure it to react to different senses of the input.) If you can
manufacture that, it seems like you can make all the other fundamental
logic gates.
pathway --#-- pathway
|
control
You can now make a buffer:
(+) --#-- Output
|
Input
For an inverter, just invert the sense of the input.
An AND gate is no difficulty:
(+) --#----#-- Output
| |
In1 In2
Similarly for an OR gate:
(+) -+--#--#-- Output
| | |
| In1 |
| |
+--#--+
|
In2
If that construction doesn't work, take an AND gate and invert all the
signal pathways. Either way, with buffers, inverters and AND/OR gates,
you should be able to construct anything.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> The difficulty is comming up with a mechanism which will connect one shaft
> to another if a third shalf is rotating, and disconnect them when it stops
> rotating.
Just get a clutch (like the one in a car) that is controlled by a third
shaft rather than your foot. The third shaft could just have some masses
attached to it that move outwards when the shaft is rotating, like a
centrifugal speed governor.
> I'm not sure that would work. A rod that's just staying still doesn't
> impart power.
Sorry I didn't explain very well, my idea was that the gates would only
update their state when you pushed the rod. It would be like a kind of
master clock for every gate. You could then have something driving the rod
back and to, and the gates would all update on each forward push of the rod.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>> The difficulty is comming up with a mechanism which will connect one
>> shaft to another if a third shalf is rotating, and disconnect them
>> when it stops rotating.
>
> Just get a clutch (like the one in a car) that is controlled by a third
> shaft rather than your foot.
Does Lego have a part like that? ;-)
> Sorry I didn't explain very well, my idea was that the gates would only
> update their state when you pushed the rod. It would be like a kind of
> master clock for every gate. You could then have something driving the
> rod back and to, and the gates would all update on each forward push of
> the rod.
...so instead of continuous rotational drive, you have continuous
reciprocating drive, basically?
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>> Just get a clutch (like the one in a car) that is controlled by a third
>> shaft rather than your foot.
>
> Does Lego have a part like that? ;-)
http://www.brickengineer.com/pages/2007/11/27/smooth-lego-clutch/
> ...so instead of continuous rotational drive, you have continuous
> reciprocating drive, basically?
Yes basically, otherwise as you stated you can't transfer any power.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>>> Just get a clutch (like the one in a car) that is controlled by a
>>> third shaft rather than your foot.
>>
>> Does Lego have a part like that? ;-)
>
> http://www.brickengineer.com/pages/2007/11/27/smooth-lego-clutch/
OK, so there's your clutch. Now, how to convert the rotational motion of
an axial into the sliding motion to operate this clutch?
>> ...so instead of continuous rotational drive, you have continuous
>> reciprocating drive, basically?
>
> Yes basically, otherwise as you stated you can't transfer any power.
Right, OK. That could work...
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
>> http://www.brickengineer.com/pages/2007/11/27/smooth-lego-clutch/
>
> OK, so there's your clutch. Now, how to convert the rotational motion of
> an axial into the sliding motion to operate this clutch?
http://en.wikipedia.org/wiki/Centrifugal_governor
Shouldn't be too hard to rig something up that works this way out of lego,
and use the linear motion to control the position of the clutch.
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |