 | | A better full adder!
|
| I realized this afternoon that I could cull a switch from my earlier design. It's now down to 12 switches from 13. 4 switches ganged onto each piston. The wiring schematic is here: (URL) on how to interpret the diagram can be found in an earlier (...) (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| Hey, that's great! Now it's nice and symmetrical with 3 columns of 4 switches... something satisfying about that. :-) Now, all I need is 4 more switches and I can build it! Cheers, Mark (R) (...) (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Kevin (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) It sure is. Thanks. Using the same pattern (setting up specific switches "backwards"), I can do a half adder with two pistons and 4 switches (2 switches on each piston). Again, no need for inter-piston bracing, just the requirement that a (...) (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Finally I can compete again! ;^) I can do a half adder with 1 piston and seven switches, plus you get dual polarity pressure. Three switches for AND and four for XOR. This says that I can do single pressure logic line with one piston and three (...) (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) AND takes 1 piston + 1 switch using your design. XOR takes 1 piston + 2 switches using my design. OR takes 1 piston + 2 switches using my design. But AND and XOR can be done using the same piston, so 1 piston + 3 switches gets you a half (...) (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Ah... it took me a while to figure out what you were doing, but I see now... (reviewing your pics) If your inputs to the XOR were A and B, you use A to drive the piston, and require the presence of both B and NOT B to make the piston move. I (...) (21 years ago, 26-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Yes. B and NOT B into the mux formed by two switches that are arranged to never release. (...) But with LEGO pneumatics you can typically get it for free. (...) Yes. When I wrote my post up yesterday I'd forgotten that I needed NOT B as an (...) (21 years ago, 27-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Not sure what you mean. If an output isn't producing any airflow, how do you invert that? BTW, I've been wracking my brains on this for several hours now, I can't figger out how you possibly could have built a full adder with only 3 pistons (...) (21 years ago, 27-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Ahh... you are stuck in single pressure line mindset. The modern LEGO pneumatic pistons are dual ported where pressure into the port at the base makes it expand, and pressure into the port at the top makes it contract (presuming of course that (...) (21 years ago, 28-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| Kevin L. Clague wrote: > [ more brilliant stuff ] >> Stick a fork in me... I'm done. > (...) Thanks. I try. :) Actually, I do have a few more ideas I've been toying around with in my brain, but I think I'm about as done with this as I'm going to be (...) (21 years ago, 28-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| Okay, Kevin... you talked me into trying this inverted logic thingamy. (I'm just too competitive for my own good...) How's this? (URL) A and NOT A, B and NOT B, and Carry in as inputs and produces A+B+C, NOT A+B+C, and Carry Out as outputs. I saw no (...) (21 years ago, 28-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Very impressive! Two pistons! (...) Well, I can eliminate one switch on the Cin/A/NOT A piston. You don't need to throw the switches the full range do you? You can get the release from A/NOT A, and have to have it come from the local switches, (...) (21 years ago, 28-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Thanks. Here's how it works: Consider only the right piston for a moment. The output of this subsystem (which will be driving the left piston) is simply (A&~C)|(~A&C), which is an XOR operator, or a half-adder. So what drives the left piston (...) (21 years ago, 28-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Oops... after rereading what I wrote, I see made a typo there, and I can see how it could be very, very confusing. What I meant to say was that the bottommost switch on the left piston will already be down to receive A & _C_, which is high and (...) (21 years ago, 28-Jun-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) While working with Mark's full adder design, the XOR gate in particular, I found something which may be useful in some pneumatic constructions. In Mark's fulladder (URL) just the XOR gate (the 3 switches on the piston on the right). There are (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Egad! Another madman joins the fray! ;) (...) Am I misunderstanding you or are you suggesting that the pressure created by the expansion of one piston would actually be sufficient to compress *TWO* others? Unless the pistons had no loads at (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) I think the idea here is that they are mutually exclusive - only one will be expanded at any one time, so when a different one is expanded, it effectively only has to retract one other - the one that was previously expanded. ROSCO (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) I've alway found low level circuit design interesting. Using pneumatics and LEGO pieces to create them seems like a good challenge. I have some ideas that might turn into workable designs. (...) I've tested it, and it does work. I don't know (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) I wasn't trying to imply that you hadn't actually tried it... I was only suggesting that I didn't think that pistons being driven solely by the compression of air created by the expanding of other pistons would be strong enough to compress (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) No offense was taken. My appologies if I gave that impression. I just completed some load testing on the compressing piston. It easily switched 2 switches, was just barely able to switch 3 switches, and clearly couldn't do 4 switches. Since (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Geat job Mark! I started out with a 14 piston, seven switch full adder, and by piggybacking off each others ideas, you've gotten us down to 2 pistons and a handfull of switches. Absolutely amazing! Unless I come up with some revolutionary (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Not sure if there's an algorithm, per se... but here's how I do it. Conceptually, I think of the backwards switch arrangement as an XOR gate. Input (from the output valves) will only be accepted from either the top or bottom valve. In general, (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Yes. Thanks. I'll have to try some of my own designs to really comprehend it. I learn by doing (and failing ;^). (...) Kevin (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Like I'm done? Just wait til I get my hands on some real ones... :) >> Mark (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) Nope.... I was pretty sure you were not done. Neither am I. I'm just grateful for what you've done so far. It was pretty lonely there for a while in pneumatic computing land. I'm happy for the company and the inventiveness. (...) I ordered an (...) (21 years ago, 1-Jul-03, to lugnet.technic, lugnet.robotics)
| | | | | Re: A better full adder!
|
| (...) As a follow-up, over time the air pressure in the closed compression loop drops due to slow leakages, so it requires periodic recharging. One way is to remove and reseat the tubing, but a beter way is to add another T and a hand pump to the (...) (21 years ago, 2-Jul-03, to lugnet.technic, lugnet.robotics)
| |
