Talk:Logic gate

The images are for the American standard for gate symbols. The European, which I belive is the real standard (approved by ISO/ANSI or something), standard looks different. An AND gate is a square with an ampersand in it, OR is a square with >= 1 in it and so on. I couldn't find any pics for them though. :-(

Sounds like IEC 617-12. Here's a page with some pictures (http://perso.wanadoo.fr/logidoc/iec61712/lbe_glossaire.htm) (in French). --Zundark 22:31 17 May 2003 (UTC)

Contents

1 Tesla patents 2 very misleading 3 tri-state logic 4 diagrams 5 Logic gates and quantum physics 6 The real history

Tesla patents

The article mentions Nikola Tesla having patented logic gates as US645576, but I looked up that patent (http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/srchnum.htm&r=1&f=G&l=50&s1=645576.WKU.&OS=PN/645576&RS=PN/645576), "System of transmission of electrical energy", and it appears to describe a means of transmitting electricity through the air, but nothing to do with logic gates. I've taken the reference to that patent out of the article. --Arteitle 19:51, Mar 7, 2004 (UTC)

The relevant patents are US 723,188 and US 725,605 granted in 1903. They were filed as one application in July 1900 but subsequently divided. I have added them to the text. Conception of the idea was at least as far back as 1896, '...I have an instinctive knowledge of the same...' Tesla said during the interference deposition for the patent, (US Patent Office Interference No 21,701, Systems of Signaling, Nikola Tesla vs. Reginald A. Fessenden, 1902, reply to question 9, August 5th 1902)--Smark33021 21:32, 23 Dec 2004 (UTC)

I've looked at the patents and they are for a signalling system, not a logic gate. Can someone tell me what relevance these patents have to logic gates? There's no mention of a logic gate in the claims, and the signalling system seems to rely on signals at different times, not at the same time, to operate a receiver. I wish I could point to an example of an "and" function used before the dates of these patents, I'm pretty sure this pre-dates electronics in general - if nothing else the Strowger switch must have had logic at least as complicated as an "and" gate and it was patented a decade earlier. --Wtshymanski 02:31, 27 Apr 2005 (UTC)

very misleading

I'm new here, so I don't want to make any changes. I shall note, however, that logic gates used in digital integrated circuitry do not operate in this manner. The author is conveying the idea that a logic gate either allows a signal to pass (in his diagram from left to right) or doesn't allow a signal to pass. While this seems sensible to the novice, it is not correct.

The reason that the author's conception of logic gate cannot be used in digital circuitry is that when a signal is not allowed to pass--for instance, if one transistor in his AND gate is OFF--the signal (technically the voltage level) on the OUT end of his logic gate is ambiguous, or "floating" in electrical engineering parlance.

I agree with you. The inaccuracy could be fixed in two ways: (1) delete the switch diagrams and the text that refers to them, or (2) correct the switch diagrams by adding pull-up resistors (as in open-collector logic) or complementary switches (as in CMOS and TTL logic) or adding loads to the circuits (as in relay logic). Option 2 sounds horribly complicated, so I am tempted to go for option 1. Anybody else care to comment? -- Heron 21:32, 5 Jul 2004 (UTC)

I take the point, but I'm not sure I agree. The diagrams are not meant to indicate how an actual logic gate necessarily works, but what it is doing in principle. I think that's a better intention for the article - to show how an AND function for example can be illustrated in a manner that is fairly intuitive to most people. Most people can understand the operation of simple switches, whereas the operation of a transistor is less clear to most. In addition, there are hundreds of ways to actually implement gates in real integrated circuits - a TTL circuit will differ from a CMOS, so which one would you use? None of them represent a single, "one true way" to make a gate, even if CMOS for example might be possibly the most usual way. Perhaps the wording should be altered to make it clear that the diagrams are only conceptual (though in terms of logic they could be built and would generate the truth tables given - so they are also practical in that sense). The article can provide links to the pages that deal with particular technologies, and those pages could include diagrams showing how logic gates are built with those technologies. Graham 22:52, 5 Jul 2004 (UTC)

I just took a chainsaw to the article and separated the content into a switch/relay logic section and a semiconductor logic section. This way, I hope to prevent any mistaken idea that the two types of logic might be interchangeable, whilst preserving the content that user:Dysprosia painstakingly added. -- Heron 09:07, 6 Jul 2004 (UTC)

This site is awesome. That was my first post, and I certainly didn't expect responses!

Please let me share my perspective. A typical person has no idea how a digital circuit does the things it does, but a curious and intelligent person, when presented with the opportunity, would be eager to find out.

I believe an essential feature of a logic gate is that it is cascadable--the output of the logic gate is able to serve as an input to another logic gate. When a reader discovers this, and when he sees how switches/transistors can perform logic operations such as AND, his computer is no longer magical--it is merely complex. If a logic gate is treated as a device that obstructs or allows the flow of electricity from point A to point B, the idea of cascadability is lost, or at least postponed.

I understand the difficulties of treating this accurately. Many readers will not be familiar with concepts like voltage, current, and resistance. But I do think that an ideal writeup must convey the idea of cascadability somehow. I think I will concede that the logic gate currently in the writeup is pedagogically acceptable even without a pulldown resistor--technicalities could be covered in writeups on digital logic schemes (e.g. CMOS). My fundamental complaints are that the inputs and outputs are undefined and that it is unclear how the output will be used at the next logic stage.

A good point. Switches and relays are cascadable though, so it could easily be conveyed by that, with a note to the effect that it is important and applicable to digital solid state circuits too. By the way, great job Heron! Graham 12:56, 6 Jul 2004 (UTC)

Thanks :-) I added a note on cascading semiconductor logic, but I haven't done the same job for relays yet. I have a feeling that this article is a sleeping giant, with the potential to turn into a "How Computers Work: from Babbage to Qubits" essay. If it starts to get out of hand, we may need to split it up. -- Heron 13:57, 6 Jul 2004 (UTC)

I think that should be resisted. The article is simply about the Logic Gate as a functional concept - it actually isn't about how they are implemented by particular technologies. There are better places for much greater expansion of that, which can link back here for those that need to know the fundamental principles. I see this as a foundation article - interested readers can follow links to other places to get more detail, or follow a link back here if they were overwhelmed by to much technicality. Graham 23:40, 6 Jul 2004 (UTC)

I agree. I was stating a fear, not an intention. It would be good, though, to have a list of all the implementations we can think of, with links to the relevant articles. -- Heron 08:28, 7 Jul 2004 (UTC)

I agree. The concept of "cascadable" logic is key. And logic gates do *not*, in general, simply pass or not-pass a signal. Someone please make a nice picture that reflects reality better. (Or else I threaten to inflict my feeble efforts to draw them). --DavidCary 05:57, 15 Jan 2005 (UTC)

tri-state logic

Where is the tri-state article ? This is clearly not the right article to talk about tri-state tristate 3-state logic gates (such as the 74356, the 74HC125 (http://rocky.digikey.com/WebLib/Texas%20Instruments/Web%20data/SN74HC125.pdf), etc.). Is "transmission gates" really the best place ? Where in the encyclopedia is the right place ?

Looks like a separate article doesn't exist yet. I'd suggest tristate logic should be the title. However, note that the term itself is inaccurate - tristate logic is not a three-state logic system at all (i.e. trinary), but just binary logic which can be connected or disconnected to e.g. a common bus line. While this ability is important in practical electronic logic systems, it's not really all that exciting - maybe a two or three line paragraph here would be sufficient. Graham 21:53, 23 Dec 2004 (UTC)

Excellent. If the term is inaccurate, then this is an excellent opportunity for this encyclopedia to teach me a more accurate term by defining the inaccurate term as a redirect pointing to the more accurate term. And the more accurate term would be ... what ? --DavidCary 05:57, 15 Jan 2005 (UTC)

Rightly or wrongly, it has always been called '3-state' or 'three-state' logic. Note that 'tri-state' is a registered trade-mark of National Semiconductor (as pointed out in this PDF datasheet (http://cache.national.com/ds/CO/COP8SAA5.pdf), for example, and mentioned in this comment on a book by Sid Katzen (http://www.engj.ulst.ac.uk/sidk/quintessential/chapters/ch2_8.htm)), so we should probably avoid using this as an article title. Everybody else (Texas Instruments, Fairchild, Philips etc.) calls it '3-state'. In the field of multivalued logic, the term '3-value' is used to describe trinary logic. [1] (http://www.multivaluelogic.com/) --Heron 12:21, 15 Jan 2005 (UTC)

diagrams

Am I seeing things or are the 'alternate AND' and 'alternate OR' labels switched on the diagram? For example: De Morgan's law says that P AND Q = NOT ((NOT P) OR (NOT Q)), i.e. there is an OR gate in the 'alternate AND' gate and vice versa. Also, following the circuits visually, it looks like they are switched. Varuna 22:22, 2005 Mar 10 (UTC)

Aha! I'm not the only one to think that. I'm going to switch the captions. --Wtshymanski 16:38, 12 Apr 2005 (UTC)

Logic gates and quantum physics

I know that from several citations I've seen, that it is known that an "AND" gate will produce more heat than a "NOT" gate, because under quantum theory, information is destroyed and is released as heat (information is a very weird property in quantum theory), whereas an "NOT" gate merely converts information. Would it be good to try to elaborate some of this issue here? -- Natalinasmpf 03:35, 19 Apr 2005 (UTC)

My opinion is not to clutter the article up with deeply theoretical and abstract ideas - put it in an article linked to quantum theory but I think it has no place in a discussion of real-world logic. I'd like to see the cites - can you provide a link that explains this in terms accessible to a non-physicist? Since anything you can do with an OR gate you can do with enough ANDs and inverters, and the reverse, I don't really see the fundmental reason for this observation. --Wtshymanski 19:45, 19 Apr 2005 (UTC)

Whoops, I meant a "NOT" gate, not an "OR" gate,. But logic gates in essence deal with logic itself - I think perhaps some mention about quantum theory (which is real-world, after all) because it entails what the nature of logic really is. And you may accomplish a function with a different setup, but the question in physics becomes, is it efficient? The theory goes that quantum logic gates must eliminate the wasted heat (as information is destroyed in an AND gate: two inputs produce an output where you cannot reverse back to deduce the input), else it would create inteference which is more of an issue for a quantum processor. Oh, I found this: [2] (http://www.themilkyway.com/quantum/FinalReport/QuantumGates.html), it cites the heat produced from destroyed information. And this too [3] (http://beige.ucs.indiana.edu/M743-talk-2/node1.html), this [4] (http://www.optics.rochester.edu:8080/users/stroud/presentations/muthukrishnan991/LogicGates.pdf), and this [5] (http://patmos2001.eivd.ch/program/Repro%5CArt_9_2.pdf). All these cite the production of heat due to the fact information is destroyed (heat being a byproduct of electrical resistance being a secondary product, heat produced by information destruction is way more). (Oh some of the proposed improvised quantum gates to make up for the elimination of such gates are also trinary gates). -- Natalinasmpf 23:39, 19 Apr 2005 (UTC)

The real history

Somewhere, probably not on a Web page, there's got to be some real history of the logic gate. The Tesla and Stowger patents referred to in previous versions of this article do not claim anything I can see called a "logic gate" or a functional equal. A mechanical logical-AND function must go back at least as far as the first lock with a key in it. I don't think you can patent the idea of logic alone, but I'd be happier if someone cited a patent that had as a major claim the idea that logical functions can be performed automatically. It may not be patentable at all. Babbage's difference engine predates even Tesla. --Wtshymanski 22:05, 19 May 2005 (UTC)

Somewhere? It was in the reference links! BTW, Andy Kessler and Leland Anderson both cite Tesla as the original. Twenty First Century Books publishers is a lesser link (but another one). If you cannot understand the patents nor the links ... do not remove the information just because you don't like the information. PLEASE provide a link or a book to refute these sources!.

PS, put in Babbage's machine if you feel like it ... I think it may be a good addition.:

The cited patents are for a signalling system, not for logic gates. None of the claims mention an electronic implementation of a logic function, possibly because Tesla would have considered this of no novelty and unpatentable. Surely the first nameless tinkerer who wired two knife-switches in series has as much claim to "inventing" an AND gate. I don't see the relevance of the cited patents to the claim that Tesla invented the AND gate. The "Twenty First Century" reference seems to be highly partisan in advocacy for Tesla and does not explain the relevance of the cited patents. I dispute the accuracy of the claim. My reference is the cited patents, which do not claim an AND gate. --Wtshymanski 00:42, 21 May 2005 (UTC)