Schematics & standards
in [Repair]
|
Prev: Sound Technology ST-1700B distortion analyzer measurement pegs meter on low range.
Next: Simple hack to get $1500 to your home.
From: Adrian Tuddenham on 19 Jun 2010 17:50 David Nebenzahl <nobody(a)but.us.chickens> wrote: > On 6/19/2010 1:19 PM Adrian Tuddenham spake thus: > > > David Nebenzahl <nobody(a)but.us.chickens> wrote: > > > >> (And I don't much like their transistor symbols either) > > > > For junction transistors they are incorrect, I agree, but I have become > > used to them. I find it takes me a while to get my mind around the > > correct symbols because they are so rarely used nowadays > > > > Just for fun, I've replaced the point-contact symbols in that drawing > > with the correct junction ones: > > > > http://www.poppyrecords.co.uk/compton/images/BassAmplifier2.gif > > > > The wrong symbols have become so well-estabilshed nowadays that I doubt > > if most people even noticed they were wrong. > > Now that's just plain *weird*. I thought you would find them interesting. > Since when are the *conventional* symbols for (junction) transistors > considered to be for the old, obsolete point-contact ones? When junction transistors were first introduced there was a need for a new symbol to distinguish them from the point-contact type which the 'conventional' symbol represented. Several eminent journals and text books changed over to the new 'junction' symbol, but, by then, the point-contact symbol was so well established that the change never caught on. >Every single > schematic that uses transistors--modern silicon ones, not ancient > point-contact germanium ones--uses the conventional symbols, like the > ones in the first drawing you posted. You will find the 'junction' symbols in some Acoustical Quad circuit diagrams, Peter Walker was a stickler for getting things right. They also appeared in Wireless World for a while and are used in "The Foundations of Wireless" by M.G. Scroggie (8th Edition) specifically to distinguish the two different types of transistor. > > I've *never* seen symbols like the ones in your "new, improved" drawing. > Those are just plain idiosyncratic, non-standard and weird. They look > kind of like diodes with an elongated anode. They certainly look strange when you have been used to the point-contact symbol, but you must admit they give a clear representation of a junction transistor. > > I'll stick with the tried and true standard symbols, thank you very much. At least you will be able to recognise the other types if you ever encounter them again. -- ~ Adrian Tuddenham ~ (Remove the ".invalid"s and add ".co.uk" to reply) www.poppyrecords.co.uk
From: PlainBill47 on 19 Jun 2010 18:29 On Fri, 18 Jun 2010 13:18:41 -0700, David Nebenzahl <nobody(a)but.us.chickens> wrote: >Someone else made a comment in another thread here about weird >schematics (like for home appliances). > >Wanted to get a small discussion going on that topic. My take: there are >good and bad standards for schematics. Personally, I can't stand the >ones that use rectangle shapes for resistors, instead of the traditional >zigzag that [insert name of deity here] intended to be used. (And even >here there are lots of variations, like old-fashioned schematics that >took this symbol rather literally and sometimes had ten or twelve zigs >and zags, as if an actual resistor was being constructed on paper). > >Likewise the wire-connecting/jumping convention: here I much prefer the >modern approach, which is to use a dot for a connection and no dot for >no connection, rather than the clumsy "loop" to indicate one wire >jumping over another with no connection. > >Regarding resistor values: Who the hell came up with that new way of >specifying resistance values, like "10R" "or 5K6" or whatever? And why >use this system? I've always used the plain value of the resistance: 10, >56, 5.6K, 56K, etc. Simple, obvious, requires no interpretation. Is this >some kind of Euro thing? > >In general, some schematics just look and feel nicer than others. A >well-drawn schematic is a pleasure to read. A bad one--lines too thin or >too thick, misshapen symbols, idiosyncratic interpretations, etc., just >don't look right. > >Feel free to add your own schematic pet peeves here. Well, the worst schematics of all are those which you cannot find. Even the most miserable scratching on a crumpled piece of paper is better. While I learned using the 3.3K style, I fiend the 3K3 eminently satisfactory, especially because of the redundancy. As was mentioned,, this is important when dealing with a PDF of a poorly scanned, poorly printed original. I prefer the 'old' style - zig-zag lines for resistors, parallel lines for non-polar capacitors, etc. Lines should be drawn with the little loop when crossing lines do not connect, a dot when they do. Again, redundancy. Tags indicating the signal connecting to an IC should have an arrow indicating if the signal is an input or an output, double arrows for a bidirectional bus. And when a signal goes off the page, the description should be accompanied by the page and grid location of the destination, as in < SYNC 3E5 indicating the SYNC signal is coming from page 3, grid location E5. As a bonus, the location of each component should be tabulated, either on the schematic, or in a separate chart so it is possible to determine that IC205 is on the bottom side of the circuit board at grid location J12. PlainBill
From: Rich Webb on 19 Jun 2010 20:10 On Sat, 19 Jun 2010 15:29:45 -0700, PlainBill47(a)yahoo.com wrote: >Lines should be drawn with the little loop when crossing lines do not >connect, a dot when they do. Again, redundancy. Crossings (four-way intersections) never connect. Three-way intersections always connect. Stick with that convention and neither the humpie or a dot are needed, although dots do "look right." There is an authorized reprint of H&H's "How to Draw Schematic Diagrams" from AoE Appx E over at http://opencircuitdesign.com/xcircuit/goodschem/goodschem.html -- Rich Webb Norfolk, VA
From: Arfa Daily on 19 Jun 2010 20:42 "David Nebenzahl" <nobody(a)but.us.chickens> wrote in message news:4c1d00fe$0$2388$822641b3(a)news.adtechcomputers.com... > On 6/19/2010 8:35 AM Adrian Tuddenham spake thus: > >> David Nebenzahl <nobody(a)but.us.chickens> wrote: >> >>> Someone else made a comment in another thread here about weird >>> schematics (like for home appliances). >>> >> [...] >>>... wire-connecting/jumping convention: here I much prefer the modern >>>approach, which is to use a dot for a connection and no dot for >>> no connection, rather than the clumsy "loop" to indicate one wire >>> jumping over another with no connection. >> >> I find the 'gap' convention is easy to draw (with a computer) and >> extremely easy to read. It also looks tidy. Four-way junctions which >> could be mistaken for crossings should never be used, they should be >> staggered instead. >> >> e.g. >> http://www.poppyrecords.co.uk/compton/images/BassAmplifier.gif > > BZZZZZZT! Fail. > > While the gap thing looks OK for non-crossing wires, I have to ding the > drafts-person of that schematic for the following: > > o Idiosyncratic symbols for electrolytic cazapitors[1] > o Idiosyncratic ground symbol (one horizontal line????) > o And no, I disagree about those offsets for connecting wires. > > That's totally unnecessary here: it would be quite obvious that all those > vertical wires connect to what is obviously a bus or rail. A well-drawn > dot is all that's needed there. > > (And I don't much like their transistor symbols either) > > > [1] With apologies to J. Liebermann. > I must say that I don't really like the staggered connections, but what's wrong with the transistor symbols ? And the single heavy horizontal line for the 0v rail, is very common this side of the pond. 0v rails always used to be shown as a heavy horizontal line right across the schematic, sometimes with a chassis symbol attached as well. These days, most schematics are so complex, that the 0v line is now left out, and 'abbreviated' to individual short heavy lines at each connection point on the schematic. The electrolytic symbol is not, however, the one commonly used here, which is a pair of rectangles, one filled in for the -ve side, and the other open for the +ve side. Sometimes, the American convention of one straight and one curved plate, is used. Arfa
From: Cydrome Leader on 19 Jun 2010 23:53
David Nebenzahl <nobody(a)but.us.chickens> wrote: > Someone else made a comment in another thread here about weird > schematics (like for home appliances). > > Wanted to get a small discussion going on that topic. My take: there are > good and bad standards for schematics. Personally, I can't stand the > ones that use rectangle shapes for resistors, instead of the traditional I find rectangles obnoxious, unless somebody from europe is drawing something in front of me. > zigzag that [insert name of deity here] intended to be used. (And even > here there are lots of variations, like old-fashioned schematics that > took this symbol rather literally and sometimes had ten or twelve zigs > and zags, as if an actual resistor was being constructed on paper). > > Likewise the wire-connecting/jumping convention: here I much prefer the > modern approach, which is to use a dot for a connection and no dot for > no connection, rather than the clumsy "loop" to indicate one wire > jumping over another with no connection. I was taught the half-loop shape first, then moved to the dots and no dots. It seemed like how you're taught to ties shoes in a really complex method of making two rabbit ears first, then tying them. > Regarding resistor values: Who the hell came up with that new way of > specifying resistance values, like "10R" "or 5K6" or whatever? And why > use this system? I've always used the plain value of the resistance: 10, > 56, 5.6K, 56K, etc. Simple, obvious, requires no interpretation. Is this > some kind of Euro thing? I first saw that on this newsgroup. My question is what idiots came up with it and why? |