Which is better?

[compilerbitch]

This:



or this:



In both cases, the unfamiliar symbol represents a delay. I use a Greek delta (Δ) to represent transmission line delay, and a square to represent inertial delay in equations, so I want matching circuit symbols. There aren't really any nice standard symbols, other than some quite yucky ones from the async literature that I don't really want to use for aesthetic reasons. I won't say which one I prefer, because I'd rather hear an unbiassed opinion from you lot.

[Poll #417272]

Comments

[pseudomonas.livejournal.com]

The bare triangles might look better if they were shifted up a bit, so that instead of the distance being the same above and below the line on which they are impaled, the area was the same. Also if the delta-ness could be emphasized by thickening one of the diagonals, so it looks more like a letter in a book face and less like an equilateral triangle.

Yes!

[aiwendel.livejournal.com]

ahhh maybe thats it... if it was written as a delta in fancy font or even italicised then it would be less random triangly and thus more sensicle....
:)

[aiwendel.livejournal.com]

the triangle in the box looks more like a delta, time unit type thingy, whilst the triangle alone just looks like a triangle and nothing more, and is alittle confusing as it looks like your other triangle which i've forotton the meaning of...
Anyway, the triangle alone is prettier, but makes me think "gate" rather than "time delay", ergo I'd go for the boxed one, or preferably a prettified version of the box one, maybe in a sharpened box?

but then in most things I go for function over appearance, though both is best...

xxxxxxxxxx

[pseudomonas.livejournal.com]

Just a thought - could you use, instead of a delta, a stylised clockface/hourglass symbol? It might be more immediate.

[vampwillow]

Basic problem of the upper one is clarity of thought .. there isn't any.

With the lower example it is clear that the items concerned are part of the system being graphed. They clearly have an input and an output.

In the case of the unadorned triagle (in whatever position) it isn't clear that they are part of the system indeed, if I saw that graphic in a book I'd assume that the registration was off during printing and that the triables indicated testor measurement points. They stand separate from the system shown and have no clear IO or explanation of what they might be (and a 30deg rotation would give them a totally different meaning too, which the person looking at the graphic might also think you had meant).

Takes be back to college in the mid-70's seeing that ;-)

2p'th ...

[hairyears.livejournal.com]

What she said.

[countess-sophia.livejournal.com]

The version at the top is clearer and easier to read, at least I think.

[davefish.livejournal.com]

When you say delay line are you talking about the actual analogue time it takes for the signal to propagate, or a digital delay, where the answer comes out the far side after a given number of cycles a-la shift register.

If its analogue then the first (Which looks like a component in the diagram), and if digital then the second, which you could imagine as some sort of black box system with the given shift-register property. Having the rounded rectangle seems to shout digital filter at me, which is why I am suggesting this way.

[davefish.livejournal.com]

And curses to polls that don't let you see the results when you haven't completed them.

[ewx.livejournal.com]

err, like this?

[toothycat]

My response on seeing the diagram before actually reading the post was, "The grey shaded triangles are NOT gates. Don't know what the difference to the thing using the standard NOT gate symbol is, don't know what the things in boxes are, the text will probably say." I would therefore argue for the deltas in boxes since they seem less likely to cause confusion as they bear less resemblance to anything familiar, so the reader *knows* they represent something new. This is generalising from one sample though :) Does that make sense?

[kaet.livejournal.com]

I think that people who prefer wiggly resistors and circleless transistors (all you VLSI types!) will prefer the upper, and people who prefer box resistors and encircled transistors (all you PCB types) will prefer the lower. But, I think that the upper looks too much like some kind of open-circuit interruptor type thing. Also, I'm assuming that you want to put these symbols in to represent lumped models, (like the lumped models of noise and transmission lines and stuff), so if they're lumped, they should be inside a lump!

[randwolf.livejournal.com]

Mmmmm...I think it needs a sense of length & direction & I don't think either graphic does that.

Maybe something like:

	         . .          . .          . .        
	       .    .       .    .       .    .       
   -----------.      .     .      .     .      .     .-----------
	              .   .        .   .        .   . 
	               . .          . .          . .  

Not enough direction--how 'bout:

	     ()-------------------------)      	       	
	   (    )             		  )       	
----------(--o   )			   )------------
	   (    )      	       	       	  ) 
	     ()-------------------------)   

I'm not terribly fond of that one either, hmmm.

[mageboltrat.livejournal.com]

The problem with electronics is there is in icon for big red button with "DO NOT PRESS" on it.... there isn't one for Armageddon device either... *pout*

[beckyc.livejournal.com]

The one in the rectangle looks more like the concept of a component. Which may be a good thing or a bad thing. But it could be an abstract concept of a component representing something that causes a delay, no matter what the implementation of it would be.

[feyeleanor.livejournal.com]

Yes, putting the delta inside the box definitely gives the impression of a component of which the delta is a function, whereas just putting the deltas in without the bounding boxes makes it look like the delta is an actual component.

I'm sure I remember seeing a triangle used that way in a course on control systems but that was in 1990 and I can't remember anything useful because the lecturer had the stupidest way of talking about diagrams whilst facing away from the OHP so he'd nearly always be pointing to the wrong components...

Anyway, I also liked the idea expressed earlier of making the deltas look a little more delta-like than bare triangles. I'd be tempted to go one step further and explicitly use delta-t notation, but that may be a little to much like physics and not enough like EE.

[lsur.livejournal.com]

The triangle in a box looks sufficiently different from a 'normal' digital symbol and kind of suggests that something is going on that needs to be taken into account.

Coincidentally, I mark difficult passages in my music books with a triangle. They also represent a dely as I try to get my fingers to the right fret. But that's something else.

[olithered.livejournal.com]

The triangles look like roadwork signs.

[zwol.livejournal.com]

I have, like, no EE experience whatsoever, so I'm not sure what people's expectations of these diagrams are ... but I *cannot* read your box-with-triangle as a single glyph, and this makes it much harder for me to understand that version of the diagram. (I have this problem with any diagram containing boxes with non-text symbols inside, especially if the non-text symbol is nearly as big as the box.)

[owdbetts.livejournal.com]

Well, since I've just figured out who you are, I may as well comment on an age old post. Or something.

I prefer the traingle in a box because as something in a box looks like it's a box (abitrary thing) labelled with a triangle (local convention) whereas the triangle as a circuit symbol looks wrong, because it isn't a standard one (that I'm familliar with, at least).

What did you decide on?

[compilerbitch.livejournal.com]

The triangle in a box. Hopefully the journal reviewers will like it too! :-)