Junction Dots Revisited


#21

USA Standard Drafting Practices .pdf

Pages 11, and 21-27, contain the relevant content.


#22

That one is hardly recent…
And most of the examples show dots.
Anyone have access to ISO standard “01.100.25 - Electrical and electronics engineering drawings” ?
https://www.iso.org/ics/01.100.25/x/


#23

ohhhhhh boy.

First off, that document is as old as I am.

Second, I used to work for the ETS group at NOAO in downtown Tucson (on the University of Arizona campus), and at the time, we used Accel for schematic capture and layout, and that schematic capture required junction dots to indicate wire connections.

Third, that document does have a lot of anachronisms. For example, it wants to have capacitance shown in picofarads and microfarads, which are fine, except that nanofarards are a better choice for many common values. (100 nF or 0.1 uF? I vote: don’t use that decimal point, as it might vanish on reproduction.)

Further in that document, say on page 23, I see a capacitor with value “4700.” 4700 F? That’s a big cap. See what I mean about being explicit? The reader, 51 years later, shouldn’t have to guess what was meant by something on the schematic.

I remind you that you originally said, [quote=“Sprig, post:1, topic:9027”]
Modern schematics have not used junction dots for quite some time now.
[/quote] and you use a 50-year-old standard to illustrate your point. Maybe by “modern” you mean the Mid-Century Modern architecture and decoration style that’s so popular around here?

All of that is, of course, completely beside the point, which was made earlier: software cannot be ambiguous about this. It wants the user to make intentions clear so it does what is expected.

In fact, in the EAGLE document to which you link, “Pro Tip #1” is “Clearly Show How Your Wires Connect,” where we read:

One can complain about many of Kicad’s usability quirks, but its use of the industry-standard dot to indicate wire junctions is not one of them.


#24

@Andy_P, any chance you can tone down the personal attacks towards me? It is unlikely that you have been involved in my very specific line of work to see what I have seen. I provided a link to you, while old, does contain several pages of schematics drawn without junction dots to US standards. It is in no way my fault that you have never seen such drawing style before. Due to NDA requirements I will not provide the documents to prove that this style is still in current daily use in certain industry elements.


#25

I would hope engineers are familiar with batch sampling, and required sample sizes to get a meaningful result. I guess there are hundreds of thousands of companies creating schematics, extrapolating from a sample of a handful to industry wide practice is simply absurd. You barely have a big enough sample size to establish a coincidence.

Reference documents are largely irrelevant to industry practice, in such discretionary areas as appearance. IME when an engineer talks about “best practice” they usually mean “local practice”. Since we are trading anecdotal evidence, I can say I never seen a policy of not using junction dots in all the companies I’ve worked for.

Obviously personal attacks should be avoided, but you do seem to have some ideas that are 180 degrees different to my experience, I think this is one of them. This is also one of the threads heading in a futile and pointless direction that I will not be returning to.


#26

I would argue that as ECAD now adds the dot automatically, unlike the old manual days of the extra operation with the drawing stencil and pen, that the dot makes more sense than ever.


#27

By “newer convention” you are referring to the 1966 document? There is obviously nothing “new” about any of this. The document you posted makes it quite clear, junction dots are optional. It goes on to show various connections both with and without dots. But nowhere does it say never use dots!

The only thing to have changed over the decades in this regard is that once upon a time crossing wires were indicated with one looping over the other, that is now discouraged in favor of lines crossing with no dot. Placing a dot on connected crossing lines is perfectly acceptable as is having two connections slightly offset. A matrix of wires with connections, such as a bus, is far easier to read when drawn as a matrix with a dot placed at the connections.

In addition, the point of these methods of indicating connections is clarity, not eliminating dots. Even when connections are drawn as offset ‘T’ connections dots are still optional.


#28

This again is a personal attack; these are not my ideas.

These are my experiences, sames as yours.

I have provided, to the best of my ability, that my experiences are based upon factual information in the link I could find that did not violate my NDA.


#29

Seriously Jim? That is far from a personal attack, he simply stated his experience is opposite to yours, however, unlike you, he did not try to promote his experience as being standard or conventional.

No, but you are the one promoting them here today as if they are something new, and a standard we all should not only adhere to but write ECAD software to enforce.


#30

Reaffirmed 1973; little bit newer. This is what I could find to share without having to violate my NDA.

Some pages mentioned show the non-junction-dot as US drafting standard even prior to 1973.

I’ll admit that I can’t currently recall how a matrix was drawn without junction dots. I don’t think that sort of matrix was in any product I worked.


#31

Some pages show non-wire lines too, what is your point? We are talking about dots used to indicate connections.


#32

Yes, and probably reaffirmed many times since then, but that doesn’t make it any newer. :wink:

I’d also like to point out some text on the cover of that document,

PROPOSED USA STANDARD FOR CONNECTION DIAGRAMS

So as of 1973 at least, is was only proposed as “standard”.

Keep in mind these “standards” came about long before ECAD software was being used to draw schematics.


#33

I can still read all of it without question.


#34

Maybe.

Maybe NOT.

I have experience to suggest otherwise.


#35

Calm down everybody. There is no definitively right or wrong way.
Ideally KiCAD would support both options to suit end user requirements


#36

Personally I would never choose a non-explicit connection over an explicit one.
It would be kind of nice if crossing non connected wired would make the jump like in altium… but removing the dot would have never crossed my mind.

Imagine our rage if we need to work with that schematic and we end with several nets disconnected because the designer chose a style that none of us even heard of.


#37

Other people need bigger junction dots. Another standard. I prefer explicit dots for T and cross connections.


#38


#39

There have been no personal attacks in this thread. If you consider a disagreement, especially one which rebuts points made, to be a personal attack, then I am done.

And on that note, gang, I bow out of this discussion and the forum.

Good night and good luck


#40

Yes, connectivity IS implied. But schematic diagrams are intended to communicate intentions to other PEOPLE, not to a set of perfectly-implemented algorithms in some machine or software program. Long before my eyeballs were fitted with trifocal glasses, my mind tended to imagine that power rail connected to the “A5” signal (Pin 1). In my mind, the (admittedly) redundant connection dot creates some kind of psychological barrier to imagining such a connection - and therefore increases the likelihood that the original designer will accurately communicate his intentions to me.

Dale