Standardizing symbols for PCB mirroring pads

Discuss hardware-related topics, such as development cartridges, CopyNES, PowerPak, EPROMs, or whatever.

Moderators: B00daW, Moderators

User avatar
Quietust
Posts: 1565
Joined: Sun Sep 19, 2004 10:59 pm
Contact:

Re: Standardizing symbols for PCB mirroring pads

Post by Quietust » Sun Jul 05, 2020 10:21 am

Whatever the symbols may be, the V symbol ideally should not be identical to the H symbol rotated by 90 degrees.
Quietust, QMT Productions
P.S. If you don't get this note, let me know and I'll write you another.

krzysiobal
Posts: 730
Joined: Sun Jun 12, 2011 12:06 pm
Location: Poland

Re: Standardizing symbols for PCB mirroring pads

Post by krzysiobal » Sun Jul 05, 2020 10:53 am

Whatever you will chose, add one letter to eliminate disambiguity
Attachments
foo2.png
foo2.png (612 Bytes) Viewed 706 times

User avatar
tokumaru
Posts: 11742
Joined: Sat Feb 12, 2005 9:43 pm
Location: Rio de Janeiro - Brazil

Re: Standardizing symbols for PCB mirroring pads

Post by tokumaru » Sun Jul 05, 2020 10:59 am

krzysiobal wrote:
Sun Jul 05, 2020 9:34 am
No, the checkboard means "different", like solid=A, checkboard=B
That's my point... solid vs. checkerboard means one thing in the first 2 icons (memory vs. mirror) but another in the third (memory A vs. memory B). EDIT: I see what you mean now. Solid is A, checkerboard is B. That's consistent in the first 3 icons, which's good. The 4th icon is a little weird though... each screen has the checkered area in a different position to illustrate that they're all unique? I'm not sure that's a good way to represent 4 different screens.
Quietust wrote:
Sun Jul 05, 2020 10:21 am
Whatever the symbols may be, the V symbol ideally should not be identical to the H symbol rotated by 90 degrees.
True! Even though cartridge PCBs have an obvious orientation, it would be better if the icons were recognizable from any perspective.

Fiskbit
Posts: 120
Joined: Sat Nov 18, 2017 9:15 pm

Re: Standardizing symbols for PCB mirroring pads

Post by Fiskbit » Sun Jul 05, 2020 7:43 pm

krzysiobal's latest suggestion is pretty good. Alternatively, if the main issue here is just that mirroring and alignment are ambiguous on H and V solder pads, I feel it's pretty simply and concisely solved by just writing HA and VA.

Pokun
Posts: 1485
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: Standardizing symbols for PCB mirroring pads

Post by Pokun » Thu Jul 16, 2020 9:22 am

I think any letters are very hard to interpret. AB means nothing and using H and V like that doesn't solve the ambiguity problem that is actually addressed. H and V stands for H-scroll and V-scroll on original boards (which simply indicates what axis you can scroll in, in order to get to the other nametable), not "arrangement" so HA and VA are also just confusing use of newly invented terminology.

Krzysiobal's patterns confuses me because I thought dotted pattern meant mirrored VRAM, but it actually makes sense if you want both screens to have different symbols (I'm not sure they do though?). The 4-screen pattern also confuses me, why is there chequered pattern in it?


I prefer Tepples' initial idea, but I also fully agree with Tokumaru and prefer his updated version of Tepples' design. That's actually exactly what I also had in mind, since Tepple's single-screen mirroring image is confusing without the explanation to it.
So I also vote Tokumaru's second picture.

User avatar
Jeroen
Posts: 986
Joined: Tue Jul 03, 2007 1:49 pm

Re: Standardizing symbols for PCB mirroring pads

Post by Jeroen » Fri Jul 17, 2020 11:03 am

Although I like the checkerboard symbols. I have to wonder if they'll be very legible on boards produced by some of the cheaper PCB houses. (Obviously if you make it big enough any board house will be able to do it, but I mean at smaller sizes)

Drag
Posts: 1314
Joined: Mon Sep 27, 2004 2:57 pm
Contact:

Re: Standardizing symbols for PCB mirroring pads

Post by Drag » Fri Jul 17, 2020 2:42 pm

How about this? The dot indicates nametable 1, much like in engineering where it shows where pin 1 on an IC is.

Edit: These are non-graphical, so at any line thickness:
Vertical Mirroring: Two horizontal lines, dot in upper left corner.
Horizontal Mirroring: Two vertical lines, dot in upper left corner.
Single-screen: Four dots, slashes, squares, or other markings arranged in a square formation.
Four-screen: A square.

Edit 2: I confused myself too. :P Got h and v swapped.
Attachments
mirroring.png
mirroring.png (295 Bytes) Viewed 469 times

tepples
Posts: 22014
Joined: Sun Sep 19, 2004 11:12 pm
Location: NE Indiana, USA (NTSC)
Contact:

Re: Standardizing symbols for PCB mirroring pads

Post by tepples » Fri Jul 17, 2020 7:16 pm

I'm liking the dot to disambiguate rotations of the symbol. Here's what the dot looks like combined with the symbols that tokumaru and I suggested.
mirroring_symbols_with_dot.png
mirroring_symbols_with_dot.png (294 Bytes) Viewed 447 times
But then each symbol becomes 17x15 pixels, losing the ability to fit each into a 16x16 pixel box. I see two solutions: either shrink each screen within the symbol...
mirroring_symbols_smaller.png
mirroring_symbols_smaller.png (282 Bytes) Viewed 447 times
...or make the dot bigger and overlapping the symbol's top left corner.
mirroring_symbols_thick_tlc.png
mirroring_symbols_thick_tlc.png (302 Bytes) Viewed 447 times

lidnariq
Posts: 9491
Joined: Sun Apr 13, 2008 11:12 am
Location: Seattle

Re: Standardizing symbols for PCB mirroring pads

Post by lidnariq » Fri Jul 17, 2020 7:52 pm

I'm still having this problem with the symbology that I don't see any of the symbols as unambiguously "shade is a copy of solid" vs "shade is the other option from solid"

User avatar
Jeroen
Posts: 986
Joined: Tue Jul 03, 2007 1:49 pm

Re: Standardizing symbols for PCB mirroring pads

Post by Jeroen » Sat Jul 18, 2020 4:21 pm

Drag wrote:
Fri Jul 17, 2020 2:42 pm
How about this? The dot indicates nametable 1, much like in engineering where it shows where pin 1 on an IC is.

Edit: These are non-graphical, so at any line thickness:
Vertical Mirroring: Two horizontal lines, dot in upper left corner.
Horizontal Mirroring: Two vertical lines, dot in upper left corner.
Single-screen: Four dots, slashes, squares, or other markings arranged in a square formation.
Four-screen: A square.

Edit 2: I confused myself too. :P Got h and v swapped.
I like this, should avoid any major printing issues.

Pokun
Posts: 1485
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: Standardizing symbols for PCB mirroring pads

Post by Pokun » Sat Jul 25, 2020 10:54 am

lidnariq wrote:
Fri Jul 17, 2020 7:52 pm
I'm still having this problem with the symbology that I don't see any of the symbols as unambiguously "shade is a copy of solid" vs "shade is the other option from solid"
Yeah none of the symbols presented so far are totally unambiguous and universally understandable without any insight.

rox_midge
Posts: 84
Joined: Mon Sep 19, 2005 11:51 am

Re: Standardizing symbols for PCB mirroring pads

Post by rox_midge » Sat Jul 25, 2020 8:18 pm

I have no horse in this race, but:
tepples wrote:
Fri Jul 17, 2020 7:16 pm
But then each symbol becomes 17x15 pixels, losing the ability to fit each into a 16x16 pixel box.
Given that these icons will be used in meatspace, whether or not the design fits into a pixel grid should be a distant secondary concern after making sure that's it's unambiguous and readable when silkscreened onto a board next to some pads that will be soldered together.
Pokun wrote:
Sat Jul 25, 2020 10:54 am
lidnariq wrote:
Fri Jul 17, 2020 7:52 pm
I'm still having this problem with the symbology that I don't see any of the symbols as unambiguously "shade is a copy of solid" vs "shade is the other option from solid"
Yeah none of the symbols presented so far are totally unambiguous and universally understandable without any insight.
There's lots of symbols that aren't universally understood without any additional context, like the power symbol for PCs. That doesn't mean they're bad symbols, it just means that you need to know what you're looking for.

In this case, the pattern tables of the PPU are most commonly visualized in a 2x2 ordering, which is why most of the options so far have four boxes next to each other. My opinion is that that 2x2 grid needs to prominently feature in the resulting design, because it already evokes the PPU for people who know the NES platform.

The names of the four options are fairly standard already - horizontal mirroring, vertical mirroring, single-screen mirroring, and four-screen "mirroring". My vote would be to include the letters to make the options completely unambiguous, either in the center or on a consistent side (above, below). Including the letter will also help to orient the symbol.

I think just adding numbers to the symbols might be enough, maybe with the H/V/1/4 off to the side. Please excuse my rushed hatchet job on your work, Tepples:
stuff.png
stuff.png (1.22 KiB) Viewed 278 times

User avatar
tokumaru
Posts: 11742
Joined: Sat Feb 12, 2005 9:43 pm
Location: Rio de Janeiro - Brazil

Re: Standardizing symbols for PCB mirroring pads

Post by tokumaru » Sat Jul 25, 2020 8:40 pm

Going against my own original suggestion, I'm not sure that drawing mirrors in dotted lines is the best option. When memory is mirrored, there's nothing saying which instance is the original and which are repeats: the same set of rules make the same memory visible in multiple places, none of the slots is more "legit" than the others.

For this reason, I vote for us to find symbols that represent different chunks of actual memory, and that they be drawn without modification everywhere that the same physical memory is mapped. Like krzysiobal suggested here.

lidnariq
Posts: 9491
Joined: Sun Apr 13, 2008 11:12 am
Location: Seattle

Re: Standardizing symbols for PCB mirroring pads

Post by lidnariq » Sat Jul 25, 2020 9:53 pm

rox_midge wrote:
Sat Jul 25, 2020 8:18 pm
There's lots of symbols that aren't universally understood without any additional context, like the power symbol for PCs. That doesn't mean they're bad symbols, it just means that you need to know what you're looking for.
But that's exactly the problem. We already have symbols "H" and "V", and they're already confused because there were two different ways of describing the nametable layout. Any representation that isn't obvious whether we're talking about "layout=arrangement" or "duplication=mirroring" fails to solve the problem.

OSHpark says they can reliably deal with roughly 400dpi, and ideally the symbol would be tiny - maybe 6mm square - so we probably have much more than 16 pixels square. (Maybe 100?). But maybe it'd be better to aim for 100dpi and 32px.

I might go for some sort of visual representation with a symbol that represents the CRT face, perhaps something like this bit of crappy programmer art:
vertical-scrolling.png
vertical-scrolling.png (147 Bytes) Viewed 258 times
horizontal-scrolling.png
horizontal-scrolling.png (150 Bytes) Viewed 258 times
There are three thoughts I have in these:
1- Rounded sides and 4:3 aspect imply screen
2- Rounded sides and 4:3 aspect give orientation clue
3- 12:5 or 6:10 aspect ratio of nametables show nametable orientation

On a physical PCB, we're just not likely to have an opportunity for the 4-screen layout symbol, so we can't rely on its presence to disambiguate the other symbols. Most of the time the PCB will be either "most" or "all" discrete logic mappers, so whatever we want to use should be clear from just the A10 (vertical mirroring, horizontal layout) and A11 (horizontal mirroring, vertical layout) symbols. Making a symbol for 4-screen layout is a nice way to finish the exercise, but the only place it'd warrant a solder pad is in contrast to mapper-controlled nametables, not in contrast to fixed 2-screen layouts.

rox_midge
Posts: 84
Joined: Mon Sep 19, 2005 11:51 am

Re: Standardizing symbols for PCB mirroring pads

Post by rox_midge » Sun Jul 26, 2020 7:48 am

tokumaru wrote:
Sat Jul 25, 2020 8:40 pm
Going against my own original suggestion, I'm not sure that drawing mirrors in dotted lines is the best option. When memory is mirrored, there's nothing saying which instance is the original and which are repeats: the same set of rules make the same memory visible in multiple places, none of the slots is more "legit" than the others.
But remember that we're not just talking about what's true from the machine's perspective, we're also talking about human perception. In this case, the dotted lines to me indicate that you've already seen this symbol once.

Here's a refined version of my suggestion, which uses a lighter line for the mirrored regions, keeps the cutouts to indicate continuity between sections, and adds human-readable text underneath each symbol:

mirroring-symbols.png
Attachments
mirroring-symbols.zip
(4.12 KiB) Downloaded 6 times

Post Reply