Pygmalion

Correction: \textup should be used for subscripts.

This also shows that David Carlisle advised me wrongly. \text should be used for scripts and not \textrm as the body font should be translated to subscripts.

​I must also confirm that "roman type" is often synonymous with "upright type". I looked in ISO 80000, in the section on subscripts, and although it talks about "roman (upright) type", the example clearly shows a sans type, the same one used for body font.

Oh, I did not know that you use math font for units. Now I understand why degree symbol must be in \text.

Or I could argue that ° is inherently a unit for technical/scientific (non-mathematical) situations. I hope you don't mind me bothering you with these details.

I understand the part about manually coding units. I just don't understand why ^\circ is semantically more appropriate. In the case of °C, the degree symbol is part of the unit and units should not be in mathematical mode. However, the problem arises with the pure degree symbol °. Is it maths or a unit? Of course, I could argue that the purely mathematical description of angles is in radians, so degrees already fall into the realm of arbitrary human units.

second better=>worse in the penultimate comment

(TBH I seriously doubt that any reader will ever notice the difference, so I must be crazy to lose so much time to make it consistent.)

So essentially you are saying {}^{\circ} is typographically better, but it does not work well for scaling. So I must choose between typographically better solution that scales badly or typographically better solution that scales well. Correct?

OK, now I have done a test and I see that \degree also uses {}^{\circ}. So if I want to be consistent, I should follow your new setting. BUT then I have a scaling problem. Is there a solution that is both consistent with \degree and has no scaling problem?

As you've probably noticed, I'm very careful about styling, but honestly, I don't understand half of what you're saying. I'll stick with \textrm instead of \text and \textdegree instead of {}^{\circ}, in the hope that I'm not committing some grave sin against typography. Thanks!

Great! With this redefinition, everything is fine and back to the way it was before. Why was the default setting changed to {}^{\circ}? Also, I was told by other smarter people on TeX.SE that I should never use \text, but \textrm instead. Why do you use \text instead?

I'm sorry, but I don't understand what exactly I'm supposed to do. Also, why didn't this problem exist in the previous version of siunitx? Can I set siunitx to behave like before?

@JohnRennie I do not take it personally, but I hate administrative rigour, which I think is the biggest obstacle to progress. I have already made up my mind on the question, so there is no need for discussion here any more. The only loss is for all the people who will not be able to see the question.

@naturallyInconsistent I apologise to anyone who wants to help me but can no longer see my question. Enough members of the SE community felt that this important question should be killed because of some petty SE rules, so let it be dead.

@naturallyInconsistent I deleted it out of spite and not to improve my numerical reputation, which I care little about.

@naturallyInconsistent The question was closed, so I deleted it.

Well bye everyone

Rigor cleans the window through which intuition shines.

@Mad OK, so why isn't the same argument applied to cross section or ground plan? They are also not inherit to the object?

@Mad Isn't that the same I wrote?

I am aware this is looks to all of you too pedantic, but I really put enormous effort into my books and want to get every detail right. Maybe I just have to accept the defeat in this detail...

As someone who has been writing books on physics for years, I find this question legitimate and interesting, but I do not know who to address it to. Do you have any ideas?

Because nobody seems to be present at the moment, when you have time, please visit the question I need answer to here: physics.stackexchange.com/questions/771797/…

Hello fellow physicist. I have a multidisciplinary question on physics and English language and I would ask you to advise me where to post it.

@samcarter Thanks!

@JosephWright I found few sites that discuss font design, but no site that discuss page design.

I will proceed with my second question: I have a few questions regarding best page design, which might not be directly connected to LaTeX, therefore possibly inappropriate for tex.SE. Is there any other forum on web where questions regarding the page design, partially in relation to LaTeX could be posted?

So you say as soon as I edit an "inappropriate" question, it gets on a queue and those who have the right to reopen it will see it?

BTW, how do you get it into the queue?

I already edited it: tex.stackexchange.com/questions/689634/vertical-spacing-in-koma

Hello, I have one question. It was suggested to me to edit my question, but before I was able to do that the question has been closed. How can I now drag the attention so that the question could be reopened?

@BruceAbbott To run DFPlayer for sound effects.

@MarcusMüller I don't edit answers from other people because I don't find that appropriate. BTW, you also used 200mOhm instead of 100mOhm inductor in your area estimate. This seems OK to me, as there is only small voltage drop involved. The point is I read all posts carefully. I am probably slow to you, because I have no training and little experience in electrical engineering, but I am trying hard.

@MarcusMüller In ti.com/lit/ds/symlink/tps560200.pdf page 20, chip itself is 3mm x 3mm, with pins included. KiCAD gives 0.5mm in one direction and 1.5mm in other, which seems reasonable.

@MarcusMüller 5 mm × 2.5 mm is not 7.5 mm². At least in Slovenia. Maybe you have different multiplying table in Germany. SOT23-5 is 3.5mm x 4.5mm in KiCad. TO220 is about 2cm x 1cm. And I have 7805 TO220 in my permanent stock so I know very well how big it is. I designed PCBs with KiCAD before and I found their default sizes quite correct.

@MarcusMüller TO220 alone is 200mm2 in horizontal position. Plus your calculation in your answer is wrong. 1. You calculated area for inductor wrong, 2. you most certainly used wrong dimensions for IC, 3. in KiCAD machine soldering areas are approximately the same size as your hand soldering sizes. 4. When I actually lay all the components in KiCAD I have more than 60mm2 with spaces between because of unideal conditions.

@KyleB No, unfortunately this is too big.

@MarcusMüller I was not aware that switching regulator is WAY SMALLER than linear regulator. Very hard to solder but at least it might work for me.

@KyleB I am already using both sides :)

@Hearth Sorry, didn't know that, as I am not an electrical engineer and also my native language is not English.

@Unimportant I was searching mouser to find a 5V switching regulator that only need an inductor and has exposed pins so it can be hand soldered and couldn't find any.

@MarcusMüller If I use linear regulator, there is no space problem - you only need linear regulator and two capacitors. But there is too much heat dissipation. If I use switching regulator, there is very low heat dissipation, but you need six additional elements, so it needs too much space. I can't open your link because I don't have TI account.

@MarcusMüller OK, I got it now - linear regulator is out of question, thanks! I was hoping to use around 50mm2, which is 1/4 of total PCB size. And I would like to avoid 0201 and 0402 elements as they are too small for hand soldering.

@MarcusMüller What exactly do you consider under "buck converter"? Switching regulator like TPS54202H? This is rather small but it requires six additional elements, so in total takes too much space for my design. I cannot hand-solder reliably elements smaller than 0805.

@Andyaka Thanks for the suggestion. This is why I already took a look into MCP1702/1703 data sheet and put the numbers into the question. My calculation shows that required heat dissipation is twice of maximum available heat dissipation.

@Andyaka So you think using SOT89 for buck voltage from 12V to 5V at 250mA will probably be fine, and manual is too cautious?

@MarcusMüller TO220 was suggested by Kyle B and it is actually a regulator module (PCB + chip + additional components) of the same size as TO220 linear regulator... which again proves my point...