12

I keep running into formulas in datasheets and articles that appear to be abbreviating pi as 'p'. An example from Ask The Applications Engineer-25: Op Amps Driving Capacitive Loads shows "1/2p..." in many places when calculating frequencies of poles and zeros.

Formulas and plots from ADI

I also see it in a TI datasheet for the OPA627 (see figure 32).

This confused me, because I thought of the SI prefix for pico. Is this common in circuit literature?

Peter Mortensen
  • 1,676
  • 3
  • 17
  • 23
Ben Froelich
  • 123
  • 1
  • 5
  • That's a first. – DKNguyen Sep 28 '22 at 20:54
  • 15
    Is this common? No. Could be a font issue. Maybe original author used symbol font, publisher substituted normal font. – Mattman944 Sep 28 '22 at 21:02
  • Somebody was just lazy, or didn't have ready access to \$\pi\$, or made it their personal practice and decided that since it was good enough for them it would be good enough for their readers (and if a reader complained the author would just quip, "So it confuses you? You should leave the profession then. It's meaning is obvious to any informed reader! And I don't care about uninformed readers.") – jonk Sep 28 '22 at 23:45
  • 8
    Actually there is a link to the pdf version of the tutorial, in which the pi randers fine as greek pi. – Mike Sep 29 '22 at 04:32
  • 4
    @Mike which of course suggests a font issue when the figures were rasterised for the webpage. My suspicion is that the original author used neither Windows Symbol nor a more modern approach with a font containing Greek characters, but an unusual symbol font that then wasn't included in the PDF and was substituted wrong. – Chris H Sep 29 '22 at 13:52
  • That doesn't explain the datasheet where it's copyable text in the PDF (and the PDF apparently contains a font with "symbol" in the name). That leads to my other suggestion: the whole thing written in one font (e.g. Times New Roman) + Symbol, and someone highlighted all the text then changed to a sans serif font, forgetting about the Greek chars – Chris H Sep 29 '22 at 13:53
  • @Mike I don't see any π in the linked PDF, but I do see "p" and Ω. Ubuntu 18.04 here. – Izkata Sep 29 '22 at 14:31
  • @NickMatteo Ah nevermind I was looking at the wrong graph (one of the similar ones that doesn't have π). If I search for "π" the only result is in a paragraph of text (what prompted my comment), but searching for "2p" does find the "2π", so yeah, fonts. – Izkata Sep 29 '22 at 14:40
  • 1
    From "A History of Pi", IIRC, P was used for Pi, before Pi became standard. So, older texts may have that. – Zenzizenzizenzic Sep 29 '22 at 15:01
  • 9
    There's also a weirdly placed dollar-sign ($) in the formula in the upper right... more suggestion that this is a font issue. – kgutwin Sep 29 '22 at 16:01

3 Answers3

44

No, it's not common.

While I can't be 100% sure, I believe this is a font issue with whatever tool was used to render the figures for display in the web version of the article (the PDF version displays the pi correctly).

In the old days, people represented Greek letters in their technical documents by selecting the "symbol" font. While nowadays we tend to use Unicode instead, the old symbol font sticks around for backwards compatibility.

In the symbol font, the code positions that would represent Latin letters in a normal ASCII font are instead used for Greek letters. Specifically code position 0x70 is π in the symbol font and p in ASCII.

So if a character that was intended to be rendered with the symbol font is instead rendered with normal font then the intended Greek letter will be replaced with a Latin one.

Peter Green
  • 21,158
  • 1
  • 38
  • 76
  • I would go as far to say it's *never* done except in error. I've never, ever, seen anything other than "pi" _used on purpose_ when Greek letters are unavailable, except by typo or font or glyph substitution error. (Not in maths or engineering; you sometimes see Greek written in the Latin alphabet, in which case "π" is always just "p".) – jonathanjo Sep 29 '22 at 13:35
  • It is a font rendering issue. This is how Wine renders LTspice IV help file: https://i.stack.imgur.com/r5WVG.png. It used to like so be even for XVII, until r ecent updates. – a concerned citizen Sep 29 '22 at 13:39
  • 4
    I would argue that it's more of a font selection issue than a font rendering isssue. – Peter Green Sep 29 '22 at 14:07
21

'π' is 0x70 in the symbol font and 'p' is 0x70 in ASCII

Similarly, 'Ω' is 0x57 in the symbol font and 'W' is 0x57 in ASCII -- you frequently see 'Ω' improperly rendered as 'W' in PDFs.

Something to look out for if there is possible confusion.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
  • 7
    Ohms vs. Watts is a more insidious version of the error. I've seen a diagram with a heater drawn as a resistor and specified with a "W". I can't remember what it was meant to be, only that we had to figure it out from the fact one interpretation would have done bad things to the PSU. – Chris H Sep 29 '22 at 13:56
11

No, I've never seen it before. Maybe my beard isn't grey enough to make a blanket statement for the community, but I stand behind it.

If you can track down a link or email of the datasheet maintainers I would suggest complaining, because the last thing I'd like to see is little 'p's popping up in all my circle and frequency equations.

Bryan
  • 2,156
  • 1
  • 6
  • 11