WP-34S galUS-> l Precision

[Claudio L.]

(12-15-2016 09:08 PM)Nigel (UK) Wrote:  Thank you for taking the time to check the conversion factors; I've looked at them again as a result. This is what I've found.

• galUK (including flozUK) A google search yields several sites which state that one UK gallon is 4.54609188 litres, but without (so far as I can see) giving a source. My source was the UK Weights and Measures act 1985, which is rather old. I find the same definition given in the 1995 amendment to this act, and (more authoritatively) at this NIST page where the 4.54609 value I have used is listed as being exact. In the absence of a source contradicting this, I'll leave it unchanged.
• pc The sources on which this is based (a 2009 redefinition of the au by the International Astronomical Union (IAU), ratified in 2012, and an exact definition of the parsec as being exactly 648000/pi au, introduced by the IAU in 2015) are both more recent than the definitions in force when the HP-50g was released and so I'll leave this unchanged as well.
• ly The issue here is the definition of "a year". I've used the Julian Year, exactly 365.25 days, in accordance with this recent document from the IAU. The HP-50g value appears to use the Mean Tropical Year, or 365.24219 days. I think that it would be reasonable to stay with the IAU definition.
• mmHg The value given by the HP-50g is exactly 1/760 of 101325 Pa (1 standard atmosphere), which is generally known as 1 torr and not 1 mmHg. Reference 4 on the mmHg Wikipedia article is to an NPL page that (in footnote 2) lists the standard density of mercury that I've used to derive my value for mmHg. Interestingly, footnote 3 on the same page states that there is a concerted effort to remove units such as mmHg, either by excluding them from conversion tables altogether or by restricting their precision! If we aren't going to remove mmHg from the WP-34s altogether I think it should stay as it is.
  

So I think the values listed should stay as they are, with the possible exceptions of mmHg and inHg which (it could be argued) belong to an age that is past.

Nigel (UK)

I just double checked the definitions above and match 100% what I used in newRPL (see post #11), including the mmHg definition. So I agree these are the best definitions we have at the moment.