[time-nuts] time-nuts Digest, Vol 63, Issue 52

[Magnus Danielson]

Arnold Tibus wrote:
> On Sun, 11 Oct 2009 12:10:36 +0000, Mark Sims wrote:
> 
> 
>> Alas,  if there was only an FM to R...  there is some useful information in the revision history/comments near the beginning of the file heather.cpp
> 
>> The OSC graph defaults to OFF because it tends to be a very jagged and noisy looking graph that gets rather annoyingly in the way of things.  The next rev of the program has a display filtering option that makes that plot look a lot more tame.
> 
>> The OSC param is shown in PPB in the status info at the top of the screen since that is the way it comes into the program.   It is shown in PPT in the plots since that gives values that are much easier read against the scale divisions on the screen.   I have considered converting to PPT in the status info,   can't remember why it stayed PPB...
> 
>> ---------------------------------------- 
> 
>> RTFM comes to mind:-) 		 	   		  
> 
> 
> Mark,
> 
> personally I have some promlems with the expressions as in LH 
> ppb, ppt etc. used because there are different meanings about 
> around the world and this is therefore misleading, error-prone.
> If I search in the Internet I do find lots of discussions about.
> Is there no way for an improvement, no standardization? 

There is an international standard for it. It is part of the SI standard 
which is also realized in ISO 31.

http://en.wikipedia.org/wiki/International_System_of_Units

However, the first map could indicate why you have troubles.

> I learned that in Germany (and all over continental Europe?)
> 1 thousend = 1.000				= E3
> 1 million = 1.000.000			= E6
> 1 milliard = 1.000.000.000			= E9
> 1 billion = 1.000.000.000.000			= E12
> 1 billiard = 1.000.000.000.000.000		= E15
> 1 trillion = 1.000.000.000.000.000.000		= E18
> 1 trilliard = 1.000.000.000.000.000.000.000	= E21
> and so on with
> quartrillion, quartrilliard, quintillion, quintilliard, sextillion, sextilliard....
> 
> and in the US and some more countries it is
> million = 1,000,000
> billion = 1,000,000,000
> trillion = 1,000,000,000,000
> and further ...? ( fantastillions acc. Donald Duck ;-))  )
> 
> wouldn't it be more scientific and less error-prone  to agree at least to
> xE-3	instead of 	(m)
> xE-6	instead of ppm	(µ)
> xE-9	instead of ppb	(n)
> xE-12	instead of ppt	(p)
> xE-15	instead of pp? etc.	(f)
> xE-18			(a)
> (how do one express parts per mili...(E-3)?)
> 
> or if not wanted perhaps then this way :
> x10^-6
> x10^-9
> x10^-12
> etc.
> 
> or could one type eg. m, µ, p, f, a for milli, mikro, nano; pico, femto, atto?

These later two are standard representations. A particular problem is 
that not all text formats contains the my symbol, so u has been an 
accepted shorthand and so far this have not been result of confusion.

The 1.0E-6 form is adapted to ASCII computerized form, any should maybe 
be avoided if possible.

> When used these numbers in calculations we anyway have to convert 
> these ppm, ppb, ppt etc. to scientific numbers using exponents

4,6 µ can be confusing compared to 4,6 ppm as part-per- indicate a 
normalized relative measure. This is as handy as in percent, promille, 
part-per-milion etc. I can't recall a suitable means to handle it, but 
you can write it in the ugly for of 4,6 µHz/Hz which would only convey 
the normalization part and not the relative aspect (f-f0)/f0.

> There are too often discussions and misunderstandings 
> because the ignored case sensitivity of units (b for bit, B for Byte,
> m for milli, M for Mega...). 

This can only be solved by means of education and correction.

Notice that 1 MB is to be interprented in 1.000.000 Bytes in SI 
standard. If you want 1.048.576 Bytes you shall now write it as 1 MiB.

http://en.wikipedia.org/wiki/Mebi

It's ugly, i know, but solves the problem and is consistent.

> Btw. I remember to all these strange mmH, µµF etc. when I collected 
> rare inductors, capacitors revovered from vintage MIL- equipment in 
> the end fifties/ early sixties of last century ... :-)

You still see stuff like 0,001 uF and 0,010 uF regularly, even if they 
should be baned as their propper form of 1 nF and 10 nF is what should 
be used.

I find myself using the shorthand of inserting the prefix number in 
place of the decimal point, as this can easilly become hart to read, so 
a 2k2 resistor is easy to convey in schematics and ASCII formats, while 
2,2 kOhm is more propper.

The carefull reader will discover my use of the "," for decimal place 
and "." for digit separation. The US convention works the other way 
around. It is also part of the US adaptation of the SI standard, so care 
should be taken not to interprent the NIST publication as conveying the 
correct detail for certain things, they are only to be viewed as local 
interpretation to the USA, possibly only recommended use.

> I believe that Time Nuts prefer precise and clear expressions!? ;-)
> What do you think about it?

There is still some debate to be had. I don't recall that there is a 
propper "SI style" relative normalized form. Need to check the docks, 
but if someone could enligthen me that would be super.

Cheersm
Magnus