Why do Americans not use nanofarads?

[gbaddeley]

I used to use .1uF etc but switched to 100nF because it avoided the use of a decimal point, which could become lost in photocopies or poor quality prints.  Values like 4.7n I would write as 4n7 or 4700p to avoid the decimal.

[David Hess]

I used to use .1uF etc but switched to 100nF because it avoided the use of a decimal point, which could become lost in photocopies or poor quality prints.  Values like 4.7n I would write as 4n7 or 4700p to avoid the decimal.

I always include a zero in front of a leading decimal point, so 0.1uF.

[gamalot]

I just checked, on the DigiKey website, some of the capacitance values in the list are in nF, but only for some E12 series values. Mouser still doesn't use nF as a unit for capacitance at all.

[paulca]

At least on Windows, AltGr+m produces µ. I have finnish/swedish keyboard layout so YMMV.

Regards,
Janne

I was going to say it's hard to enter µ because the forum software eats it, but apparently you have succeeded.

As a test, here is 10 microfarads written with the Greek letter mu and not the graphic character mu: 10 ?F
And here is 1000 ohms: 1000 ?

Let's see which of them survive.

(Apparently neither of them.)

I'm not sure if it is due to forum software but choice of character encoding in the database containing this forum. Perhaps it is set to some ISO-charset instead of UTF. Maybe Dave will tell us and could it be possibly changed?

It can be anywhere in the chain really.  The users browser locale settings, the underlying OS locale settings, the HTTP request encoding, all the way through the forum software down to the database collocation settings on the database, table or columns.

The difference in survivability may come down to UTF-8 versus iso-8859-* (or windows codepages) encodings from the browser/local OS.

When I was dropping Turkish special characters in a video UI generation application it took me a week to "normalise" the full flow into the turkish iso-8859-? encoding before I finally seen end to end correctness.

These days if you ever get asked about what encoding you would like, just pick UTF-8 unless you want to write some exotic or rare pacific languages or dialects where you might need to go the rarer character encodings.  Japan for example of 15 different encodings, IIRC

EDIT:
If you look here: https://www.compart.com/en/unicode/U+03BC
There are multiple options with different unicode ids.
From a bot, as I got lazy:

In the ISO-8859-1 character encoding, the Greek small letter mu (μ) is represented by the character code 236, which corresponds to the Unicode code point U+03BC.
 This character is also present in other character sets such as ISO-8859-7, where it is similarly encoded as μ.
 The ISO-8859-1 standard, also known as Latin-1, defines a set of 256 characters used primarily for Western European languages and is considered the default character encoding for the World Wide Web.

[0xdeadbeef]

In German, it seems that "Nano" and "Giga" were already used (for distances) in 1928. But of course the SI was only defined in 1960.

[aeg]

From a bot, as I got lazy:

In the ISO-8859-1 character encoding, the Greek small letter mu (μ) is represented by the character code 236, which corresponds to the Unicode code point U+03BC.
 This character is also present in other character sets such as ISO-8859-7, where it is similarly encoded as μ.
 The ISO-8859-1 standard, also known as Latin-1, defines a set of 256 characters used primarily for Western European languages and is considered the default character encoding for the World Wide Web.

Complete word salad. Totally wrong. The letter μ corresponds to the Unicode codepoint U+03BC; the character code 236 does not correspond to the Unicode codepoint U+03BC. The character does exist in ISO-8859-7, but what on Earth does it mean for μ to be encoded as μ? And there are not 256 characters in ISO-8859-1, there are only 191.

[UnijunctionTransistor]

I also hate decimal points in schematics.
I have dealt with my fair share of older schematics, either copied or scanned, where the decimal point has either disappeared or is confused with a dust spec. It gets worse, because the leading zero use is not standardized; for instance 10 nF is sometimes written as 0.01 uF but other times .01 uF, drives me nuts.

[UnijunctionTransistor]

As an American, I'd just like to say that we don't use nanofarads because it's another stupid power-of-ten. We'd much prefer to use twelfths of a microfarad, or 1/5280th of a millifarad or something like that.

D*mn metric system.

You read my mind. ISO units are socialist units, that is the meaning of the “S”. 
Stop their World domination, long live freedom units.

And continuing with this thought process, why stop there?
Similar to a mile and nautical mile having a different value depending on the use, let’s also define semiconductor ohms, real resistor ohms, and other ohms.

[SiliconWizard]

I also hate decimal points in schematics.
I have dealt with my fair share of older schematics, either copied or scanned, where the decimal point has either disappeared or is confused with a dust spec. It gets worse, because the leading zero use is not standardized; for instance 10 nF is sometimes written as 0.01 uF but other times .01 uF, drives me nuts.

Yes, decimal point on schematics is a no-no in my book and that of many "serious" companies. And yes, the additional dropping of the 0 in front of the decimal point makes it even worse. (And when I read some comments that more or less admit that it's an almost purely "political" stance, that's crazy. I'm all for independence and freedom, but I don't think politics has a place in science & engineering.)

I strictly use "infix" notation for decimal values when needed, such as "4k7", "2µ2", etc. Which avoids any ambiguity regarding "points". And when there is no multiplier involved, which rarely happens with other values than resistance values, I use "R" in the infix position (such as "1R5" for 1.5 ohm). This is all pretty standard. Just my 2 cents.

[jpanhalt]

Re. decimals:

No one writes ".1 uF".  Or, a least they shouldn't.  The standard is 0.1 unit, which even without a visible decimal point should not be ambiguous.  In some fields, that standard is written.

[ejeffrey]

That's largely archaic.  There are still people out there that will go straight from pico to micro and you even still see microfarads written as "mfd" but generally use of nanofarads is now relatively common in the US.

If I had to guess, it' would be  added on from  capacitor values were marked before everything got too small to mark.  Film and ceramic capacitors were often marked without units at all, but with a three digit code similar to resistor color bands. They were marked in picofarads with an exponent digit like 471 = 470 picofarad, 682 = 6800 picofarad. 

[TimFox]

Re. decimals:

No one writes ".1 uF".  Or, a least they shouldn't.  The standard is 0.1 unit, which even without a visible decimal point should not be ambiguous.  In some fields, that standard is written.

I agree about leading zero before decimal point, but try to shift prefixes to keep the value greater than one.
My professor (from Italy) used to confuse people by reading 0.1 as “zero one”.

[paulca]

Scientific notation, IIRC has certain rules around precision communication when dealing with measured values.

So
0.6x10^10 is MORE valid in many situations than 6.0x10^9 is.

The former has it's sig.figs such as may match it's measuring accuracy.  0.6 - the 6 is your limit of precision.  0.60 would exceed it.  Thus 6.0x10^9 affords an order of magnitude higher precision than the measurement held.

It matters when you start to calculate with multiples of these and need to track the precision of the values to determine the precision of the output.

[gamalot]

Scientific notation, IIRC has certain rules around precision communication when dealing with measured values.

So
0.6x10^10 is MORE valid in many situations than 6.0x10^9 is.

The former has it's sig.figs such as may match it's measuring accuracy.  0.6 - the 6 is your limit of precision.  0.60 would exceed it.  Thus 6.0x10^9 affords an order of magnitude higher precision than the measurement held.

It matters when you start to calculate with multiples of these and need to track the precision of the values to determine the precision of the output.

Is 6x10^9 illegal?

[SteveThackery]

Is 6x10^9 illegal?

Not really. There is a custom in engineering that the exponents jump in threes - you will find this almost everywhere. However, in science and mathematics this custom is not normally observed and the exponents is chosen in the way Paul describes.

[gamalot]

Is 6x10^9 illegal?

Not really. There is a custom in engineering that the exponents jump in threes - you will find this almost everywhere. However, in science and mathematics this custom is not normally observed and the exponents is chosen in the way Paul describes.

I think you might have misunderstood me. What I meant was, why would he write it as 0.6×10^10 instead of the simpler 6x10^9.

[SteveThackery]

Is 6x10^9 illegal?

Not really. There is a custom in engineering that the exponents jump in threes - you will find this almost everywhere. However, in science and mathematics this custom is not normally observed and the exponents is chosen in the way Paul describes.

I think you might have misunderstood me. What I meant was, why would he write it as 0.6×10^10 instead of the simpler 6x10^9.

Oh, I'm sorry - I did misunderstand you. I think Paul covered your question, but I will leave it to him to respond.

[BILLPOD]

I always include a zero in front of a leading decimal point, so 0.1uF.
[/quote]

I agree with David Hess on this.   I once had a physics professor that insisted on a leading 'zero', and would mark a test question as wrong if none were used.  He would say:  "How would I know that a fly didn't land there and leave his mark".  That has stayed with me over 50 years, now.

[ejeffrey]

Scientific notation, IIRC has certain rules around precision communication when dealing with measured values.

So
0.6x10^10 is MORE valid in many situations than 6.0x10^9 is.

In standard scientific notation you would write 6x10^9 if you have one significant figure.  You always have a single non-zero digit to the left of the decimal point. 

Although to be honest that convention isn't always strongly followed. For instance it's pretty common to fall back to normal decimals when the exponent is small (I.e., 34.6 not 3.46e+1). 

As always if you need to be precise about the uncertainty you just call it out specifically.  The two ways to do that are with a +/- X, or to underline the uncertain digits.

[floobydust]

A very long time ago maybe 1980 I'd asked an old timer why 0.001uF or 10mmF or 100,000uF  etc. on schematics.
He replied it was for an engineer's intuition. Keeping the order of magnitude, everything referenced to the microfarad, told you what/where the capacitor was being used, what it was being used for and ballpark frequencies of interest/coverage.
From the old days of radio, I guess. I still don't use nF, or mF - the IEC can go suck it.

OT - I was taught to use one leading zero before the decimal point i.e. 0.1uF
Displaying extra leading zeros in metrology pisses me off, I hate it and it conveys no additional information other than for idiots who don't know what range they are on.
My last rant about it was here: https://www.eevblog.com/forum/chat/why-do-digital-multimeters-not-have-zero-blanking

[jpanhalt]

OT - I was taught to use one leading zero before the decimal point i.e. 0.1uF
Displaying extra leading zeros in metrology pisses me off, I hate it and it conveys no additional information other than for idiots who don't know what range they are on.

In some instances, e.g., reporting clinical laboratory results that is effectively the law.  That is, a lab must be accredited to get paid, and accrediting agencies demand that single zero.  Plus, it makes sense. The practical requirements in clinical testing effectively require the use of specific, non-SI units.*  NEJM tried to enforce SI units in the early 90's.  That brain fart by a young editor-in-chief did not last long.

As for suppressing superfluous and misleading leading zeros, it's so easy to do, it ought to be a felony not to do it.

*As an example, physicians will refer to a patient's BUN being "18" or serum glucose as being 98, etc.  A large percentage of them cannot tell you the actual units (personal observations).  In addition, when I retired (2006), reporting systems required us to use a single unit for each test.  You couldn't change that unit for some values and not for others.  I assume that is still the case, and it's reasonable.

[SteveThackery]

A very long time ago maybe 1980 I'd asked an old timer why 0.001uF or 10mmF or 100,000uF  etc. on schematics.
He replied it was for an engineer's intuition. Keeping the order of magnitude, everything referenced to the microfarad, told you what/where the capacitor was being used, what it was being used for and ballpark frequencies of interest/coverage.

That doesn't work for me, I'm afraid. For me, micro-, nano- and picofarad are totally logical and natural, and I have the intuitive "feel" of them in my gut, so to speak. Things like 0.001uF drive me nuts.

Weirdly, I haven't yet picked up the habit of millifarads. I still write 10,000uF. However, thanks to this thread I'm going to change my ways and use mF. It is the obvious and logical thing to do.

[jpanhalt]

It is the obvious and logical thing to do.

The only obvious and logical units I know are the foot and inch.  Welcome to the club.

[paulca]

Is 6x10^9 illegal?

Not really. There is a custom in engineering that the exponents jump in threes - you will find this almost everywhere. However, in science and mathematics this custom is not normally observed and the exponents is chosen in the way Paul describes.

I think you might have misunderstood me. What I meant was, why would he write it as 0.6×10^10 instead of the simpler 6x10^9.

Oh, I'm sorry - I did misunderstand you. I think Paul covered your question, but I will leave it to him to respond.

Nope.  I have been trying to remember the rules, but I can't even remember how to add/sub/div and multiply scientific notation "properly".

It was through those rules they taught us how to account for precision.  Back in 2002!

Like multiplying 10.0 and 10 or( 1*10^1 * 1.0*10^1 ), what is the answer?  From recall only I believe the answer is 1*10^2  (no decimal), the output tracks the lowest precision, because multiplying it does not change the precision of the inputs.   

If you need this defacto, look it up, I have not had to even touch this stuff in decades.