Capacitor labelling

[JohnH]

Was it really a 0.22 uF cap, or a 0.022 uF (223) cap?

This from ChrisK prompts me to convey my understanding of labelling conventions for small caps in case others may be confused as I was, particularly that format such as 223, which is not too obvious.

taking that typical tone cap as an example:

0.022uF = 0.022 micro Farad
=22nF = 22 nano Farad
=22,000 pF = 22,000 pico farad

so check this - if one is labeled 223, we interpret that as 22 x 10^3 picofarads?

This type of notation seems most common on small caps

John

[sumgai]

John,

Yeah, that's about right. More or less, the right-most digit is the number of zeros you add onto the first two digits. It shouldn't take a propeller-head to see that if it's a small disc or something similar in size, the value is in pf.

224 would be 220,000 pf (which translates to 0.22 mf)
225 would be 2,200,000 pf (which translates to 2.2 mf)
etc.

However, there is sometimes a letter in the middle of that designator. IIRC, that's a multiplier in the traditional Greek sense. That yields:

2K3 is 0.022 mf (the conversion is already done, so to speak)

and I think we all get the gist of this, eh?

Additionally, if the 3rd digit is 8 or 9, then you multiply the first two digits by 0.01 or 0.1 respectively. No 3rd digit at all is the base number in pf. If the cap is large enough, the actual designator will be used, such as 10 uf, for example. Obviously we won't be doing any multiplying on that one.

And finally, a letter after the designator is a tolerance code. If you see 3 letters off by themselves, that's a temperature co-efficient - usually it's NPO, the opposite of a normal capacitor's reaction to heat. When things get hotter, the capacitance usually goes up, and vice versa. An NPO type performs in just the opposite manner, so it's usually more expensive to produce.


sumgai

[JohnH]

sumgai said:

2K3 is 0.022 mf (the conversion is already done, so to speak)

and I think we all get the gist of this, eh?

sumgai

Maybe not quite all of us sumgai!..Can you just run that by me again? How is 2K3 equal to 0.022microFarads?

with thanks

John

[fobits]

2K3 is 0.022 mf (the conversion is already done, so to speak)

Are you sure about that? I've never seen or heard of 'K' used with capacitors.

A letter is sometimes used in place of the decimal point, because it's easier to see. A 2.2k resistor might be mistaken for 22k, but there's no doubt if it's written as 2k2.

In the same way a 3.3nF cap can be written as 3n3, and 4.7 uF as 4u7, etc.

But 'k' on a capacitor? If I saw a thingie like that, I would assume that it was an odd-shaped resistor.

[ChrisK]

There are a number of "conventions" afoot. One of the things that we have to be careful of is the continuing use of mF for microFarad (the original way). PicoFarads were originally designated as micro-microFarads (mmF). In the days of old, terms were oft pre-limited by what was thought to be realizable. My father still won't believe that a 1" diameter 0.5" high device can be a Farad or more (to him, the Sprague 1 Farad 3 volt cap was the size of a quart of milk). He also won't believe me when I talk about multi-layer PCBs. He worked on PCBs in the 50's where they actually shrank during processing. "Growing plated thru buried vias in a 24 layer PCB, NAH, yer makin' it up laddie."

Today, m indicates milli as in milliFarads, which would see use in double layer caps that can achieve up to several Farads per cubic inch. There are ultra caps that have capacitance up to thousands of Farads in relatively small packages, an electric car so powered set some king of record (perchance acceleration) about a year ago.

There are many types of capacitors. Ceramics alone can come in Z5U for supply bypass, terrible for analog circuits, X7R for general analog use, NPO for tempco uses, and COG for those requiring the really good. Capacitor markings are actually less specific or uniform than resistors. Usually it's two digits and an exponent in base10, although I have seen "K" used as well. Since capacitors are rarely available better than 5% (usually 20% and sometimes10%), the three digit marking works well. (I do have a multi-decade assortment of film caps that are 1%, but those are golden.)

pF picoFarad 10^-12 Farads
nF nanoFarad 10^-9 Farads
uF microFarads 10^-6 Farads (?ucroFarads?)
mF milliFarads 10^-3 Farads

Some day we'll need something for kiloFarads and megaFarads.

Resistors are usually color coded, but it's been years since I've dealt w/ these on a regular basis. Everything went surface mount in my world many years ago, so I live in the land of digits. 103 is 10K, 101 is 100 Ohms, 1002 is 10K 1% or better and 12R0 is 12 Ohms 1%. I actually use 1% resistors everywhere and rarely 5%.

The method of specifying resistors w/ the embedded "K" or "R" is best. There is no mistake with 10K0, 12R0, or 2n2F, and no faint decimal point to "resist copying".


The value division of resistors (and caps as well) is unknown to many. There actually are 96 1% values (per decade), 48 2% values, 24 5% values, 12 10% values and 6 20% values.

Unknown to most is that the 1, 2, 5, 10, and 20% tolerance bands are just the rounding to an integer, of the actual tolerances per step.

The step for "1%" tolerance is actually the 96th root of 10 or 1.024275 (or +/- 1.21376%). Still rounds to "1".

The step for "2%" tolerance is actually the 48th root of 10 or 1.0491397 (or +/- 2.45699%). Still rounds to "2".

The step for "5%" tolerance is actually the 24th root of 10 or 1.100694 (or +/- 5.03471%). Wow, real near "5", rounds to "5".

The step for "20%" tolerance is actually the 6th root of 10 or 1.467799 (or +/- 23.38996%). Oops, starting to get out of hand again.

When the old Fender schematics talk about a +/- 20% variation, they were optimistic.


If this got a little to "mathy" for y'all, remember that yer playing an instrument that realizes all of the 12th roots of 2 all day long.

Fret spacing = 2^[1/12].

[sumgai]

Gang,

I think Chris cleaned it up nicely, don't you? I shouldn't have to go any further in answer to John or Frank, right?

Yes, the K in a cap is rare, but I've seen it. And Chris is equally right when he warns us that using the "old" convention of 'mf' (or 'mF') to mean micro Farads is in danger of extinction. Since the early days of using Roman designators for binary math in the computer world, the conventions have been increasingly moving into the electronics world as well, replacing the Greek designators that us old farts grew up with. It would have been nice if we could get the Greek letter that looks like a lower-case "u" to use for a 'micro', but alas, 'twas never meant to be.

I think that by the time I hang up my soldering iron for good, most of the conversion, if not all of it, will be over, and the Roman system will be fully functional for all manner of marking everything. Couldn't happen too soon for me.

Chris, nice breakdown on the values/tolerances business. More than John was asking for, but I'll lay odds no one here will complain. ;D


sumgai

[ChrisK]

Just goes to show that in "olden" days, in some cases, there were actual plans afoot!