Does it matter in puffs on a crystal? - SOLVED

Hello, I always wanted to know something that I never figure out and really wasn't on my mind into now. But in crystals Always seeing 16mhz or 20mhz crystals on the 328p but at the end i see like 18pf or 20pf. So my question is does it really matter about the puffs? I'm just trying to figure it out. Also how do you know how many puffs is right for your project?

Joseph

If you look on the data sheet for the processor , it will tell you capacitor sizes needed when running at a given frequency .
A Atmega328 for example at 16Mhz is 22pF .
You do need them, and the correct value, or the processor clock will not run . They should be close to the processor too.

Data sheet should always be your first port of call - open with a good pdf reader !

Hello, Thank you for that. Me and datasheets don't see eye to eye. Always confusing me. That's all on me. So what does the puff really do and mean? I was curious to know.

Joseph

Crystals can oscillate at different frequencies, especially multiples of their rated frequency, which would be too high for the chip to run. The caps help to suppress those other frequencies.

You may find that the circuit will run ok without the caps, which might make you think they are not needed. But then you move the circuit from breadboard to PCB, or add extra components, or the weather changes, or... and suddenly it could stop working and you could waste a long time trying to figure out why it stopped working. It's not worth the risk - add the caps at the start. Same goes for decoupling caps for chips.

And they are capacitors “ picofarads” in value.
1pF = 1/1,000,000,000 Farads.

Data sheets are your best friend ( well maybe) , if you open in a good reader , you can search through them

So that means the crystal itself has capacitance like a capacitor has? That is what I was always confused on.

Joseph

Are you sure? Isn't 1pF = 1.0E-12F?

Of course it has, but it also has inductance! :astonished:

The crystal behaves as a tuned circuit with an extremely high "Q".

The capacitors function as part of a Colpitts oscillator.

picoFarads 1E+12
nanoFarads 1E+9

Tom... :grinning: :+1: :australia: :coffee:

Yep …lot 000000’s

Sorry got it wrong !

Hello all, Thank you for helping me to understand about more detail on crystals and the whole puff system.

Joseph

This article may help.

Lots of professionals have difficulty setting the correct capacitors. I have had to alter designs where the client was having their products being unreliable due to oscillators failing to start. If you get one in a thousand failing to start under some condition, e.g. temperature. Then if you may half a million/annum the failures soon mount up.

It is rumoured that the EU is about to limit the amount of surprises any theatrical or movie performances is permitted to inflect. The standard unit of surprise is the BOO and is defined as the amount of surprises needed to kill a man.

Of course this is way too high to use in practice, so it is standard to give ratings as 1E-12 of a BOO or as it is better know, a Pico BOO.

Oh dear, everyone's having trouble with this.
pF = picofarad = 1E-12 farad = 0.000000000001 F
nF = nanofarad = 1E-9 farad = 0.000000001 F

IE negative powers of ten! And its the farad (symbol F), not the Farad.

Fun fact: before SI prefixes like nano and pico were around the pF was called the µµF and values were only ever given in µF or µµF, there being no equivalent of the nano prefix. Even today you see crazy values like 0.001µF or 10000pF being widely used instead of 1nF or 10nF.

And just as bad, a 1000 µF electrolytic which is actually simply 1 mF. :roll_eyes:

Of course, the reason for the hang-ups is that only recently (to me anyway) have 1 Farad and more capacitors actually been real components.

In the Dark Ages mF was used for microfarad. I think for this ambiguity using mF is avoided.