common decoupling cap - why 100nF

Thread Starter

bug13

Joined Feb 13, 2012
2,002
Hi guys

So I always use decoupling cap where possible, and as close to a chip as possible. I was told 100nF is usually good enough.

But why 100nF, not 47nF, or other values?

Thanks guys!
 

Sensacell

Joined Jun 19, 2012
3,432
They are small, cheap and readily available. It's not critical that it be a specific value, but 100 nF is about right.
Large capacitors tend to have larger internal impedance at high frequencies, longer lead structures with higher series inductance- going bigger is not necessarily better. You will often see a range of capacitors across the power supply bus, big electrolytics, tantalum caps - and ceramic caps near each chip.

The combo has lower impedance than any one could provide alone.
 

ErnieM

Joined Apr 24, 2011
8,377
Because in most every application 0.1uF works just splendidly. It's time to quote Alexander Pope once more:

Alexander Pope said:
Whatever is, is right.
They give coverage over a good range of frequencies. If some higher frequency protection is needed, add in a 0.001uF cap in parallel. For low frequency coverage switch to tantalum or electrolytic (tant is better but more expensive) of a value that depends on the low frequency current draw.
 

ScottWang

Joined Aug 23, 2012
7,397
You want to know why?
Connecting the +V and Gnd of power supply to the breadboard.
To measuring the +V and Gnd on terminal of breadboard before you plus the cap and after, comparing the noise and voltage amplitude.

Changing the different cap and measure them repeat again.
 
Last edited:

alfacliff

Joined Dec 13, 2013
2,458
if you use too much decoupling cap, the initial turn on current is very high. .01 is about right to give enough cap without a high surge when turning on, especially when there are a lot of decoupling caps.
 

ErnieM

Joined Apr 24, 2011
8,377
Generally speaking bigger is better, but also costs more.
These are ceramic. Electrolytic are different.

I like that graph too as it shows bigger is not necessarily better.

The orange line is for a 2.2 uF cap, the largest of the bunch. Note it stops working as a capacitor (where the line turns back up) between 3 & 4 MHz. That's because it is looking inductive at those frequencies. 3 or 4 MHz is still within the response of many digital and analog components

Our 0.1uF cap is in blue and is capacitive way out till 100 or 200 MHz. That is beyond the response of many components and why it is a go-to value.
 
Top