HPA L2 noise

• Gaussian
• 

It indicates the normalised power of the signal

Its the same as the root mean noise with units of Volts dissipated over 1ohm. It gets squared so that the units end up as power

It describes the noise content in a 1Hz bandwidth. Its got units of V²/Hz.

• You end up with the total mean square noise. Notice i didnt say root mean square, so the formula look like this
• 

The equivalent noise bandwidth is the frequency span of a noise amplitude equal to the peak value and with same total integrated area.

• Shot noise
• Thermal (johnson) noise
• Flicker noise
• Burst noise
• Avalanche noise

Current flow across a potential barrier. blah blah

where n is the no. of electrons per second

• It is noise arising from the thermal fluctuations of free electron.
• 
• where k is boltzmann's constant

Spectral density function function of thermal noise is flat so it 'white'.

• An ideal noiseless component in series with a thermal voltage vn
• Or by an equivalent current generator in parallel with a resistance.

• Low frequency phenomenon exhibited by all electronic devices, but nature not understood.
• Could be due to surface traps capturing and releasing electrons in a random fashion

Si_f(f) = k(I^a/f^b) A²/Hz

i_f² = k(i^a/f^b)Δf  A² 

where a is around 0.5 and 2 , b is about  and k is an unknown constant.

Occurs in all semiconductor devices and is sudden changes in the average level of the already existing noise (noise of noise)

If  a noise signal is amplified and fed to loudspeakers, you get a constant and random crackle.

Burst noise is common in monolithic IC amps but its causes are unknown.

Produced by zener/avalanche breakdown during the reverse biasing of a pn junction. 

Carriers in the depletion region create electron/hole pairs by colliding with silicon atoms. 

Avalanche noise is much greater than shot noise, so zener diodes are avoided in low-noise designs.