High Quality White Noise Gen

Sl>>

transistors. Then

than

current

low

random

best.

few mA

signal

from

If you are determined to be a PITA, yes.

John

Sl>

transistors. Then

better

better than

current

low

random

best.

few mA

signal

other from

Sorry about the tone. I didn't sleep well last night. ... or perhaps yes, I am determined to be a PITA. maybe with more coffee I will know which it is.

Sl>>

transistors. Then

better

better than

current

at low

mathematical random

probably best.

a few mA

signal

other from

from a

to

I've had my coffee, two cups of Peets. Snarl. Snap.

Noise amplifiers need, well, infinite headroom, so we can never get it right.

John

(Jim Sl>

noise

transistors. Then

better

better than

current

at low

mathematical random

probably best.

at a few mA

the signal

other from

from a

bigger

DC

sensitve to

see

Peet's smells rotten to me. I think they ferment it.

If the noise is to be white and not contrived, yes, there is no way to really do it. I have had a few cases where I needed noise that was more like a a uniform probability over a band of values and also a flat spectrum over a band. It can be done.

The crappy version is just the usual white noise thing feeding into a pair of diodes. The log like curve of the diodes does not that bad of a job.

(Jim Sl>>

noise

transistors. Then

better

google > >> >> >> >>> >> >> diode>

better than

high current

oscillate at low

mathematical random

probably best.

at a few mA

the signal

other from

theorem.

from a

bigger

Volt DC

sensitve to

Gaussian,

Central

voltage

the

ever

see

to

At least it hasn't passed through rodents.

John

com (Jim Sl>

ite noise

d transistors. Then

with better

google > >> >> >> >>> >> >> diode>

any better than

e, high current

ice!

cillate at low

rrents.

thematical random

s probably best.

iased at a few mA

add the signal

d the other from

theorem.

ry.

noise from a

mes bigger

sensitve to

of

aussian,

Central

mit_theorem

voltage

d of the

ave you ever

her and see

summing

ikely to

ues.

act

an

The OP wanted white noise, not necessarily gaussian noise. Your signal may be white but just exhibit two values: Va and -Va. "White" relates to the spectral power density and "gaussian" to the probability density function.

Pere

PD. Our internet provider has now dropped usenet. Sorry for posting from google groups!

On a sunny day (Fri, 11 Jun 2010 01:16:32 -0700 (PDT)) it happened oopere wrote in :

Use news.aioe.org

Thanks! Now I have real news again :)

Pere

oopere

signal

There is also eternal-september.org. Plus a few low cost commercial providers, like where i post from.