High Quality White Noise Gen

Hobbs

photo-detection.

includes

of

No, but the optput capacitor impedance continues to decline.

There are a few frequency zones:

Dc to where the base lowpass filter kicks in: 0 dB ripple attenuation.

A region where the Early thing works, roughly -50 dB.

A slope downward, beginning at the corner frequency set by Re and the output filter capacitance Cf.

Some high frequency where the attenuation is Cf/Cce

More or less.

I was interested in that second one, where simple calculations or Spice aren't necessarily predictive.

John

Hobbs

photo-detection.

includes

accuracy

orders of

is 10

Just imagine a resistor from collector to emitter. The question is, what's the value?

John

Hobbs

photo-detection.

includes

orders of

is 10

Tulgey.

I can recite "Jabberwocky" by heart, and do sometimes if the beer or wine are of sufficient quality.

This Bandit chardonnay here is actually pretty good.

John

=A0 =A0wrote:

rold

-detection.

nces includes

accuracy

six orders of

time is 10

um

than

AD's -

see

-wing

ur

bsence

hout

st

ct

My wife recently found an excuse to open some of our 2004 Gosset Polish Hill Riesling - last year she thought that the 2003 was better, but this year the 2004 does seem to come into its own.

Happily, it doesn't inspire either of us to recite poetry. If it did, this might be more appropriate than "Jabberwocky".

-- Bill Sloman, Nijmegen

Hobbs

photo-detection.

includes

accuracy

orders of

time is 10

absence

How can a wine critic know that the wine tastes of "glacial gravel, slate and shale"? Does he munch on rocks for reference?

I've seen a lot of "rock" references in wine criticism lately. "Hints of cherry and apricot mold" are passe. Wine snobs are like audiophools. Double-blind testing shows them as the delusional fatheads that they are.

There is a trend in California to buy tasty cheap wine. Makes sense to me.

The other trend is bars and restaurants that serve "draft" wine, right out of the barrel.

John

Hobbs

photo-detection.

includes

accuracy

orders of

is 10

Early effect changes the base WIDTH, thus the current gain.

Stick to sales, John :-) ...Jim Thompson

=A0wrote:

ld

etection.

es includes

curacy

ix orders of

ime is 10

han

's -

ee

ing

ence

ut

Ahh excellent. It's this one that makes it not work so well when I load the output too much. (Or do I just need a bigger cap?)

Thanks so much for spelling it all out.

George H.

=A0 =A0wrote:

rold

-detection.

nces includes

accuracy

six orders of

time is 10

um

than

AD's -

see

-wing

ur

bsence

hout

st

ct

My daughter (age 10) recited it at the last company/ holiday poetry reading. She may make a mistake or two, but she=92ll kill ya with charm.

e:

Herold

rote:

to-detection.

rences includes

g accuracy

six orders of

ad time is 10

lium

st than

SPAD's -

- see

ht-wing

your

absence

ithout

most

lect

B at

gets

Current gain is modulated by the collector-base voltage? That seems to explain it.

George H.

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

quoted text -

Hobbs

Herold

photo-detection.

references includes

accuracy

orders of

time is 10

than

SPAD's -

see

right-wing

absence

without

For the purposes of measurement and modeling, an equivalent c-e resistor value is perfectly appropriate. That's how it behaves here.

Without real numbers, all models are worthless.

John

Hobbs

Herold

photo-detection.

references includes

accuracy

orders of

time is 10

than

SPAD's -

see

right-wing

absence

without

In your dreams, dorkfish :-) ...Jim Thompson

Hobbs

photo-detection.

includes

accuracy

orders of

is 10

At modest currents, emitter dynamic resistance Re is inverse on current, actually about 25 ohms divided by emitter current in mA. So the corner frequency of Re * Cl changes with load. I don't know how the Early feedthrough changes with load current.

Capacitor ESR also forms a voltage divider with Re, so yet more ripple blasts through as Re goes down at higher currents.

Not a simple circuit!

John

signal

other from

voltage

the

spaced

spacing

The nice thing about PIN photodiodes is that there's a first-principles relationship between the DC and noise currents. That's a great calibration principle for instruments.

I have about 100 InGaAs APD/preamp modules that I got for about 75 cents each--probably 0.5 cents on the dollar.

Cheers

Phil Hobbs

Oh, that's just great! More Juvenile crime. ;-)

=A0 =A0wrote:

Herold

oto-detection.

erences includes

ng accuracy

ood six orders of

ead time is 10

elium

ust than

SPAD's -

s - see

ght-wing

your

e absence

without

most

llect

Max Coltheart is a professor of cognitive psychology who has a long- running experiemnt in which he gives a talk about wines and wine tastng to academic audiences - including a variety of this kind of wine-tasting terminology - tests his audience's performance as wine tasters and records the results. He hasn't published it yet - and I took part in one such experiment nearly thirty years ago. Informally, I've heard from him that some phrases are more useful than others, but the communication is not all that analytical

Wine-tasters are anything but fat-headed, but their language is imprecise, if evocative.

It would. You aren't great of acquiring background knowledge.

Modern wines are often made to be instantly drinkable. This doesn't mean that they can't get better with age, but many don't. If the wine- maker has a gas-chromatograph - ideally with a tandem mass- spectrometer as a detector - they can have a very clear idea of what their wine tastes and smells like, and the good ones know how to create an attractive mixture of sensations, and have a pretty good idea of how that mixture will change and develop with time,

It an art, and there are some great artist out there. Wolfgang Blass in Australia was an early pioneer of this approach, and while he isn't a great artist, I've drunk some very pleasant wines sold under his name.

-- Bill Sloman, Nijmegen

erold

ote:

o-detection.

-3-2232

ences includes

accuracy

ix orders of

d time is 10

ium

t than

PAD's -

- see

-680,

t-wing

our

absence

thout

ost

ect

at

Yup, It's all making sense. A few years ago a made a bunch of RC lowpasses and measured ESR's. I came to the conclusion that you couldn't expect anything much lower than 0.1 ohms. But I didn't try any really big Al electro's (>100's uF)

George H.

e:

add the signal

the other from

e voltage

id of the

idge

ital

t feeds a

ally spaced

ncy spacing

the sine

an Nyquist

never

. =A0 The

nd my

s that

the layout

and I

ybe you

clip the

ll in a

andwidth

tuff

too.)

or less

do shot

he

m it

=A0So

th

I find a

ith

lly

hat

f them.

hort

nts

he

ee

of

Yeah, Lafe Spietz does shot noise thermometry with tunnel junction diodes. I've got an idea for a similar circuit trick that will, balance shot noise and johnson noise to measure temperature... The shot noise gives you the circuit gain and bandwidth, which then drop out in a ratio.... looks good on paper but I still have to try it. (You run the same DC current from the shot noise through the resistor that gives you the johnson noise and it's DC voltage is then the thermal voltage. (or maybe two times it.)

George H.

t -

Can a GC tell that the vines grew in water that had been filtered through glacial gravel, slate and shale?

John

Hobbs

photo-detection.

references includes

accuracy

orders of

time is 10

than

SPAD's -

right-wing

absence

The 120 uF polymer aluminums I'm using are 25 milliohms typ. I use one to LC filter the +15 from the wart, another at the c-mult output. Each takes about a quarter of a square inch of PCB surface, not too bad.

John