1: 1/f noise and ttl - levels 2: triode speed: semiconduct

[Arne Rosenfeldt]

1: 1/f noise and ttl - levels
since
integral( 1/f, x..intfinity) -> infitiny for x->0
a transistor has infinite noise power.
You can only use it for high frequency above 1kHz,
absolute DC TTL levels are nonsenese

So where is the nonsense here?



2: triode speed: semiconductor vs vacuum
who is faster?
semiconductor has higher carrier density
vacuum has higher carrier mobility

or what is the point?

Expecially if I have a high impendance source, wouldn't vacuum be better?

 

[Dimitrij Klingbeil]

Arne Rosenfeldt wrote:

1: 1/f noise and ttl - levels
since
integral( 1/f, x..intfinity) -> infitiny for x->0
a transistor has infinite noise power.
You can only use it for high frequency above 1kHz,
absolute DC TTL levels are nonsenese

So where is the nonsense here?

Somehow I don't get what you mean. Are you sure that the noise ratio is

really ~~ 1/f. If so, in which case? Do you have any schematics or more
detailed descriptions where this shall apply? I can't imagine any case.
Absolute DC TTL levels are far from nonsense (if them were so, you would
hardly be reading this on-screen), and (excuse my limited experience) I've
never seen noise ratios rising significantly with lower frequencies. In any
case, could you please quote an example from where your formula is from.

2: triode speed: semiconductor vs vacuum
who is faster?
Semiconductor devices are usually faster because of their smaller size and

lower internal capacity, however there are still special vacuum devices for
very high frequencies. It depends on where you want to use the stuff and
what it is supposed to do. Besides, try to imagine a PC processor with
vacuum tubes . In any case, a more precise description of what use of the
corresponding devices you mean could probably help.

semiconductor has higher carrier density
vacuum has higher carrier mobility
Right.

or what is the point?

Expecially if I have a high impendance source, wouldn't vacuum be better?
What sort of source? Modern Op-Amps do have pretty high impedances, vacuum

tubes need a resistor at the input (for discharge), which cannot have an
infinitely high resistance, so in many cases they shall mostly equal. A
major advantage of vacuum tubes is that they can sustain much higher
overvoltages in an 'unsafe' environment but as operating voltages tend to
get smaller these days, you can safely forget about vacuum tubes unless you
are to handle microwave, radar and this like sorts of things. In nuclear
physics, there are still lots of uses for vacuum devices, but these are
sort of special. Also if the source impedance is really exceptionally high
(in the GigaOhm range or above), there may be a need to use vacuum-tube
amplifiers, but things like that tend to be very special (e.g. vacuum
current multipliers), you won't run across them very often.

P.S. In case my School-English is incomprehensible, feel free to say so,
I'll try to express myself differently. It looks that you are from Uni
Muenster, so you can write me an e-mail in German as well. Here, at TU
Berlin, we do speak German, though it sometimes sounds a bit 'englished'.

Dimitrij

 

[Kevin Aylward]

Dimitrij Klingbeil wrote:

Arne Rosenfeldt wrote:

1: 1/f noise and ttl - levels
since
integral( 1/f, x..intfinity) -> infitiny for x->0

Indeed. This is the usual formula.

a transistor has infinite noise power.

Since a transistor doesn't usually start glowing white hot, where do you
think your error is?

You can only use it for high frequency above 1kHz,
absolute DC TTL levels are nonsenese

Oh. You don't say? Strange. One can slap a TTL device on the bench and
measure quite low "DC" noise levels. Why do you think this is?

So where is the nonsense here?

Somehow I don't get what you mean. Are you sure that the noise ratio
is really ~~ 1/f. If so, in which case? Do you have any schematics or
more detailed descriptions where this shall apply? I can't imagine
any case. Absolute DC TTL levels are far from nonsense (if them were
so, you would hardly be reading this on-screen), and (excuse my
limited experience) I've never seen noise ratios rising significantly
with lower frequencies. In any case, could you please quote an
example from where your formula is from.

The formula is essentially, correct. However, one has to know how to
apply formulas.

Expecially if I have a high impendance source, wouldn't vacuum be
better?
What sort of source? Modern Op-Amps do have pretty high impedances,
vacuum tubes need a resistor at the input (for discharge), which
cannot have an infinitely high resistance, so in many cases they
shall mostly equal. A major advantage of vacuum tubes is that they
can sustain much higher overvoltages in an 'unsafe' environment but
as operating voltages tend to get smaller these days, you can safely
forget about vacuum tubes unless you are to handle microwave, radar
and this like sorts of things. In nuclear physics, there are still
lots of uses for vacuum devices, but these are sort of special. Also
if the source impedance is really exceptionally high (in the GigaOhm
range or above), there may be a need to use vacuum-tube amplifiers,

Note that, tubes can have significant grid leakage.

Kevin Aylward
salesEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

 

[Roy McCammon]

Arne Rosenfeldt wrote:

1: 1/f noise and ttl - levels
since
integral( 1/f, x..intfinity) -> infitiny for x->0
a transistor has infinite noise power.
You can only use it for high frequency above 1kHz,
absolute DC TTL levels are nonsenese

So where is the nonsense here?

The nonsense is that you never actually use zero
frequency. If your device runs for a year continuously,
then take the lower frequency as 1/year. If it runs
continuously for 100 years then the lower frequency
is .01/year. Even for these low but non zero frequencies,
the total rms fluctuation will be sufficiently low to
expect proper operation.




--
local optimization seldom leads to global optimization

my e-mail address is: rb <my last name> AT ieee DOT org