Zener-noise (current)
Bosken
I have made a noise generator based on a zener diode. But I want to know
how the noise reacts when I increase the current through de zener diode.
With PSpice a simulated the circuit and it seems when the current
increased, sigma (RMS voltage) decreases..
But when I do the practical test, it seems when the current increased,
sigma (RMS voltage) also increased..
Does anyone have any explanation for this?
Thanks in advance.
---------------------------------------
Posted through http://www.Electronics-Related.com
MooseFET
wrote:
> Hi all,
>
> I have made a noise generator based on a zener diode. But I want to know
> how the noise reacts when I increase the current through de zener diode.
>
> With PSpice a simulated the circuit and it seems when the current
> increased, sigma (RMS voltage) decreases..
>
> But when I do the practical test, it seems when the current increased,
> sigma (RMS voltage) also increased..
>
> Does anyone have any explanation for this?
In absolute terms, the current noise in the zener rises as you
increase
the current. The rate of rise is less than linear so the portion of
the current that is noise decreases.
The voltage noise on a zener ideally decreases with current. If there
is a resistance in series, the combined effect can be an increase in
noise voltage with increasing current.
Does this help?
George Herold
wrote:
What was the Zener voltage. I looked at the current noise from an
8.2V zener. I was hoping to see big current spikes. Instead I
observed that the noise increased till you reached the knee in the I-V
curve. As the voltage approached the knee I started to see little
current pulses of fixed height. The width of the pulses changed till
at the 'center' of the knee I was seeing what looked like random
telegraph noise. (random steps between two voltages.) The size and
width depended on how diode was biased. (a series resistor and
parallal capacitor) As the voltage was increased above the knee the
noise first went down... But then increased at higher currents. You
should remember that the zener impedance changes rappidly near the
knee.
George H.
Robert Baer
zener diode at any given current - from nanoamps to tens of milliamps.
Firstly, a given 1N zener will have different noise characteristics
at a specified current from vendor to vendor.
Secondly, noise and usually concurrent negative resistance will vary
all over the map for a given diode as current is varied from near zero
to (say) 20mA.
Usually, noise and negative resistance starts very close to zero
current and usually dies out in the 5-20mA region (and does not return
even up to burnout).
And there are a number of exceptions to put the lie to the word
"usually" above.
Some seem to have no visible negative resistance over some
(nondescript) current range, either abruptly starting as current is
increased, or abruptly stopping as current increased (the other "end"
being "graceful"), and others are "graceful" startup / stopping.
Variables: manufacturer, zener voltage rating, current drive.
change any one and you are not only in a different ballpark but could
be in a different city or even planet!
Bill Sloman
> Bosken wrote:
> > Hi all,
>
> > I have made a noise generator based on a zener diode. But I want to know
> > how the noise reacts when I increase the current through de zener diode.
>
> > With PSpice a simulated the circuit and it seems when the current
> > increased, sigma (RMS voltage) decreases..
>
> > But when I do the practical test, it seems when the current increased,
> > sigma (RMS voltage) also increased..
>
> > Does anyone have any explanation for this?
>
> > Thanks in advance.
>
> > ---------------------------------------
>
> One cannot use any version of Spice to even guesstimate noise of any
> zener diode at any given current - from nanoamps to tens of milliamps.
> Firstly, a given 1N zener will have different noise characteristics
> at a specified current from vendor to vendor.
> Secondly, noise and usually concurrent negative resistance will vary
> all over the map for a given diode as current is varied from near zero
> to (say) 20mA.
> Usually, noise and negative resistance starts very close to zero
> current and usually dies out in the 5-20mA region (and does not return
> even up to burnout).
> And there are a number of exceptions to put the lie to the word
> "usually" above.
> Some seem to have no visible negative resistance over some
> (nondescript) current range, either abruptly starting as current is
> increased, or abruptly stopping as current increased (the other "end"
> being "graceful"), and others are "graceful" startup / stopping.
> Variables: manufacturer, zener voltage rating, current drive.
> change any one and you are not only in a different ballpark but could
> be in a different city or even planet!
There's also the distinction between Zener didoes that work by the
Zener mechanism - breakdown voltages below about 6.2V - and the hgiher
voltage Zener diodes that work by an avalanche mechaism, which is
noisier.
8.2V zener's are avalanche devices. At low currents, there is a finite
chance that the avalanche will fail because none of the electrons
trvalling theough the - tiny - avalanche region produce a secondary
electron and the current stops dead for a few microseconds until
thermal noise injects anither electron into the junction to get the
alavanche going again.
It's all fun stuff, and Spice doesn't model any of it.
The 1997 thread on "ZENER DIODE OSCILLATION" has lots of good
information on the subject.
--
Bill Sloman, Nijmegen
George Herold
>
> - Show quoted text -
Thanks for the 'hint' Bill. I've found some of the old papers from
the 50's and will read them over. Sounds like you all had some fun
back in the summer of '97.
George H.
George Herold
>
> - Show quoted text -
This is Great! Just finished K.G. McKay "Avalanche Breakdown in
Silicon", Phys. Rev ('54) 94, 877. The discussion of Breakdown
Instabilty at the end of the paper is exactly what I observe. (Nice
to know I can see what was done before I was born at Bell labs.) John
L. I know you've measured zener voltage noise, but have you looked at
the current noise?
McKay also talks about a temperature dependence which sounds like more
potential fun!
I'd really like avalanche zeners to behave like PMT's and give shot
noise that is much larger than that for single electron events, but
still finite. I guess I have to run them below the 'knee' but at a
point where there is still some multiplication, just not infinite.
(Hmm, thermal stability might be an issue???)
George H.
Bosken
>> > > > Hi all,
>>
>> > > > I have made a noise generator based on a zener diode. But I want
> know
>> > > > how the noise reacts when I increase the current through de zener
>iode.
>>
>> > > > With PSpice a simulated the circuit and it seems when the current
>> > > > increased, sigma (RMS voltage) decreases..
>>
>> > > > But when I do the practical test, it seems when the current
>ed,
>> > > > sigma (RMS =A0voltage) also increased..
>>
>> > > > Does anyone have any explanation for this?
>>
>> > > > Thanks in advance.
>>
>
>This is Great! Just finished K.G. McKay "Avalanche Breakdown in
>Silicon", Phys. Rev ('54) 94, 877. The discussion of Breakdown
>Instabilty at the end of the paper is exactly what I observe. (Nice
>to know I can see what was done before I was born at Bell labs.) John
>L. I know you've measured zener voltage noise, but have you looked at
>the current noise?
>
>McKay also talks about a temperature dependence which sounds like more
>potential fun!
>
>I'd really like avalanche zeners to behave like PMT's and give shot
>noise that is much larger than that for single electron events, but
>still finite. I guess I have to run them below the 'knee' but at a
>point where there is still some multiplication, just not infinite.
>(Hmm, thermal stability might be an issue???)
>
>George H.
>
Hi all,
Thanks for the reaction!! I did some extra test on temperature and I see
similarities with what you all are saying.
But now I only measure the RMS Voltage so I get an idea about the Thermal
Noise.
How can I get some information about the Shot Noise, is there also a simple
way to measure this??
Thanks in advance.
Bosken
George Herold
>
> - Show quoted text -
Hi Bosken, You still haven't told us the voltage of the Zener, nor
what current you are running it at. Just because you are measuring
voltage I wouldn't call it thermal noise. A few months ago I would
have been tempted to call it shot noise, but I would have been wrong.
Let's just call it zener noise. You can measure the zener voltage
noise as you have been doing. To measure the current noise from a
zener you can put a resistor in series with the zener and measure the
voltage across it. You need to keep the value of the resistor below
the impedance of the zener at the operating point... determined from
the I-V slope at that point. You can also run the zener into an opamp
configured as a current to voltage converter... 'course then your
bandwidth will be limited by the opamp. Depending on how you are
biasing the zener you may also need to 'short out' the bias resistor
with a capacitor. (You need to supply a low impedance path for the AC
noise current.)
Have fun,
George H.
Bosken
><jens.bossaert@n_o_s_p_a_m.n_o_s_p_a_m.kahosl.be> wrote:
>> >> > > Bosken wrote:
>> >> > > > Hi all,
>>
>> >> > > > I have made a noise generator based on a zener diode. But I
want
>> > know
>> >> > > > how the noise reacts when I increase the current through de
>r
>> d=3D
>> >iode.
>>
>> >> > > > With PSpice a simulated the circuit and it seems when the
>t
>> >> > > > increased, sigma (RMS voltage) decreases..
>>
>> >> > > > But when I do the practical test, it seems when the current
>> >ed,
>> >> > > > sigma (RMS =3DA0voltage) also increased..
>>
>> >> > > > Does anyone have any explanation for this?
>>
>> >> > > > Thanks in advance.
>>
>> >> - Show quoted text -
>>
>> >Silicon", Phys. Rev ('54) 94, 877. =A0The discussion of Breakdown
>> >Instabilty at the end of the paper is exactly what I observe. =A0(Nice
>> >to know I can see what was done before I was born at Bell labs.)
>> >L. =A0I know you've measured zener voltage noise, but have you looked
at
>> >the current noise?
>>
>> >McKay also talks about a temperature dependence which sounds like more
>> >potential fun!
>>
>> >I'd really like avalanche zeners to behave like PMT's and give shot
>> >noise that is much larger than that for single electron events, but
>> >point where there is still some multiplication, just not infinite.
>> >(Hmm, thermal stability might be an issue???)
>>
>> >George H.
>>
>> Hi all,
>>
>> Thanks for the reaction!! I did some extra test on temperature and I
see
>> similarities with what you all are saying.
>>
>> But now I only measure the RMS Voltage so I get an idea about the
Thermal
>> Noise.
>>
>> How can I get some information about the Shot Noise, is there also a
>le
>> way to measure this??
>>
>> Thanks in advance.
>>
>> Bosken
>>
>> Posted throughhttp://www.Electronics-Related.com- Hide quoted text -
>>
>> - Show quoted text -
>
>Hi Bosken, You still haven't told us the voltage of the Zener, nor
>what current you are running it at. Just because you are measuring
>voltage I wouldn't call it thermal noise. A few months ago I would
>have been tempted to call it shot noise, but I would have been wrong.
>Let's just call it zener noise. You can measure the zener voltage
>noise as you have been doing. To measure the current noise from a
>zener you can put a resistor in series with the zener and measure the
>voltage across it. You need to keep the value of the resistor below
>the impedance of the zener at the operating point... determined from
>the I-V slope at that point. You can also run the zener into an opamp
>configured as a current to voltage converter... 'course then your
>bandwidth will be limited by the opamp. Depending on how you are
>biasing the zener you may also need to 'short out' the bias resistor
>with a capacitor. (You need to supply a low impedance path for the AC
>noise current.)
>
>Have fun,
>
>George H.
>
Hi George.
The zener that I use is a 6.2 voltage zenerdiode, the current will be
around 1mA..
Do you say that the contribution of Shotnoise also can be measured by
measuring the Noise Voltage? It seems strange to me because Shotnoise comes
from the current thought the potential barrier. Or is my approach to
literally? I just want to get an idea of what I am measuring...
I will test the current measurement with a resistor, thanks for the tips!