Vacuum tube noise source

[matt]

This circuit uses a 2D21 thyratron as a noise source, the noise is
generated by placing a permanent magnet on the envelope and causing
"fluctuations in the dense layer of positive ions near the cathode."

http://cgs.synth.net/tube/noise.html

Does anyone know if the same effect could be generated from a
longitudinal magnetic field, rather than a transverse one? Say from
putting the tube inside a solenoid?

[Alie...@gmail.com]

On Monday, September 19, 2016 at 2:51:28 PM UTC-7, matt wrote:
> This circuit uses a 2D21 thyratron as a noise source, the noise is
> generated by placing a permanent magnet on the envelope and causing
> "fluctuations in the dense layer of positive ions near the cathode."

I don't get the physics of a DC field causing fluctuations, unless it somehow interferes with striking and extinguishing the arc cleanly.

Anyway, now that you've got me interested...

> http://cgs.synth.net/tube/noise.html
>
> Does anyone know if the same effect could be generated from a
> longitudinal magnetic field, rather than a transverse one? Say from
> putting the tube inside a solenoid?

The 1957 Sylvania datasheet has this to say about it (using a different thyratron designed for your intended use):

http://tubedata.milbert.com/sheets/137/6/6D4.pdf

"In this application, the tube is placed in a steady magnetic field which is directed through the tube at right angles to the conducting arc. Proper magnetic field alignment will be obtained if the field is directed parallel to a center-line through the tube base. This center-line passes half-way between pins numbered 3 and 4. Preferably, the north pole of the magnet should be adjacent to pin 7. A magnetic field strength of approximately 375 +/- 20% gauss should be used."

You were thinking of a solenoid co-axial to the tube so you can turn the effect on and off or modulate it? I don't know offhand how the tube elements are arranged, but I don't see a plate cap in pictures of the 2D21, so you'll have to figure out the direction of the arc inside the tube. If it's transverse to the tube axis (as implied by the magnet stuck to the side of the tube in your link), it should work. Beware of where the external part of the field goes- you might need an iron can to contain it. There's interfering with air cooling to consider, too.

The Sylvania sheet says 375 gauss, which isn't trivial, though your tube *may* require less, *or more*. You can try any handy solenoids that fit but I have the feeling you're gonna hafta wind some coils.


Mark L. Fergerson

[John Larkin]

GR did that in the 1960's.

www.ietlabs.com/pdf/Manuals/GR/1390%20Noise%20Generator.pdf

Looks like the mag field was perpendictular to the tube axis.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

[Tim Williams]

"John Larkin" <jjla...@highlandtechnology.com> wrote in message
news:3v91ubtlrt8g6mu2g...@4ax.com...

> GR did that in the 1960's.
>
> www.ietlabs.com/pdf/Manuals/GR/1390%20Noise%20Generator.pdf
>
> Looks like the mag field was perpendictular to the tube axis.

Elgenco did as well;
http://seventransistorlabs.com/Images/Gaussian_Noise1.jpg
http://seventransistorlabs.com/Images/Gaussian_Noise7.jpg
(1 thru 7, give 'em a browse)

Never found a service manual for it, so you'll have to scan the circuit
in-situ. It's 3.5 sigma and a few MHz, IIRC.

As you can see, it uses two 6D4s, which solves GR's statistical problem.

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Contract Design
Website: http://seventransistorlabs.com

[George Herold]

Tim, Any idea where the noise comes from? Amplifying the
shot noise of the tube?

George H.

[John Larkin]

On Tue, 20 Sep 2016 04:18:54 -0500, "Tim Williams"
<tiw...@seventransistorlabs.com> wrote:

>"John Larkin" <jjla...@highlandtechnology.com> wrote in message
>news:3v91ubtlrt8g6mu2g...@4ax.com...
>> GR did that in the 1960's.
>>
>> www.ietlabs.com/pdf/Manuals/GR/1390%20Noise%20Generator.pdf
>>
>> Looks like the mag field was perpendictular to the tube axis.
>
>Elgenco did as well;
>http://seventransistorlabs.com/Images/Gaussian_Noise1.jpg
>http://seventransistorlabs.com/Images/Gaussian_Noise7.jpg
>(1 thru 7, give 'em a browse)
>
>Never found a service manual for it, so you'll have to scan the circuit
>in-situ. It's 3.5 sigma and a few MHz, IIRC.
>
>As you can see, it uses two 6D4s, which solves GR's statistical problem.
>
>Tim

I think GR did a later generator, with two anti-phase zeners to
improve the histogram symmetry.

[Alan Folmsbee]

Perm mags domaim browmiam motiom, atoms move Fe

[whit3rd]

On Tuesday, September 20, 2016 at 6:36:39 AM UTC-7, George Herold wrote:
> On Tuesday, September 20, 2016 at 5:17:41 AM UTC-4, Tim Williams wrote:

> > Elgenco did as well;
> > http://seventransistorlabs.com/Images/Gaussian_Noise1.jpg
> > http://seventransistorlabs.com/Images/Gaussian_Noise7.jpg
> > (1 thru 7, give 'em a browse)

> Tim, Any idea where the noise comes from? Amplifying the
> shot noise of the tube?

It's usually a gas tube biased into the proportional region, so avalanche
multiplication does the job of a preamp. There's similar devices that
live in a waveguide, give out microwave white noise.

[boB]

Reminds me of Barkhausen noise. But the magnet has to move I think to
create the noise, then can be picked up and amplified.

Related maybe ?

boB
K7IQ

[srober...@gmail.com]

Plasmas almost always have some form of thermal motion.

Three effects,
A. The plasma is hot, so you have thermal noise radiated.
B. You have free electrons rotating in the magnetic field, so you get magnetron style effects. The fill pressure on noise tubes, both microwave band and in the Thyratrons, is very low, encouraging effects that would be suppressed in a dense plasma.
C. You have impact effects from hot electrons hitting metal.


Steve

[Robert Baer]

1) It is a gas tube, so it will inherently generate more noise than
an electron tube; 2) the ions are forced to move in loops via the
magnetic field, and so have a much higher probability of colliding with
another ion.

[Phil Hobbs]

If E and B are perpendicular, as it appears from the quotes from the
manuals, then the electrons want to move at a constant velocity
perpendicular to both. It's called "E cross B drift", and is
independent of both charge and mass, so the whole plasma drifts in the
same direction. See e.g.
<https://www2.warwick.ac.uk/fac/sci/physics/research/cfsa/people/pastmembers/peeters/teaching/lecture5.ppt>
and
<https://svs.gsfc.nasa.gov/4265>.

When E and B are inhomogeneous (as in the tube case) it gets a bit more
complicated, of course.

The magnetron motion (spiralling round *B*) should damp out pretty fast
in a gassy tube. but in any case, the magnetic field allows you to use a
much higher bias voltage without the tube firing, and so increases the
particle energy, which in turn increases the multiplication gain.

Fun--I's never looked into how noise tubes worked.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

[John Larkin]

I think the magnet near the thyratron breaks up periodic oscillations,
improves the spectrum, somehow.

[George Herold]

Yeah I think I read that in one of the manuals you posted.
I didn't know about thyratrons... avalanche's!
https://en.wikipedia.org/wiki/Thyratron

The 6d4 manual mentions the magnet, but doesn't say what it does.

http://tubedata.milbert.com/sheets/137/6/6D4.pdf

The noise doesn't go much beyond 1 MHz.

George H.

[John Larkin]

Probably too much capacitance. Gas discharge tubes have been used in
waveguides, way up in the microwaves, as reference noise sources.

I wonder if a neon bulb is a very optically-noisy source, as in above
photon shot noise.

[George Herold]

I don't know much about neon bulbs, why would you think they would be noisy?
Do you also get avalanches.. with a bunch of photons coming at the same time?

I've made an RF discharge lamp that is mostly at the shot noise limit.
(mostly 'cause when the bulbs are new you get some "aging" effects or
something that gives big spikies... well almost every thing is big compared to
shot noise.)

George H.

[Phil Hobbs]

Probably a bit over shot noise. You do get some plasma fluctuations,
and the ion current is sustained by avalanche, whose gain is stochastic.

The upper-state lifetime of the emission line provides some lowpass
filtering, of course. The main effect is that your collection system
will have an efficiency much less than unity, so I would guess that the
photodetection statistics generally dominate with low-pressure bulbs.

High-pressure arc lamps are super noisy (+- 10% power fluctuations are
common). The beam pointing is also unstable, because in some bulbs the
arc travels aimlessly across the electrodes at speeds up to 50 m/s or so.

[John Larkin]

On Wed, 21 Sep 2016 13:44:16 -0400, Phil Hobbs

Neons are very electrically noisy, so some of that probably spills
over into light.

>
>The upper-state lifetime of the emission line provides some lowpass
>filtering, of course. The main effect is that your collection system
>will have an efficiency much less than unity, so I would guess that the
>photodetection statistics generally dominate with low-pressure bulbs.
>
>High-pressure arc lamps are super noisy (+- 10% power fluctuations are
>common). The beam pointing is also unstable, because in some bulbs the
>arc travels aimlessly across the electrodes at speeds up to 50 m/s or so.
>
>

One more thing to try on a dull rainy afternoon.

--

John Larkin Highland Technology, Inc

picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

[Michael A. Terrell]

George Herold wrote:

>
> John Larkin wrote:
>>
>> I think the magnet near the thyratron breaks up periodic oscillations,
>> improves the spectrum, somehow.
>
> Yeah I think I read that in one of the manuals you posted.
> I didn't know about thyratrons... avalanche's!
> https://en.wikipedia.org/wiki/Thyratron
>
> The 6d4 manual mentions the magnet, but doesn't say what it does.
>
> http://tubedata.milbert.com/sheets/137/6/6D4.pdf
>
> The noise doesn't go much beyond 1 MHz.

Fair Radio used to sell a surplus noise jammer for aircraft. It
transmitted a broadband noise to keep the enemy planes from talking
between themselves. They used a phototube and a florescent lamp to
generate the noise. That was followed by a broadband power amp. I think
the output was about 20 watts.


--
Never piss off an Engineer!

They don't get mad.

They don't get even.

They go for over unity! ;-)

[John Larkin]

On Thu, 22 Sep 2016 14:53:09 -0400, "Michael A. Terrell"
<mike.t...@earthlink.net> wrote:

>George Herold wrote:
>>
>> John Larkin wrote:
>>>
>>> I think the magnet near the thyratron breaks up periodic oscillations,
>>> improves the spectrum, somehow.
>>
>> Yeah I think I read that in one of the manuals you posted.
>> I didn't know about thyratrons... avalanche's!
>> https://en.wikipedia.org/wiki/Thyratron
>>
>> The 6d4 manual mentions the magnet, but doesn't say what it does.
>>
>> http://tubedata.milbert.com/sheets/137/6/6D4.pdf
>>
>> The noise doesn't go much beyond 1 MHz.
>
>
> Fair Radio used to sell a surplus noise jammer for aircraft. It
>transmitted a broadband noise to keep the enemy planes from talking
>between themselves. They used a phototube and a florescent lamp to
>generate the noise. That was followed by a broadband power amp. I think
>the output was about 20 watts.

Some of the jammers used a 931A PMT, in the dark. They get real noisy
as the voltage is cranked up.

--

John Larkin Highland Technology, Inc

[Michael A. Terrell]

John Larkin wrote:
> On Thu, 22 Sep 2016 14:53:09 -0400, "Michael A. Terrell"
> <mike.t...@earthlink.net> wrote:
>
>> George Herold wrote:
>>>
>>> John Larkin wrote:
>>>>
>>>> I think the magnet near the thyratron breaks up periodic oscillations,
>>>> improves the spectrum, somehow.
>>>
>>> Yeah I think I read that in one of the manuals you posted.
>>> I didn't know about thyratrons... avalanche's!
>>> https://en.wikipedia.org/wiki/Thyratron
>>>
>>> The 6d4 manual mentions the magnet, but doesn't say what it does.
>>>
>>> http://tubedata.milbert.com/sheets/137/6/6D4.pdf
>>>
>>> The noise doesn't go much beyond 1 MHz.
>>
>>
>> Fair Radio used to sell a surplus noise jammer for aircraft. It
>> transmitted a broadband noise to keep the enemy planes from talking
>> between themselves. They used a phototube and a florescent lamp to
>> generate the noise. That was followed by a broadband power amp. I think
>> the output was about 20 watts.
>
> Some of the jammers used a 931A PMT, in the dark. They get real noisy
> as the voltage is cranked up.

The florescent lamp gave it a broader bandwidth. It had to wipe out
VHF AM communications.

[George Herold]

On Thursday, September 22, 2016 at 4:32:30 PM UTC-4, John Larkin wrote:
> On Thu, 22 Sep 2016 14:53:09 -0400, "Michael A. Terrell"
> <mike.t...@earthlink.net> wrote:
>
> >George Herold wrote:
> >>
> >> John Larkin wrote:
> >>>
> >>> I think the magnet near the thyratron breaks up periodic oscillations,
> >>> improves the spectrum, somehow.
> >>
> >> Yeah I think I read that in one of the manuals you posted.
> >> I didn't know about thyratrons... avalanche's!
> >> https://en.wikipedia.org/wiki/Thyratron
> >>
> >> The 6d4 manual mentions the magnet, but doesn't say what it does.
> >>
> >> http://tubedata.milbert.com/sheets/137/6/6D4.pdf
> >>
> >> The noise doesn't go much beyond 1 MHz.
> >
> >
> > Fair Radio used to sell a surplus noise jammer for aircraft. It
> >transmitted a broadband noise to keep the enemy planes from talking
> >between themselves. They used a phototube and a florescent lamp to
> >generate the noise. That was followed by a broadband power amp. I think
> >the output was about 20 watts.
>
> Some of the jammers used a 931A PMT, in the dark. They get real noisy
> as the voltage is cranked up.
>

That makes sense (depending on the frequency.) PMT's give ~10ns pulses,
~ few ns edges ?
(Probably spendier than the gas tube.)

George H.

[John Larkin]

Some of the jammers transmitted random microwave pulses, and the PMT
was used to make those trigger pulses; it wasn't broadband noise. I
use to buy the surplus random pulser modules, a shoe-box sized
chassis, from Fair Radio Sales, to get the 931A. The whole thing was
$5 or something like that.

--

John Larkin Highland Technology, Inc

lunatic fringe electronics