Neutron Flux Modulation (By power supply ripple)

For posts specifically relating to fusor design, construction, and operation.
Post Reply
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

I’ve been eager to try this experiment since first hearing it mentioned years back in a post by Richard Hull. Now I have the tools; (good running fusor, He3 detector, etc), and the inspiration; (a recent post by Carl Willis “Another update: Carl’s Jr.” in which Carl mentions the differences between filtered and unfiltered power supplies, and my recent post, “Fusor plasma measurements” which included scope pics in which my power supply’s 120 Hz ripple was quite evident).

The setup…
Paraffin moderated He3 detector, (detector centerline to poissor = 20.7 cm).
He3 detector’s output fed into Canberra Series 35+ mca set to mcs (multi-channel scaling) mode with a dwell time of 1 ms per channel.
Fusor neutron output was around 3.5E+06 n/s, and the He3 detector was racking up just over 11,000 cps. (Whew, that’s a lot of counts)!

In the graph below one can see that the neutron output peaks every 7 to 9 ms, which coincides nicely with the period (8.333 ms) of my power supply’s 120 Hz ripple.

So, unless I have made some grave error in my test procedure it appears (as expected) that our fusor’s neutron output does indeed mimic the power supply’s input.

Questions & comments welcome.

Jon Rosenstiel
Attachments
1 ms dwell.gif
User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 7:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Carl Willis »

Hi Jon,

That was a good experiment and I will look at the data more carefully later to see what it tells us quantitatively about voltage and relative cross-section. You know what the peak-to-average ratio is for the power supply, and the peak-to-average ratio for neutron flux can be calculated pretty easily from the data in Excel. By the way, what was your grid voltage for this experiment?

Thanks for posting that.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Carl-

Grid voltage around 46~47 kV. I should probably explain a little about how I calibrated my meter. A year (or two) back I placed my SiLi x-ray detector up to the viewport (using Pb collimator with tiny a hole). At that time my meter was indicating 30 kV, but the x-ray spectrum continued on out to 32.7 keV before it dropped down to zero. At that point I decided I should recalibrate my meter so it would match what the x-ray detector was telling me. (X-ray spectrum falls to zero at 32.7 keV, meter indicates 32.7 kV). So now my meter reading should match the supply’s peak voltage. (At 30 kV, anyway)

If you (or anyone else for that matter) want the Excel file just let me know and I’ll e-mail it off.

Jon Rosenstiel
User avatar
Steven Sesselmann
Posts: 2115
Joined: Wed Aug 10, 2005 9:50 pm
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Steven Sesselmann »

Jon,

Nice work as always, you are helping all of us to get a better picture of what is going on.

Thanks..

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by DaveC »

Jon - Very interesting. Since the neutron flux drops to zero, your power supply ripple must be fairly large. Are there filter caps in the HV output? At 60 Hz, unfortunately you need some rather large capacitances to support the voltage, even with a full wave supply. A large series inductor, or best a pi section C-L-C filter would smooth things out.

Dave Cooper
User avatar
Adam Szendrey
Posts: 1333
Joined: Fri Mar 29, 2002 5:36 pm
Real name: Adam Szendrey
Location: Budapest, Hungary

Re: Neutron Flux Modulation (By power supply ripple)

Post by Adam Szendrey »

Hmm...a funny idea just popped into my mind...you know singing tesla coils..they are built by using a PWM modulated supply, usually. Now this can be a "singing fusion reactor" (though not audible) with an ampl. modulated supply ;).Imagine it, neutron communication...neat. Too bad the range would probably be...umm...not so good...And i'm not sure what the upper frequency limit would be. I suppose the x-ray flux ripples similarly?
User avatar
Steven Sesselmann
Posts: 2115
Joined: Wed Aug 10, 2005 9:50 pm
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Steven Sesselmann »

Jon,

What sheilding if any do you have around your Fusor when you are putting out an isotropic emission of 3.5 million neutrons per second?

I make this a flux of around 30 neutrons per cm^2/sec @ one meter.

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Starfire »

Jon - Is there a second component in the modulation? The primary is pseudo sine but I wonder if a sub-harmonic is also present - would be good to see the input waveform as a second graph - the synchronism and phase could also be seen.
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by DaveC »

Good point, John -

There may actually be two (or more) other modulations.
It's pretty clear there is another lower frequency modulating the peaks of Neutron counts. If the primary modulation is at the mains frequency, this one is several times slower. Some type of pressure -thermal cycle is suggestive, here.

The other seen in several of the peaks, and might just be a sporadic current burst, which pulls the output voltage down. Or it might be a higher frequency ( couple or more times the basic power supply frequency.)

Nice to have some interesting data to scrutinize. Thanks, Jon!..

Dave Cooper
Verp
Posts: 171
Joined: Fri May 02, 2003 3:27 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Verp »

Is the flux strong enough to measure the higher frequencies that fusors might resonate at that we have talked about in other threads? Is there a reasonable way to build a power supply that would be quiet enough to actually see such things?

Rod
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Dave- My x-ray power supply has no filtering. Actually, I’ve worked towards reducing the capacitance as much as possible, going as far as removing the outer shield from the x-ray cable that connects the supply to the fusor. (I was having a lot of trouble with EMI / RFI affecting my instrumentation). Shortening the cable, removing the outer shield, and installing ferrites on just about every cable in my system helped tremendously. I've been considering temporarily wiring in a filter cap and doing the experiment again, but the thought of what may happen to my instrumentation is scaring me.

Adam- Don’t see why it wouldn’t work, but as you said, the range wouldn’t be too good. What could we call it, neutron phone? I would think that the x-ray flux would ripple just the same.

Steven- Lead shielding for the x-rays, and distance (as much as possible) and time (as little as possible) for the neutrons.

John & Dave- (About the second, or third, component in the modulation) I assumed that the peaks were of different heights because of the difference between the mcs timing (1 ms) and the 120 Hz ripple period (8.33 ms) The peaks would be highest when the 1 ms mcs “window” happened to straddle one of the ripple peaks. Does that make sense? The only other way that I can think of explaining it is that it’s like two slightly different frequencies "beating" together, sometimes they reinforce each other, and sometimes they cancel each other. Then again, it’s entirely possible that there is something else going on here. Wouldn’t be the first time our fusors surprised us.

I'm not sure which is more "fun", the doing of the experiment, or the discussion of the experiment. Lots of interesting stuff.

Jon Rosenstiel
Frank Sanns
Site Admin
Posts: 1973
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Neutron Flux Modulation (By power supply ripple)

Post by Frank Sanns »

Nice work again Jon.

If you filter your output your time averaged neutron numbers will be at least 300% higher. The area under a sine curve is 0.707 for RMS vs Peak and you are running the reverse of that with IEC since cross sections increase rapidly as the voltage increases.

Now for the bad news, your wonderful graph is around 6 orders of magnitude too slow to show much of what is going on with ions and collisions in the fusor. You need something with a resolution of tens of nanonseconds or better to really be able to track what is going on in the fusor.

One thing you could do with your existing setup is to do FFT analysis of the peaks and you would have a superb signal to noise ratio for voltage vs neutron output. Actually you could just sum at a reapeating time as a function of voltage. Simple Excel file should do that in quick time.

Frank S.
User avatar
Adam Szendrey
Posts: 1333
Joined: Fri Mar 29, 2002 5:36 pm
Real name: Adam Szendrey
Location: Budapest, Hungary

Re: Neutron Flux Modulation (By power supply ripple)

Post by Adam Szendrey »

Neutron phone...neat ;). Not very practical indeed :). It would be one of those really neat and fun things to show off with. A gamma phone would be even better!
User avatar
Steven Sesselmann
Posts: 2115
Joined: Wed Aug 10, 2005 9:50 pm
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Steven Sesselmann »

Jon, just a bit of wild speculation... what if this is not power supply ripple, but rather a natural harmonic, that occurs in the poissor? If you scaled up the poissor to the size of the sun, would you end up with a wavelength of 11 years?
If so, you can't blame the power supply for that one!

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Frank-
I believe I'll stay away from filter caps... that is unless Carl forces my hand.

FFT analysis, indeed! I barely know what it means, let alone how to do it. How about I send you the Excel file, you do the analysis, then return it to me?

Jon Rosenstiel
Frank Sanns
Site Admin
Posts: 1973
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Neutron Flux Modulation (By power supply ripple)

Post by Frank Sanns »

Hi Jon,

Fast Fourier Transformations is a mathematical algorithm that are used in near real time to deconvolute waveforms. Fourier Tranforms can take a complex waveform that is the superposition of many other frequencies and alow you to work backwards to figure out what pure waveforms and amplitudes must have been present to produce the complex waveform. Send me the file and I will return the result of fusions as a function of voltage if you tell me the peak voltage on the run. Please also send a file of the HV waveform if you have it. No need to sync the two. Raw data is fine.

Frank S.
Frank Sanns
Site Admin
Posts: 1973
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Neutron Flux Modulation (By power supply ripple)

Post by Frank Sanns »

And the answer is attached as a jpeg.

For 2020 data points this is the graph of voltage vs neutron production (fusion rate). I did not have the actual waveform of your power supply so there is some error in that. The frequency for the 2020 data points was 120.02 +/ 0.13 Hz.

This raw data graph clearly shows the shape of the voltage vs fusion rate and the noise (width of the curve). There is power in statistical analysis. Especially with repeat measurements like this one. Replicate data is ALWAYS good and 2020 data points gives great confidence in the result. The really cool part is that the entire experiment that generated 2020 data points occured in 16.8 seconds!.

Frank S.
Attachments
FusionRate.jpg
User avatar
Richard Hull
Moderator
Posts: 14038
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Neutron Flux Modulation (By power supply ripple)

Post by Richard Hull »

It appears I have come late to the party. I have been in NC with Electric Spacecraft for a board and Trustee meeting ove the last five days.

I am 100% positive Jon's results are power supply sine response related. You can forget some sort of natural period other than the natural period of the full wave rectified line voltage.

Caps are not what is needed to obtain stable operation and elimenate the sine. (boost output)

What is needed is a 3kw 50khz HV supply with stack. In this fashion only would the instruments calm down with continuous power at a fixed DC level.

One might find that a rigidly fixed DC level would be a whole new beast or regime of operation with the old voltages no longer attainable due to continuous power application with no dead time.

The big filter caps on a 60hz mains supply is NOT the way to go as Jon wisely senses and suggests. Old hands at fusor operation who use mains rectified supplies are fully aware of Jon's logic and reasoning. Actually this reasoning and full discussion goes way back to the old "Songs" site and postings.

If anyone is looking for fixed, flawless DC to place on a big neutron production device, they need to perpared to peel off the bucks or be very talented in the HV switcher electronics and magnetic design department and very good at electronic assembly.

Good work Jon!

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
User avatar
Steven Sesselmann
Posts: 2115
Joined: Wed Aug 10, 2005 9:50 pm
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Steven Sesselmann »

Just for everyones amusement..

Let us assume that the poissor (little star in Jon's fusor) is 1 cm in diameter and that it has a pulse rate of 7 milliseconds.

If the pulse is a natural harmonic, we assume that scaling up the size of the poissor in Jon's fusor will increase the wave length of the frequency.

What would the frequency be if Jon scaled up his poissor to the size if the sun?

The sun's diameter is around 1.3x10^11 cm or 1.3x10^11 times bigger.

The hypothetical frequency in seconds would then be..

(7 x 10^-3) x (1.3 x 10^11) = 9.1 x 10^8 seconds

As there are 3.1x10^7 seconds in one year, the expected frequency would be 29 years.

Unless I have made some silly mistake above somewhere, this is remarkably close to the 11 year solar cycle that we observe.

Give or take a millisecond on Jon's poissor frequency and it would be spot on.

Suggestion for an experiment would be to do the same experiment on a much bigger fusor and look for a lower frequency.

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by DaveC »

I think we will get farther with understanding the electrical behavior of the fusor, by reviewing the V-I curve for a standard glow discharge. A good example is the simple neon lamp.

You cannot run a neon lamp from a fixed voltage supply, without a ballast (or current limiting ) resistor. Placed across a sufficiently powerful constant voltage supply, the lamp will self destruct in milliseconds, or less. The reason is found in the nearly vertical portion of the V-I curve. Without current limitation, the plasma forms with a decreasing resistance as the current increases, holding the voltage across to a nearly constant value.

With unlimited current, lamp almost instantly passes from a glow discharge to a spark or arc discharge.

The fusor is just a large Deuterium Glow Discharge lamp... that happens to produce neutrons in the process, if you run it at a high enough voltage.

Jon - I understand about the cable capacitance. The voltage instability and EMI/ RFI problem can be intensified by the stored energy in the cable capacitance unless... a current limiting resistor is present. One common value is about 1 ohm per volt. of output voltage... so a 40 kV operating level would need about 40K ohms. This will limit major surges to 1 amp peak . At 40 mA, the resistor would dissipate about 64 watts, which is not trivial. The terminal voltage would also sag about 1600 volts at full load.

Given the unstable nature of the high voltage gas discharge in the Fusor, it would seem useful to also employ some reactive current limiting during transients. So, and L-R limiter might be the quietest. But then the PI section filter, with a damping resistor might also sidestep the nose.

I didn't realize you were using the MCS for data acquisition. .Clearly, this will provide some sort of sampling- like effect. With a storage or digital scope, you should be able to see if it is and artifact or real.

Interesting data.

Dave Cooper
User avatar
Richard Hull
Moderator
Posts: 14038
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Neutron Flux Modulation (By power supply ripple)

Post by Richard Hull »

The secret to operation with a filtered supply be it 50khz or main frequency at stable voltage is a "bungie cord" current limiter or LR filter. The LR filter is just not in the cards for most here so the obligatory ballast resistor is the usual solution. More relief is granted by unfiltered full wave operation.

As I note in my original post two prior to this, all this was hashed out in the early "Songs" stuff. The second person to do fusion with a fusor was my friend Scott Little at EathTech. He used a wonderful combo LR circuit which is quite funtional for the, then, very low level operation. He used a 15kv neon sign transformer secondary in series with his HV supply These things have a hundred henry inductance and about 27-40k ohms of resistance in the miles of #38 or #40 wire used. This is just fine for those trying to push 20kv into a fusor and perhaps even 25kv if you don't need the transformer afterwards.

Scott, realizing that the secondary was center tapped to case, wisely sat the transformer out of range quite high and out of reach on a shelf atop a large chunk of plexiglass about 1" thick. Naturally, touching the body of the transformer would have proven instantly fatal. But, his lashup was temporary and designed to prove the fusor concept to his doubtful employers.

The neon transformer worked amazingly, if not frightfully, well in smoothing out the cantankerous plasma system operation of the fusor while using his "stiff" and well regulated HV supply. Alas, the world is not awash in 100 henry 30ma 60,000 volt insulated chokes and so a well crafted and successful, PROPER type of L-R filter is just not realistic with finished goods.

My fusor works just fine, though not in the rarified atmosphere where Jon's works, with a simple 50k ohm 300 watt ohmite wire wound resistor in oil. It is a little larger than a stick of dynamite and is also a very rare find at a hamfest, but was only $1.00. I doubt if one is currently manufactured and on a dealer's shelf.

As Dave notes, the resistor disappates. Mine, at 28kv applied and with 10 ma flowing to the fusor, sucks the fusor voltage down to a still respectable 27.5 kv on the fusor input and disappates a mere 5 watts while letting the fusor have 275 watts of juice. Not a bad tradeoff for relatively stable operation.

It is actually the combination of a somewhat weak and limited HV transformer, (X-ray), working unfiltered, full wave rectification and the resistor that allow my fusor, and I am sure, Jon's, to operate at all!

I am not in the mood to wind a 100 henry choke and insulate it to elevated voltages just now and so the ideal L-R filter will have to be a dream for the moment.

Until you have ridden the bucking bronco of a fusor with the gas valve, vacuum gate valve and variac wheel, demanding constant attention, trying to not only get good operation, but also to avoid destruction of the device's grid, you can't really appreciate the dynamics of the system that usually makes you want to blame the supply for its seeming fitfulness.

The real culprit here is the unrully physics of electrical plasma dynamics in a changing or less than stable enviroment. This and other issues have held fusion in abayance all these years. Other normally unencountered laws of physics have conspired to make thermal fusion machines aluringly attractive, but unworkable. Actually the physics laws appear to queue up in growing numbers to thwart man made fusion such that if we ever do thermal fusion, it might be so prohibtively expensive and unprofitable as to make it the carrot we actually got hold of, but not before it rotted on the string.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Starfire »

A very effective HV choke can be constructed using HV insulated wire ( Co-axial transmitter aerial cable core? ) simple wound on a toroidal core. The larger the core ( from an old transformer ) the better. Easy to get the Henry's high.
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

To smooth things out I run both a current limiting resistor (40K ohms, under oil) and a ~13 mHy choke in series with my x-ray supply's primary winding. For safety reasons I placed the 40K resistance in my hv supply cabinet. To further suppress EMI / RFI I’ve placed 5 ferrite doughnuts over the hv cable. (At the fusor end).

Jon Rosenstiel
Wilfried Heil
Posts: 590
Joined: Sat Dec 24, 2005 7:31 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Wilfried Heil »

Jon - I assume that you are running the 40 kohm resistor on the secondary of the xray transformer, rather than on the primary side. Is the choke connected on the primary, in order to block RF ripple between the fusor and the mains power input, or is this on the secondary as well? Would a smaller resistor on the primary side (e.g. a lamp/iron/toaster) or a ballast choke work equally well? If the supply just needs to be current limited, it should not matter where the limiting resistance is. This would make it simpler because a ballast on the primary does not need to be a high voltage device.
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by DaveC »

The issue of where to place the "ballasting" impedance in a plasma device is important.

In principle, it can be placed anywhere, as will be obvious by considering an equivalent circuit of the HV transformer. In theory, it is just an energy conversion device transforming voltages and currents in inverse relationships, that keep the transmitted power constant.... in theory.

A ballasting impedance in the primary side, requires the transformer to handle not only the load impedance (the fusor's plasma) but also the volt-amperes of the ballasting impedance.

While the real ( or resistive) part of this ballasting impedance does not signifcantly affect the HV transformer heating due to loading, the imaginary part (the inductance) can and usually does.

In the secondary, an inductive impedance (R + L) will produce not only a voltage drop from the resistance, but a voltage boost when the current drops. The L di/dt term can be quite large and, in the secondary circuit, performs a very useful function. It helps keep the plasma current stable. It also supplies stored energy to the output stray capacitance of the HV circuit, minimizing the amount of stored energy that has to pass through the transformer. This becomes increasingly important as the operating frequency of the system is increased...as in the case of switching power supplies.

Plasma, being a movement of ions, is essentially a gas flow. The flow is driven by the voltage difference between the fusor electrodes and consists of many ion (charged particle) paths in parallel. The parallel nature and the fact that ions move along field lines, generally, and are affected by local magnetic fields, means that the plasma current is inherently unstable.

Its magnitude will vary widely in very short intervals of time. .

If the plasma stream is interrupted, because of a sufficient reduction in current density, the potential across that path increases instantly, diverting other nearby plasma streams. The wavy paths increase the overall length of travel resulting in a reduction in ion acceleration over the wavy path, in comparison to the acceleration in an undeviated path.

By use of an inductive element in series with the plasma current, whenever a rapid reduction in current occurs, a correspondingly rapid increase in voltage is generated by the collapse of magnetic field in the inductor. This results in an instantaneous increase in plasma drive, which minimizes, but does not eliminate, the instabilities.

A fusor, having a converging plasma configuration, might work more stably if ion sources were divided into clusters that were each limited by a separate ballasing impedance. (this is just a speculation of mine at this point) But such a device is much more complicated.


In summary, ANY plasma device requires some sort of inductive stabilizer, along with a basic current limiter in the event of an outright anode to cathode short.

A resistive ballast works, but an inductive ballast works much better.

Dave Cooper
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Starfire »

A resistive ballast with negative coefficient is best - the humble domestic incandescent lamp will work well as the resistance will change as the lamp heats ( or use a thermistor, but hard to get at these voltages and currents ) - build a ballast plane of lamps - the mains voltage here is 250v { on a good day } so I use four lamps per kilovolt ( 15 watt lamps will give 60ma at full voltage or 25 watt for 100ma - a kilowatt of light on 10kv :) I also use them for HV cap bleeders
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Wilfried-
Sorry, wasn't very clear on a couple of points. The 40K resistor is in the secondary circuit and the choke (actually a ballast choke) is in the primary (mains) circuit. I found the choke in a surplus store. I don’t know its original purpose, but the seller did tell me that it came from a decommissioned TV transmitter. Printed on the choke is: “15 MHY, 16 AMP”. What is really cool about it is that it’s fairly easy to unbolt the “I” core and change the gap, thereby changing system response. I suspect it would be quite the challenge controlling a fusor without a ballast choke somewhere in the circuit.

Jon Rosenstiel
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Great post, Dave. Thanks.

About inductive ballast in the secondary.... Richard mentions needing inductance values in the 100's of Henrys, but I don't follow. Here’s why: Right now my system is fairly controllable with about 15 mHy of inductive ballast in the primary circuit. My seat of the pants reasoning is telling me that if 15 mHy works well in the primary, something considerably less would work equally well in the secondary circuit. Or am I thinking about this bass-ackwards?

Below comments added 11-2-06, 1310 hr.

Ok, I see the error in my thinking.... for the inductive ballast to function properly in the secondary circuit it will have to have a high inductive reactance, (~40 Kohm) which means lots of inductance. (Over 100 Henries)

Jon Rosenstiel
JohnCuthbert
Posts: 339
Joined: Mon Aug 19, 2002 4:30 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by JohnCuthbert »

Not really my field, but I'm pretty sure the HT side is also the high impedance side so to make a difference you nead a bigger reactance ie a bigger inductor. With the added fun of trying to insulate an HT inductor, I think the primary side inductive balast is a good idea.
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Right you are, John. You posted as I was in the process of editing (correcting) my post. Thanks

Jon Rosenstiel
User avatar
Richard Hull
Moderator
Posts: 14038
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Neutron Flux Modulation (By power supply ripple)

Post by Richard Hull »

There is no set rule on which side the inductor should go on as it is taken on a case by case basis. There are a boat load of questions to be asked to decide. What is the load impedance of you transformer?... Is it magnetically shunted in any way? What are the dynamics of the load? Is this a current and voltage limited supply? How fast does the inductive protection have to act? What kind of slower transients do you wish covered, if any? Will a resistor be used in concert with the inductor? Are there any large energy storage capacitors or stray capacitances present? Primary inductance usually works best in mult-kilowatt pulse situations.

Scott Little needed high side inductance with his setup as it was the quickest fix at hand.

The only optimum solution is a fast HV switcher with super fast acting current and voltage limit sets. Quite a tall order for the experimenter.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by DaveC »

Just to add a few more thoughts to the discussion. As we noted earlier, you can add the ballasting impedance almost anywhere in the drive circuit. The primary side will require a high(er) current, low(er) voltage component and the secondary side the reverse - high voltage, low current. In either case the energy storage requirement is more or less the same.... 1/2 L i^2.

There will be some particular volts/turn value for whatever you decide on for the inductor, that will not change whether it is in the primary or secondary circuit. Since inductance L is proportional to N^2, the increased inductance of the secondary side inductor just compensates for the reduced current (squared). So all remains the same.

At issue is how much voltage boost or buck you need to stabilize your particular plasma (or fusor). That will depend on current level, operating voltage, gas density (or fill pressure) and temperature...at least!.

These are all dynamic variables, with time constants that are will be in the microseconds or faster, out in the ion cloud itself, all the way down to 100's of milliseconds in the shell, and electrode temperature changes. And most ion gages struggle to respond in tenths of a second, usually. The actual voltage and current signals will probably be the fasted responding in the setup.

Since the inductor's induced voltage depends on the rate of change of the current, then both instantaneous current level and it's rate of change will set the instantaneous voltage the inductor will produce.in response to a current change at that point in time.

Thus for small rapid changes, you won't need a huge inductor, but you will need to guess about how stabilizing voltage change is needed. I would guess about 50% variation, minimum, of whatever voltage level you plan to operate. The required inductance then goes up with operating voltage and down, with operating current. ( More current, more Ldi/dt generated voltage... and etc.)

It IS a case by case anyalysis, as Richard notes, and the best arrangement for one may not be ideal for another.

It's a great exercise in understanding the flow of energy around the fusor circuit, which will give a lot of insight to the appropriate measurement techniques.

Dave Cooper
MCL
Posts: 52
Joined: Sun Mar 12, 2006 7:27 pm
Real name: Mark Langdon
Contact:

Re: Neutron Flux Modulation (By power supply ripple)

Post by MCL »

A lot of useful information here. I use a long array of high-voltage
resistors to current-limit the output of my power-supply. This
seems to smooth things out, and is an absolutely necessity for
my config, as I have an old nitrogen laser DC powersupply that is
used to energize the fusor. The DC seems very smooth, but I
can't prevent the fusor from sometimes just pinning my
current-meter, as the whole fusor device will sometimes just
"short-out", when the voltage-levels are pushed up as needed.
As Richard points out, one must adjust and balance the gas flow
level, the applied voltage, the vacuum system operation and the
device radiated output. But perhaps you can make the
power-supply ripple work to your advantage. As you now
have modulated output, if you can devise another method that
also modulates the system neutron output (a mass-flow controller
to modulate the fuel gas flow?, or a magnitron or ion-gun or
RF-field applied to the grid?) - maybe it might be possible to
adjust the 2nd modulated parameter to constructively interfere
with the modulated activity that you currently show - the idea
being to boost momentarily to a higher, but unstainable, level of
neutron production.
But thanx for posting the data, and stimulating the discussion.
- Mark Langdon
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Mark- Hmmm... I like the idea of modulating the gas flow, or maybe the chamber pressure.
While operating my fusor I've often noticed that a small, quick, increase in pressure is sometimes accompanied by a spike in the neutron count. More to think about.

Jon Rosenstiel
User avatar
Richard Hull
Moderator
Posts: 14038
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Neutron Flux Modulation (By power supply ripple)

Post by Richard Hull »

Interesting, indeed. I, too, have noted a pressure change to result in a neutron burst. Abrupt current changes usually have an opposite effect! (reduced neutrons)

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.
Frank Sanns
Site Admin
Posts: 1973
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Neutron Flux Modulation (By power supply ripple)

Post by Frank Sanns »

Current is to electrons as neutrons are to ice.

Frank S.
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: Neutron Flux Modulation (By power supply ripple)

Post by Starfire »

Make a ToDo list! Jon - lest a gem is forgotten and not explored.
As usual, work and observation of the highest calibre. Much respected.
Jon Rosenstiel
Posts: 1477
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Neutron Flux Modulation (By power supply ripple)

Post by Jon Rosenstiel »

Attached pic: Power supply ripple vs. neutron output. Upper (channel 1) trace is my fusor's hv input waveform. (From capacitive pickup on hv cable). Lower (channel 2) trace are NIM positive logic pulses from a Tennelec 440 single channel analyzer. The Tennelec sca outputs one logic pulse for each neutron detected by a paraffin moderated He3 detector. Note the absence of logic pulses (indicating absence of neutrons) as the power supply ripple goes positive.

I tried to calibrate the capacitive pickup, but my system is very non-linear, particularly above 30 kV. A rough estimate (plus / minus 10%) of the input waveform voltages are as follows. DC level = negative 43 kV. Positive peak = negative 20 kV. Negative peak (at reticule centerline) = negative 50 kV.

Additional info:

Fusor operated at 40 kV, 20 mA.
He3 detector was positioned as close as possible to fusor, and was clicking off ~ 5000 n/s.
The scope’s persistence was set to 1 second in order to better display the logic pulses.
It was somewhat difficult to capture a clean trace; instability of the fusor’s plasma creates huge, ugly, nasty, 50+ V peak to peak oscillations of around 40 mHz at the scope’s input.

Jon Rosenstiel
Attachments
40 kV, 20 mA 001a.jpg
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”