FICS II Planning and construction

For posts specifically relating to fusor design, construction, and operation.
User avatar
Richard Hull
Site Admin
Posts: 10842
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: FICS II Planning and construction

Post by Richard Hull » Wed Jan 11, 2017 8:20 am

I would never laugh at any new experiment win or lose. A physical effort in experiment is the highest achievement anyone can attain. Sure, a win is nice, but we learn even in a loss.

I watched and paid close attention to the video...

I would not give up just yet. Your problem in the motorboating within the detector, (massive RF pulse into the detector), tells us you are operating a classic relaxation oscillator much as the now well known NE-2 bulb oscillator made with a NE-2 lamp and a capacitor and resistor which sets the rate of oscillation. The NE-2 gas filled bulb is non-linear and has considerable hysteresis in its on and off points. This is exactly what you are experiencing.

Two things. I assume your supply is really beefy and can easily supply a full 30 milliamps @ full voltage. If not, there is you relaxation oscillator and it will not go away until a really beefy supply is obtained. The next thing is if you are using a voltage multiplier, then that is equivalent to a weak supply unless you are running a very powerful high frequency oscillator into same.
Finally, if you do indeed have a super beefy supply, your ballast resistor may be too large.

The upshot is that assuming a sterling, ample supply at full applied voltage with good current delivery at same, there is always a point where you will and must have a stable plasma in a rarified gas. Finding that point is a matter good gear and practiced artifice by the operator. No two systems will ever function alike unless they are perfect replicas of one another.

99% of all oscillation issues involve a charged capacitor in the circuit where a large enough resistor or a weak supply, (high impedance supply= large resistor), will strike the arc plasma in the gas and imediately the supply will buckle under load. As the capacitor discharges, the voltage plunges, the plasma turns off at reduced voltage, (gas load disappears), the capacitor recharges or the weak supply recovers and the gas plasma fires again and the cycle goes on and on. Classic relaxation oscillator.

All the best in getting a stable plasma. Be it a common old fusor or and advanced FICS design, learning how to operate a plasma gas load system is an art within a hard science background.

Working on a knife's edge........

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: 2096
Joined: Thu Aug 11, 2005 1:50 am
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: FICS II Planning and construction

Post by Steven Sesselmann » Wed Jan 11, 2017 10:51 pm

Richard,
Your problem in the motorboating within the detector, (massive RF pulse into the detector), tells us you are operating a classic relaxation oscillator
Thanks for feedback and for your diagnosis, I think you are spot on.
Two things. I assume your supply is really beefy and can easily supply a full 30 milliamps @ full voltage.
Not quite a beefy 30 mA, but no chicken either, it's a Glassman 125kV 16 Ma supply, but for this experiment I am limiting it to 62 kV but still with the full 16 mA.
Finally, if you do indeed have a super beefy supply, your ballast resistor may be too large.
This is a valid point, I am using a big ceramic 5 M Ohm resistor as a ballast. The capacitance of the cathode is in the order of 10 pF so with 5 Mohm ballast this works out to a time constant of 50 µs. My calculations estimate the ions moving at 2500 km/s which means the ions travel around 125 m in 50 µs.(more than 200 times longer than the beam return path). A ballast resistor of 10K would cut the time constant down to 200 ns., in which the ions would travel around 0.5 meter. This works out to about a 5 MHz oscillator. What are your thoughts on this?

For the moment I'm just wearing my thinking cap and trying to avoid a complete rebuild.

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG

User avatar
Richard Hull
Site Admin
Posts: 10842
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: FICS II Planning and construction

Post by Richard Hull » Thu Jan 12, 2017 12:11 am

I am now concerned the Glassman is the issue here. 16 ma is nothing at even 30kv and above. At 62kv when the arc strikes you are in need of maybe 50-100ma. The supply auto cuts out or buckles. It doesn't matter which occurs, one or the other will, and apparently, from experiment, does occur!

The 4 meg resistor sounds way too high and 50-100k would be more like it. In the old days unquenched GM tubes used 10 and 100 meg reistors to quench the geiger tube when it fired. I think the 4 meg is doing just that... Still, the supply is also part of the problem.

Try this to check things out. 1. Put in a smaller resistor as noted above. 2. Put in some gas to about 4-6 microns. slowly bring up the supply. If it doesn't strike by 10-15kv, take it back down to zero volts and keep adding gas and keep bringing the voltage back up. Ideally, at some gas pressure you should get a stable gas plasma. Now it is a matter of delicately bringing up the voltage and lowering the gas pressure to keep the plasma lit. Again, it is an art.

Never ever just choose a voltage like 62kv and then bring the gas up. This is doomed to failure and an effective buckling of the supply. I would tend to think 16 ma is nothing in the size device you are working with.

Virginia Commonwealth University, here in town, had a 30kv, 20ma Spellman supply and while they did fusion they could not keep the plasma lit for more that a minute before a burp would auto-shut down the supply forcing them to inch-worm the voltage up again into the fusion region as the current approached the trip point. They have since purchased a much more beefy Spellman and are doing fine. They were working a stock 6" diameter spherical fusor.

You'll not know a thing about FICS success or failure until you get a stable plasma lit in the device and at a voltage where fusion is possible and detectable with your counter or BTI bubble detectors.

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: 2096
Joined: Thu Aug 11, 2005 1:50 am
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: FICS II Planning and construction

Post by Steven Sesselmann » Thu Jan 12, 2017 1:22 am

Richard,

Thanks for advice, I have a big 10K ballast resistor in stock, so I will give that a go first, and go for the ignition first followed by pressure reduction method.

Hopefully get a chance to try this on the weekend, so I will let you know how it goes.

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG

John Futter
Posts: 1291
Joined: Thu Apr 22, 2004 2:29 am
Real name:
Contact:

Re: FICS II Planning and construction

Post by John Futter » Thu Jan 12, 2017 3:10 am

Steven
I do not see any potential resistors on your aluminium plates setting the voltage gradient.
Without them you will have little or no focusing ie beaming. see attached pic the resistors are 50meg HV types from Welwyn
DSCN0916.JPG

User avatar
Steven Sesselmann
Posts: 2096
Joined: Thu Aug 11, 2005 1:50 am
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: FICS II Planning and construction

Post by Steven Sesselmann » Thu Jan 12, 2017 3:36 am

John,

Well observed, I actually had these on my last build, but decided to omit them this time because I didn't think they served much purpose.

My thoughts were that random ion collisions would charge the discs potentials gradually, but that might not be the case.

Relatively easy to fit some resistors all the way down the stack, I can drill small 1.5 mm holes into the edge of the 5 mm disks and push in a copper wire.

In the earlier FICS MK-1 build I used 68M resistors but they got quite hot.

Steven
FICS-1.jpg
FICS MK1
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG

User avatar
Steven Sesselmann
Posts: 2096
Joined: Thu Aug 11, 2005 1:50 am
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: FICS II Planning and construction

Post by Steven Sesselmann » Sat Mar 04, 2017 10:26 am

Hi All,

Time to post a FICS progress report again, and it's great to have some good news for a change. Since my last post I have implemented the following changes.

Voltage Divider Across the Potential Discs
Taking John Futter's advise, I drilled and tapped small 2.5 mm holes into the sides if the aluminium discs and installed 68 M ohm resistors all the way to ground, much the same way as a PMT voltage divider. The purpose of these resistors is to ensure equal potential between discs, which in turn helps to focus the beam.

Capacitors on last two stages
Richard was of the opinion that my 16 mA power supply might be a bit low on power, so I borrowed another trick from PMT dividers and added some high voltage 10n capacitors across the last 3 stages of the accelerator tubes. My thought here is that it might just store and release a little more energy at times of high current draw.

Large ring magnets top and bottom
To further help ionisation I bought a couple of giant ferrite ring magnets and mounted them top and bottom of the accelerator stack, hoping that the field will extend through the aluminium and into the top of the chamber. Maybe it helps, I don't know.

Invented new flux capacitor
After months of frustration with the floating gas cylinder and gas handling system, I came up with the idea for the flux capacitor, a plasma back flow prevention device that works like an airlock with diffusion pump oil. So I completely removed the old gas handling system and mounted the gas cylinder and the mass flow controller on the rig at ground potential and connected it via the flux capacitor to the cathode with Swagelok hose barbs and 6mm high vacuum hose.

Rewired NI-6008-USB-DAQ
To eliminate any chance of bad connections I rewired all sensors using a boxed telephone style punch down rack and cat5 cables going back to the DAQ, which now sits at the computer end, far away from the high voltage.

First Confirmed Neutrons
The above changes have improved matters and I can now reliably run the machine from my Labview virtual instrument panel without anything crashing, and yes we have neutrons, but I am not yet where I want to be. The flux capacitor breaks down at voltages over 30 kV, it usually starts with a bubble and then a channel opens up through the silicone oil with a beam of plasma shining through the oil. Although this looks incredibly cool it draws a lot of current and makes the oil go brown, so I have now ordered a new flux capacitor from my glass blower which is longer and should hold off the full 60 kV.

Using Labview to detect neutrons
I worked out a crude method of using Labview to detect neutrons by using the sound card as a data acquisition device, but it is not so easy to pick the pulses reliably based on a threshold only, so I use PRA in parallel, but the two counters don't agree. More work to do here.

Practice runs
Last weekend and today I did some practice runs to learn how to drive it, and the typical pattern I am seeing now is that the machine needs a but of warming up before the neutron counter starts popping. I still have a lot of voltage fluctuations as you can see on the attached chart, and this is due to little bursts of plasma through the flux capacitor every time it bubbles. After about 10 minutes tweaking the pressure etc, the neutron detector starts popping, at first sporadic and then more regular, and occasionally it goes crazy for a few seconds. There was a point where I had a real storm lasting around 10 seconds.

My best guess for neutron TIER is in the low 10E+5 range, which at least gives me something to work with.

Steven

flux-capacitor - 1.jpg
Flux capacitor
Fix-Chart.PNG
Practice Run
flux-MKII.png
Improved flux
Resistors.jpg
Voltage Divider
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG

User avatar
Steven Sesselmann
Posts: 2096
Joined: Thu Aug 11, 2005 1:50 am
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: FICS II Planning and construction

Post by Steven Sesselmann » Tue Apr 25, 2017 5:45 am

Hi Guys,

I received the longer flux capacitor last week, it performed better than the MK1 version, at least for the first 45 minutes.

After installing the new plasma backdraft arrestor, I did a long pump down and the chamber was still holding vacuum well, after valving off the chamber the leak was around than 1 micron per minute.

I started the run gently at around 15 micron pressure and gradually decreased the pressure as I increased the voltage. The machine is notably better behaved now that I have the pressure controller and gas cylinder at ground potential. The addition of two large ring magnets top and bottom may also help keep the plasma going.

After allowing the plasma to stabilise for around 10 - 15 minutes, I started seeing the neutron counter move upwards, but then I lost ignition for a moment and had to start again, and at one point the neutron TIER spiked somewhere around 400,000 n/s. At this point something went wrong, and when I went to check the apparatus, I noticed small bubbles coming from the side wall of the glass plasma backdraft arrestor, and on closer inspection there was a hole right through the glass.
FICS-Run-2017-04-19.PNG
Fix Run 19 April 2017
I feel a lot more encouraged by this latest run and I think a new flux capacitor made from thicker glass should do the trick. Considering I had a steady 150,000 n/s at 30 kV, I have my sights on the mega neutron mark.

Here is my latest Labview VI (Virtual Instrument), from this control panel I control my experiments from a safe 10 meter distance behind a 24" sandstone wall.

Note the addition of a neutron rate meter. From labview I am grabbing the data from the PC Line In port coming from my neutron detector (GS-Neutron-150), counting the pulses and calculating the count rate in software.

Stay posted...
VI.jpg
Virtual Instrument
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG

User avatar
Richard Hull
Site Admin
Posts: 10842
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: FICS II Planning and construction

Post by Richard Hull » Tue Apr 25, 2017 6:36 am

Nice report and I am glad you got it woking, and makiing decent neutron numbers. Sorry about the hole spoiling you continued running of the system. Back to the repair shop for thicker glass. Keep us updated.

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
Dennis P Brown
Posts: 1490
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: FICS II Planning and construction

Post by Dennis P Brown » Wed Aug 23, 2017 1:12 pm

Glad your system is/was working and providing neutrons (none too shabby generating 400 k neutrons!); that is a great start but set backs do occur. Doing any real reserach oft (more than most realize) end up like that.

Like the names you use for the system ...I too used a movie name for some of the research work I did in the past; in my thesis I developed a new material (based on glass); while its formal name (laminated Cermet) was used by me in my thesis defense, I did jokingly called it transparent aluminum (in fact, it is 99% glass, but is as flexible and supports superior tensile and strength properties compared to 7070 aluminum alloy.) It is transparent - all of 100 microns ... the committee, however, was most definitely NOT amused at all by my reference ... bad mistake on my part. Not a good idea to make those types of jokes in such formal proceedings.

Post Reply