Neutron Club Application - Sarvesh Sadana

Current images of fusor efforts, components, etc. Try to continuously update from your name, a current photo using edit function. Title post with your name once only. Change image and text as needed. See first posting for details.
Post Reply
Sarvesh Sadana
Posts: 48
Joined: Tue Aug 23, 2016 1:55 am
Real name: Sarvesh Sadana
Location: Fremont, CA

Neutron Club Application - Sarvesh Sadana

Post by Sarvesh Sadana » Fri Oct 06, 2017 1:26 am

Hi!

I’m Sarvesh. I’m a senior (17) in high school. I’ve been working on building a fusor for about a year and a half. I apologize for not keeping a post detailing my fusion efforts in this forum. I have been too lazy to do so. I do, however, have pictures of my previous iterations if needed. A fair amount of stuff is also on my website http://sadanaresearch.com/project-index/fusor-overview/

I had my first full fusion run recently, and gathered data which suggests that I have finally achieved fusion.

Fusor Mk. 1
IMG_1367.JPG
Vacuum system:
IMG_1368.JPG
Here, you can see the bulk of my vacuum system. I have a two stage mechanical pump (gets to ~100 microns), and an air cooled Veeco oil diffusion pump. The ultimate pressure after around 30 minutes is .7 mtorr. It may be slightly above or below, as I haven't had a chance to calibrate my convectron gauge tube with a system of known pressure. The foreline pump is connected to the backing pump with a rubber vacuum hose. There is an automatic (I replaced the handle with a block of wood and added a motor) gate valve between the chamber and the pumps. The convectron gauge tube is mounted on the elbow I have. Unforunately, they are both O-ring seals, so I had to machine a piece of titanium to fit between them. I would guess that this is contributes to the low quality vacuum I have.

The main chamber itself is a 6" SS sphere. I welded two 8" CF flanges to them and bolted those together. All of my gaskets are Viton. I use a 40kV feedthrough, and have a grid made of 1.5 mm tungsten mounted on that. There is also a viewport on the main chamber.

Fusor input power:
IMG_1377.JPG
IMG_1376.JPG
IMG_1439.JPG
I use a -30kV 32mA Glassman supply to power the fusor. Unfortunately, it appears to be a custom model, so I cannot find the manual for it. This means that I can't modify it to give a higher voltage at a lower current. I am using a 200Kohm 50 watt victoreen resistor as the ballasting resistor between the power supply and the feedthrough. The ends have a heavy amount of silicone gel on them to prevent arcing.

Gas input:
IMG_1375.JPG
IMG_1437.JPG
I wasn't able to acquire any deuterium, so I had to resort to making my own with electrolysis. I used a PEM cell and connected that to two reservoirs, and had the deuterium gas stored in a small beaker. The gas then travels through a regulator (I can't adjust the regulator, for some reason), and a tube of drierite. It then is let into the chamber via a needle valve.

Plasma:
Screen Shot 2017-10-05 at 3.05.54 PM.png
This is a picture of a standard plasma in air. The voltage is around 7 kV at 20 mA, and the pressure is about 15 microns.

Neutron detection:
IMG_1371.JPG
IMG_1296.JPG
I built my own proportional counting system. It's a helium-3 tube connected to an oscilloscope and a power supply. The oscilloscope is then connected to a computer, which uses PIcoscope 6 to discriminate and count the values (-12 mV was used as the trigger setting). I created a faraday cage out of lead, and put blocks of paraffin wax inside. I also used an inch of high density polyethylene to moderate the fast neutrons even more. The pictures show the final system and an earlier iteration (so you can see the wiring) respectively. The power supply is set to 1600 volts.

I'm afraid I cannot separate the moderator from the tube, as it the paraffin wax is built into the faraday cage.
IMG_1419.JPG
IMG_1365.JPG
My background count seems accurate. I have taken multiple, and they are all in the range from 2-6 CPM, depending on where I am. The first count was taken recently at UC Berkeley, and the second was taken in my home.

My fusion run:
This was my first pass at fusion. There were some problems with my camera-- specifically while I was fusing, the video feed kept cutting off. There were also random white dots. I suspect x-rays caused this problem.

Operation:
I first used the pumps to get the pressure down to .7 microns. I then applied voltage to the PEM cell, and waited until there was enough deuterium in the beaker. I slowly opened the regulator and the needle valve, and allowed deuterium to flush out the air. Once it did, I turned on the plasma to a few kV. I then used the automatic valve (I didn't want to get close) to lower the pressure to 6 microns (the real pressure was higher since deuterium has a lower molar mass than air). I turned the power supply all the way up to 30kV at 18mA-25mA. After gathering my data, I purged the deuterium and closed the gas valve, and shut everything off.

Data:
Pressure: 6-8 microns (true pressure is slightly higher)
Voltage: 30 kV
Current: 18-25 mA
Fusing time: ~30 seconds
Moderation: 1 inch of HDPE, plus 1.5 inches of paraffin wax in all directions
Faraday cage: 1/16th inch lead sheet. The face of it was covered in aluminum foil.
Power supply for the neutron detector was set to 1600 volts
IMG_1366 copy.JPG
I recorded 5600 counts in total. According to similar runs in air, it tends to generate 300 CPM on the scope due to RF noise. As you can see, when I dialed the voltage up in a deuterium rich environment, the neutron counts spiked (minute 4 and some in minute 5). Also note the spectrum on the right. I believe the rest of the data is simply due to noise.

To ensure that my data was completely accurate, I have an appointment at the High Flux Neutron Generator ( http://www.nuc.berkeley.edu/tags/hfng ) at UC Berkeley. I will test the detector and report the percent error from my counter vs their system.

I would prefer a conditional acceptance into the neutron club, pending more data. After my appointment with HFNG, I will do another run, and will record it completely. I would have submitted this afterwards, but I wanted to do it before HEAS.

Please let me know if you need any more data or info,
Sarvesh

User avatar
Dennis P Brown
Posts: 1714
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Neutron Club Application - Sarvesh Sadana

Post by Dennis P Brown » Fri Oct 06, 2017 12:25 pm

Looks interesting and that you might have achieved fusion.

I have a few questions about the post - you say your average background counts are 300 per minute ( a rounded value I assume. Would be nice to see that raw data as well.) That you obtained 459 counts per minute when you raised your voltage - while a significant level above your stated noise floor of 300 cpm I am wondering about you saying you have this count rate "When I dialed up the voltage."

However, unless that 459 counts per minute was also done under the same conditions (max voltage and current) as your average count rate I am wondering if the reading is due to neutrons or possibly, from more noise from the increased voltage/current? While removing the moderator would quickly answer this question and prove that you had a neutron count increased as measured, you indicate that you can't do this (why?) So, to clarify my question, was the 300 cpm done with the same voltage/current (in air as you did state) as with the voltage/current when you obtained the 459 cpm with deuterium?

Also, why did you make such a short neutron run and why did the signal then so quickly fall off to your noise floor? Did you lower voltage or run out of deuterium?

A few runs to get better statistics might also be appropriate. Determining data variation is important for confidence in said data.

I bring this up because noise issues can be a big problem with a fusor.

Your "Faraday cage: 1/16th inch lead sheet" does not look to be very good; one needs to electrically seal the ends and have a proper feed into the detector tube that is also electrically conductive to the shielding case system - crumpled Al foil is not going to provide electrical shielding at all. Also, 1/16 inch thick Pb is a bit thin for good RF shielding. Your use of coaxial cable is a good idea. This shielding for the detector may or may not be significant for your noise issues. Placing a vertical metal foil between a fusor and a detector is useless. RF goes everywhere and one needs to encase a detector for proper RF shielding to work.

All that aside - your data acquisition system is very nice. Making your own deuterium is also a good achievement in of itself. You have the makings of a very impressive system. The wiring could be better organized and your mechanical supports could be a bit more polished but that is a taste issue only; you have a good overall setup. I see you too (like I had too) used a fan to cool the main HV electrode - those can get rather hot. Also, your plasma looks very clean (color) and looks that you have a good leak free system. Your 30 kV and 32 ma is really a good power level for fusion. That is a very good supply.

Good idea to get a handle on your detector system's performance by having it run against a known source. While that will not address an electrical noise issue I do think that is a good idea; determining a valid neutron measuring rate for a given detector is an impressive next step.

Richard makes all determinations for the neutron club; I am just asking questions from my read of your experiment (and overall, a very good one, too) - please don't think I am questioning your actual results; I am most certainly not. You have an impressive system. I do strongly suggest that you do more runs and if at all possible, do a few runs with and without a moderator - that can side step any noise issues.

Sarvesh Sadana
Posts: 48
Joined: Tue Aug 23, 2016 1:55 am
Real name: Sarvesh Sadana
Location: Fremont, CA

Re: Neutron Club Application - Sarvesh Sadana

Post by Sarvesh Sadana » Fri Oct 06, 2017 2:27 pm

Yes, there are definitely a few problems with it.

The issue with the faraday cage is the largest by far. Currently, the paraffin wax is actually screwed into the lead body, so I can't remove it without significant damage. I agree that the aluminum foil I used to cover the entrance of the cage is less than ideal. How should I go about making proper RF shielding?

The 300 CPM I stated was more of a weighted average. The true range is from 0-500 CPM, depending on the voltage (usually around 400 CPM for 30kV, but sometimes higher). The 459 was taken at the maximum voltage. While the camera only worked intermittently, I did see an arc toward the grid, and I was worried that I significantly damaged it. I consequently ran the plasma in air during minute 7 at 30kV (unsure about the current, sorry). I will post the graphs of my air runs as soon as I can -- the laptop they are on is currently locked in a drawer at UCB.

I am concerned with my signal to noise ratio, which is why I will be visiting HFNG soon to compare my value with their instruments. I may also use a PoBe source and calibrate it with that.

The reason I had such a short neutron run is that I believe that the rate of deuterium flowing into the system was less than how much flowed out. I wasn't able to adjust the regulator (I bought it for $5 at a flea market, so), and I had to open the gate valve significantly to achieve the required 6 micron pressure. I closed the needle valve in minute 5.

The pictures of the electrolysis system were taken after my run. After the deuterium runs out, water starts to flow through the system. The drierite tube seems to be completely used up from that. Thankfully, I have 8 more on the way.

And yes, it's definitely not a very pretty system. Operation was quite hectic, mostly because it's very messy. I'm planning on making a cable organizer sometime.

I hope that answers your questions! Feel free to ask any more. I don't mind.

Sarvesh

User avatar
Dennis P Brown
Posts: 1714
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Neutron Club Application - Sarvesh Sadana

Post by Dennis P Brown » Fri Oct 06, 2017 2:53 pm

First off, I have VERY little doubt you have both fusion and are reading neutrons; the issue is proof that rises to a level Richard accepts (and you may have that already. I am just looking at possible questions.) Richard may feel that your single data run is enough to include you as a possible member of the neutron club. What I was concerned and am still concerned with is noise level compared to neutron signal.

Second, don't get discourage by my questions nor this issue relative to your data. I had a lot of noise in my data. That is why I made numerous runs and did both statistics and removed the moderator and did a series of runs (all that identical voltages, pressures, deuterium flow. Also, like you, did an air run.) While my signal was a good bit more clear than yours, none-the-less, we can be fooled by our owe bias - again, in no way am I saying this is an issue for you! You presented good data. Just very noisy - 459/300 is a signal but getting close to a point that is hard to separate for a single run; hence, my suggestion for more runs.

Third, your assembled parts and fusor are very good - please don't think I am judging by those few "quick - fix" solutions - everyone who can't buy off-the-shelf does that. I did!

Also, RF shielding is difficult and from my experience with my russian made system, and it was a terrible bear in that respect (pun there.) I used a 10 foot co-axial cable, and a battery HV supply and a fully electrically sealed case for my detector. That is not easy for the average person; so, best to use both extra runs and moderator removal to prove noise is not creating one's neutron signal.

If noise is an issue, that can be addressed - not easily but can be done. Many here can help. You have a good start with all coaxial cabling. Sealing a detector is tricky - I do think some have posted images of their's.

I am fairly certain, that once you get more runs, and runs with and without a moderator, you will very likely get a clear fusion signal even with any noise issues.

My method for the detector case: I used a heavy (4 mm thick) copper tube, thick brass end caps (one end was soldered in place, the other was a tight fit with conductive epoxy.) I used a shield coaxial cable thru a co-axial mating connector to reach the detector tube inside that case. Both my computer counter and HV detector supply operated off a battery/converter system (which can, also create serious noise!) Crumbled aluminum foil does very little. Also, lead foil isn't the answer - that is good for an x-ray shield but not for RF shielding. Rather, a metal tube with one end cap would be a good start. Also, don't include the moderator with one's detector. The detector should be removable; I use a plastic tube inside a large container that has been filled with paraffin. In that way I can remove my detector. These are just ideas and not necessarily what you need to do. Wait for Richard to comment first before changing anything!

Another point: the detector does NOT have to be inside the moderator. All one needs is a slab of moderator between the detector tube and the chamber! We are not trying to get exact room value measurements, just prove fusion. So, don't encase the detector in a moderator (like paraffin) if that makes removal an issue. So a sheet of Pb, than a slab of moderator and the detector tube all in-line (roughly) with the center line of the electrode in the chamber.

Aside: holding a plasma current steady in a fusor is difficult. Arcing happens but I wouldn't be too worried. Your heavy resistor should help protect your power supply. As your system runs and gets 'seasoned' its stability will increase. AS for running out of deuterium, maybe have a bigger reserve holding tank. I do not know how you control the vacuum to your chamber but mine is nearly fully shut and I leak a very slow rate of deuterium gas at my 3-5 micron pressure. Yes, if the gate valve (or what ever) is fully open, the rate of gas flow can be extremely high.

Also: congratulations on your fully assembled and operating fusor and neutron detector system (very nice) - very few reach that level!

User avatar
Dennis P Brown
Posts: 1714
Joined: Sun May 20, 2012 2:46 pm
Real name: Dennis P Brown
Location: Glen Arm, MD

Re: Neutron Club Application - Sarvesh Sadana

Post by Dennis P Brown » Fri Oct 06, 2017 4:43 pm

Again, while reducing noise is a good goal, see what others have to say on this subject, too. Shielding is just one approach.

My memory about my data's noise level isn't correct; it was rather noise free - shielding works. I will link to that post because I have a good image of my detector's position relative to the chamber and how I set up the shielding/moderator plates. Note how my encased detector tube is housed in a plastic tube, that in turn, is in another such tube filled with paraffin. What counts is how much (within reason) paraffin/moderator is between the detector and chamber, not having 4Pi radians of moderator around the detector.

viewtopic.php?f=6&t=11313

Our two power supplies offer very similar voltage/currents. As such, you should get similar or higher counts as mine (excluding geometry issues of the electrodes/chamber.)

Sarvesh Sadana
Posts: 48
Joined: Tue Aug 23, 2016 1:55 am
Real name: Sarvesh Sadana
Location: Fremont, CA

Re: Neutron Club Application - Sarvesh Sadana

Post by Sarvesh Sadana » Sun Oct 08, 2017 10:33 pm

Very interesting. I definitely will look for a metal tube of the type you are describing -- that would be far easier than attempting to make a proper RF cage. After speaking to some of the folks at HEAS, I believe the problem lies in the power supply cable. The RF may be traveling along that wire, amplifying it. I will use a torodal ferrite choke to eliminate some of it. I may also raise the trigger voltage to a level such that the noise levels are reduced. And yes, you are correct that I shouldn't use my paraffin radially. I will see if I can disassemble the cage and reorganize (or remove) the paraffin to give more moderation. Unfortunately, I do not have access to my detector for at least a week.

I think my problems with the deuterium were simply due to me being inexperienced with it. As I get more experience, I will no doubt be far better at operating it, and it will likely disappear.

I really appreciate all of your insights and help! If you have any more questions or comments, please fire away.

Sarvesh

Sarvesh Sadana
Posts: 48
Joined: Tue Aug 23, 2016 1:55 am
Real name: Sarvesh Sadana
Location: Fremont, CA

Re: Neutron Club Application - Sarvesh Sadana

Post by Sarvesh Sadana » Mon Nov 27, 2017 11:40 pm

I've been quite busy with schoolwork and haven't had much time to work on this. I have, however, installed the ferrite choke on the HV input line. What else can I do to reduce the noise?

Also, I'd appreciate it if Richard reviewed this.

Sarvesh

User avatar
Richard Hull
Moderator
Posts: 11487
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Neutron Club Application - Sarvesh Sadana

Post by Richard Hull » Tue Nov 28, 2017 5:08 am

I must have missed this posting as it was coincident with HEAS 2017. After review I will conditionally accept while hoping for future runs that are a bit more specific. 300CPM background tells me that something is wrong with the counter, Was this background with the fusor running, but not fusing?

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.

Sarvesh Sadana
Posts: 48
Joined: Tue Aug 23, 2016 1:55 am
Real name: Sarvesh Sadana
Location: Fremont, CA

Re: Neutron Club Application - Sarvesh Sadana

Post by Sarvesh Sadana » Tue Nov 28, 2017 8:31 am

Sure, no problem. I'll try to do another run soon when I replace my drierite tube.

Yup. The 300 CPM is the average noise floor generated by running the fusor at fusion conditions without any deuterium. I'm hoping that the ferrite will help absorb and limit the RF produced. The peak seen on the graph is from when I ran it in deuterium rich conditions. The real background ranges from 2-6 CPM on average.

Sarvesh

Post Reply