Kilometer Mean Free Paths

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
Post Reply
User avatar
Liam David
Posts: 509
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Kilometer Mean Free Paths

Post by Liam David »

I hit 9.3e-9 torr in my system today after a very thorough cleaning and bakeout. With some luck, I'll hit the mid e-9 range within the next couple dozen hours. 10/10 would recommend an MFC bypass valve.

low pressure edited.png

Ignore the crossed-out items as these values are meaningless--the hardware (PSU) isn't connected so the op-amp buffers are floating. Orange box around pressure reading.
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Kilometer Mean Free Paths

Post by Rich Feldman »

Congratulations!

What is your pressure measurement instrument?
Please excuse me if that's explained in your other system development posts,
or obvious to those skilled in the art of e-9.

Is your MFP a ballpark value for common molecular species, as opposed to the MFP for electrons?
[edit] A few minutes of investigation confirmed Liam's value, and reminded me that the bonus for electrons in gas is not orders of magnitude. I'd been thinking of tiny CRT's that don't need super high vacuum if beam length is order of 1 cm.
All models are wrong; some models are useful. -- George Box
User avatar
Liam David
Posts: 509
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Re: Kilometer Mean Free Paths

Post by Liam David »

Thanks! I'm using a Pfeiffer PKR 261 combination cold cathode + Pirani gauge. The gauge startup time at these pressures is on the order of a minute or so. The only gauges I'm aware of that measure this low are cold cathode, hot cathode, and RGAs, although I am approaching the limit of the PKR 261 which is about 3e-9.
User avatar
Richard Hull
Moderator
Posts: 14976
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Kilometer Mean Free Paths

Post by Richard Hull »

That is a very deep pumping! I can't remember anyone here talking a bottom of 10e-9 torr. Congratulations.
Normally ion pumps and Ti-sub pumps are needed to warrant vacuums this deep and $$$$$ instruments to claim depths deeper than 10e-10 torr.

Naturally vacuums of 10e-10 and below are just never found in most sciences. Most such vacuums are reserved for particle colliders and special accelerators. CEBAF needs 10e-12 torr. Its beam line is not large in diameter. Still it takes weeks to pump down. Back in the 90's we had the designer of the accelerator niobium beam line resonant multipliers, Jock Fugitt, come to three of our TCBOR monthly meetings. (morphed into HEAS in 2000)

Jock was a wonderfully plain man. Never saw him without suspenders. He brought one of the large pure Niobium tuned resonator multipliers he designed that was bathed in liquid helium, when in super conducting use at CEBAF. For those interested, I include his photo (tiny) and a copy of his paper on the cavity. CEBAF would be his last accelerator job. He worked at CERN, Berkley and many accelerators during his productive life. One thing he noted to us was once the vacuum burped during a run to 10e-10 torr and five of the niobium multiplier cavities punctured exploding their liquid helium tanks. (420 cavities in the original system)... 4 month fix. Deep vacuums are the order of business at the new 12 GeV CEBAF...Now Jefferson National Lab.

Richard Hull
Attachments
New 12GEV cavity resonator.  Jock brought the original 8GEV 5 cell unit from early CEBAF, (about 30-inches long, cells were 6-inches in diameter.  Made of ultra high purity Niobium.  Looks like a metal muffler.  Sealed in liquid helium cryostat for operation by the hundreds!
New 12GEV cavity resonator. Jock brought the original 8GEV 5 cell unit from early CEBAF, (about 30-inches long, cells were 6-inches in diameter. Made of ultra high purity Niobium. Looks like a metal muffler. Sealed in liquid helium cryostat for operation by the hundreds!
Beam line of waveguide fed cryostats...WOW!!!
Beam line of waveguide fed cryostats...WOW!!!
Jock Fugitt Pix.jpg
Jock Fugitt Pix.jpg (42.45 KiB) Viewed 4585 times
Jock Fugitt.pdf
(110.33 KiB) Downloaded 137 times
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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Liam David
Posts: 509
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Re: Kilometer Mean Free Paths

Post by Liam David »

I have two turbos evacuating the system, although one is connected through a very low conductance path (RGA) and so doesn't do much. No ion or TSP pumps needed. I have all-metal seals with 2 internal viton o-rings, a very small surface (few mm^2) of teflon, and some PCTFE diaphragms in valves. Cleaning + bakeout make all the difference...
User avatar
Richard Hull
Moderator
Posts: 14976
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Kilometer Mean Free Paths

Post by Richard Hull »

Sounds like the ultimate low pressure system. Lots of cool vacuum experiments in such a system are possible beyond fusion. Very nice. Bottoming a vacuum system to such levels is a goal in and of itself to the aspiring vacuumist.

Jock noted that even CEBAF vacuum levels had the best pressure measuring instruments fade to near worthless levels of uncertainty. Deep sub nano torr levels are discussed in some of their papers. If my thoughts are right 10e-9 torr is one nano torr. One chart shows 10e-3 nano torr. Again, Wow!

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Roberto Ferrari
Posts: 364
Joined: Tue Mar 02, 2004 12:21 pm
Real name: Roberto Ferrari
Location: Argentina
Contact:

Re: Kilometer Mean Free Paths

Post by Roberto Ferrari »

Hi

Great achievement, Liam!
Many years ago I reached those levels with a "poor-man approach". May be I already posted this but memory is feeble...
Thirty years ago, I optimized in my small lab a technology able to reach 10-9 torr, learned from....Electronics magazines from the 30 or 40s.
In those days, there were full page advertisements of vacuum pumps and a side point was the technology involved (mostly Litton Industries). If you go to a collection of Electronics, you will learn that in those days, between the diffusion pump and the high vacuum chamber, they used a "charcoal trap". It was a trap with a cage full of coconut charcoal, with a low voltage heater, with a feedthrough for the positive pole, closing the circuit through ground.
Rather weird to tame, the system is turned on in the following way:
1) mechanical pump on and charcoal heater on, as well heaters on your chamber walls
2) then you monitor low vacuum at the point between the diffusion pump and the mechanical pump
3) after a long time of evolving a lot of gas, the vacuum start to improve,
4) then you turn on the diffusion pump
5) finally, you have a good high vacuum at your chamber, around 10-6 torr...
6) then, you turn off heating to your charcoal trap...
7) and in half an hour your vacuum improved two orders of magnitude!

It is advisable to have a cooling loop around both sides of the trap, in order to protect the large o-rings at the diffusion pump and at the chamber.
In my experiments, I replaced the charcoal by 5A molecular sieve pellets, in a SS system, that every time the vacuum was broken, was filled with dry N2.
Then, after several cycles, I finally reached 10-9 torr, several hours sustainable.
I warn everybody to be patient! To start the system takes a while, because the charcoal or the molecular sieve store lots of water and will require a long heating procedure.
I employed successfully this technique to rebuild power grid tubes and x-ray tubes.
Hope this may be helpful to many amateurs.
Roberto
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Kilometer Mean Free Paths

Post by Rich Feldman »

Yay for old technology!
Didn't most diffusion pumps in those days use boiling mercury instead of oily fluids?
Kind of thing we would see in the Amateur Scientist column (by C. L. Stong) in 1960's Scientific American magazine.
All models are wrong; some models are useful. -- George Box
User avatar
Liam David
Posts: 509
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: Arizona

Re: Kilometer Mean Free Paths

Post by Liam David »

That charcoal trap must have taken just ages to pump out!

After another bake at 150 C for 4 hours, followed by overnight pumping, the pressure is down to 4.7e-9 torr. Needless to say, I'm quite happy with the result.

I've had some gauge instabilities over the last few days which caused the pressure reading to fluctuate by orders of magnitude (1e-9...1e-5) on a timescale of seconds, but having now solved the issue, I think contamination was to blame. When I initially baked the system several days ago, I did not heat the gauge (< 50 C) while the chamber was at ~150 C to monitor the pressure over time, causing contaminants to condense within the gauge. Igniting the cold cathode discharge caused the contamination to slough off and outgas. Yesterday I baked the whole system, gauge included, and the problem seems to have disappeared. This highlights the importance of a uniform bakout.

During a brief turbos off period to rearrange heaters, I was also able to pull the chamber to 1.5 mtorr with just the rough pump. The low chamber outgassing in addition to the hot and outgassed pump oil (a week of continuous pumping) makes a big difference.

I'm convinced the high e-10 torr is achievable, perhaps with a long bake (12 hrs +) at 150 C. The limitation is the viton o-rings on my two valves, otherwise I'd go to 200 or 250 C. But my gauge can't even read that low, so it is a moot point.


My procedures for cleaning the vacuum hardware are based loosely on those from various accelerator labs. Every single part, except copper and VCR gaskets, was cleaned this way, including o-rings, foreline parts, and flexible metal hoses.
  • A gross cleaning with wire wheels, paper towels, etc. to get rid of rust and grime.
  • Depending on the degree of contamination, a minimum 10 minute ultrasonic cleaning in a hot, 60 C Alconox solution with distilled water. Aluminum parts should be cleaned at a lower temperature and for less time. The cleaning solution was replaced once it became noticeably cloudy.
  • A rinse with distilled water from a wash bottle to get rid of most of the Alconox. This improves the longevity of the distilled water in step 4.
  • Immersion with agitation in three successive distilled water rinse buckets. This step dilutes any remaining contaminants by ~1e6.
  • A rise with 99% anhydrous isopropyl alcohol to remove the water, any remaining hydrocarbons, and Alconox.
  • Air drying in a dust-free environment. Wrap in clean aluminum foil. Do not let any water, distilled or otherwise, dry on any of the parts, and absolutely never put parts in foil unless they are completely dry. Don't ask how I learned that the hard way.
The turbo pumps were cleaned as instructed in their manuals, that is by immersing them, upside-down, to the top of the foreline flange in 99% anhydrous IPA for 10 minutes. They were moved up and down to flush the blades several times and then left to air dry, inverted, in a clean environment.

Everything was assembled in a clean tent with new gaskets. I used clean latex gloves and avoided touching any vacuum-exposed surfaces or gaskets.

During the initial pumpdown, a lot of IPA vapor was evolved over the course of about an hour. This vapor and the bulk of the remaining adsorbed water trashed the rough pump oil, which I replaced once the outgassing subsided.
User avatar
Sophie Yang
Posts: 30
Joined: Mon Jan 13, 2020 5:57 am
Real name: Sophie Yang

Re: Kilometer Mean Free Paths

Post by Sophie Yang »

It is relatively easy to enter the ultra-high vacuum range only by using turbomolecular pump. In fact, the single-stage turbomolecular pump can hit E-11torr, and the two-stage molecular pump can hit E-13torr (exceeding the range of Agilent uhv-24p)

The accuracy of cold cathode vacuum gauge under ultra-high vacuum is difficult to guarantee. When hitting the pressure of E-9torr, pkr261 does not even mark the accuracy range, and the error at this time will exceed ± 30%.

The Pirani vacuum gauge will outgassing  under ultra-high vacuum. The measurement of ultra-high vacuum range should use independent B-A Gauge, such as Agilent uhv-24 and uhv-24p, Granville Phillips 370 stable ion gauge.

In addition, the B-A Gauge and RGA should be degas during the baking of the chamber. The degas of B-A Gauge is carried out on the controller, and the degas of RGA can be carried out on the software, and the emission current is generally set at 10mA. This can avoid the sudden rise of pressure when RGA is opened.

After baking, the pressure quickly entered E-9torre with the decrease of temperature. When the temperature of the chamber returns to room temperature, ion gauge has exceeded the range.
MKS 358 and MKS 370 VGC
MKS 358 and MKS 370 VGC
under range
under range
When I turn off the Turbomolecular pump to 50% speed, the gas molecules flow back into the chamber through the Turbomolecular pump, and the pressure begins to rise. At this time, the ion gauge also has a reading again.
20220717_111413.jpg
20220717_111427.jpg
20220717_111449.jpg
User avatar
Sophie Yang
Posts: 30
Joined: Mon Jan 13, 2020 5:57 am
Real name: Sophie Yang

Re: Kilometer Mean Free Paths

Post by Sophie Yang »

For a sufficiently clean vacuum system, the peak behind 44amu should not appear in the RGA spectrum. Some vacuum systems are very dirty without surface treatment. However, due to the excellent pumping effect of molecular pumps for high-quality gases, these pollutants will be cleaned up in no more than 15min.
Dirty vacuum system
Dirty vacuum system
When entering E-8torr, the RGA spectrum will become very clean.
mmexport1648638644137.png
1amu: H+ (The Turbomolecular pump has poor pumping effect on hydrogen)
16amu: O+ (from H2O, CO2, CO)
17amu: OH+ (from H2O, without baking)
18amu: H2O+ (from H2O, without baking)
28amu: CO+ (from CO, Carbon deposits on the filament are oxidized, when 14amu peak does not exist, 28amu is not considered as nitrogen )
44amu: CO2+ (from CO2,ditto)
Post Reply

Return to “Vacuum Technology (& FAQs)”