Page 1 of 15

Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA-9mA

Posted: Mon Sep 07, 2015 8:11 pm
by Andrew Seltzman
With the new modifications to the spellman supply, I can now run at up to 9mA grid current at up to 40kV. After pumping on a turbo station for about 14 hours, the fusor pressure is down to 3e-4 torr as measured on an ion gauge. Running the ion injectors at low power without any grid bias lowers the pressure to about 2.8e-4 torr, but then it will rise back up to 3e-4 when the ion sources are turned off, that seems to be the limit of my system due to gas permeation of the viton o-rings / out-gassing of internal plastic parts ant the low conduction rate of my vacuum system. I also suspect the boron nitride grid ceramic insulators on the grid are out-gassing when bombarded with plasma. During operation, the pressure will rise towards 1mTorr, but will pump down pretty fast after grid bias is turned off.

With the anode layer ion sources off, at 3e-4 torr, there is no measurable current draw from the grid at -40kv. The injectors will run perfectly at this pressure with enough plasma current to easily hit the 9mA limit of the spellman. The fusor maintains rock solid discharge stability with these sources on all the way down to the 3e-4torr minimum achieved pressure(for reference, there is no ballast inductor or resistor in series with the grid supply), by varying the ion source bias, the current into the grid may be controlled up to the 9mA max as limited by the power supply.

4e-4torr, 1ma, 40kv, 4 injectors on
5e-4torr, 1.3ma, 40kv, 4 injectors on
7e-4torr, 7ma, 40kv, 4 injectors on
Vacuum during operation

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Mon Sep 07, 2015 9:35 pm
by Jerry Biehler
14 hours and you are only in the -4 range?

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Mon Sep 07, 2015 9:54 pm
by Andrew Seltzman
Yes, I just did a pump down / valve off / pump down plot to show pressure rise, it's entirely linear after around 2mTorr. I'm almost entirely convinced it's due to gas permeation of the o-rings. All my orings are viton, I have 16x 2.75" CF, 1x 6" CF, 1x 0.75" quick connect, 1x NW40, 1x NW16 on the fusor. I helium leak checked all o-rings and welds and found no leaks (~3e-8 TL/s), but that's only inspecting where the helium is blowing over(looking for discrete leaks), I'm fairly sure it is the cumulative permeability effect of all the viton in the system.

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Tue Sep 08, 2015 12:27 am
by Jerry Biehler
Unless you have a really tiny turbo you should be much lower than that in 14 hours even with a system that has been at air for a long time. You should be down to the -5 to -6 after that period. I bet you still might have one little leak someplace.

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Tue Sep 08, 2015 6:28 pm
by John Futter
Itend to agree with Jerry. We have systems at work with many tens of metres of viton o-ring. All of these systems get down to the low ten to the minus eight eventually (weeks of pumping).
But we will not use any zinc plated screws, polycarbonate, as these have extremely high vapour pressures

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Tue Sep 08, 2015 6:45 pm
by Dennis P Brown
Appears your system is leaking badly (like there ever was a good leak? OK, not very good joke.) Try a can of "Dust Off" type spray to search for the leak when the ion gauge is working - you will see a jump when it hits the area of the leak(s). Nearly pure alcohol works well, too. If you have O-rings, possible you have a scratch in a metal face that cross's over to of one and is a source of the leak.

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Tue Sep 08, 2015 6:59 pm
by Richard Hull
Based on what I am seeing and what is reported, this system will do fusion just fine provided you just flow in D2 with the pumps running and a slight valve down on the pumps. Start with a D2 pressure of 5 microns and work up. I assume you have a good neutron detection system that can readily herald increases and decreases in fusion as the pressure and voltage are varied.

I have run a terribly leaky fusor for years with a good diff pump that just cannot ever reach 10E-5 torr. My system has produced over of a million n/sec as there are many actual witnesses on this board who can attest. No one needs to go to or below do good fusion. Only vacuum head purists and the anally retentive need struggle for better vacuum levels. A laudable goal to be sure, but just not needed.

Richard Hull

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Tue Sep 08, 2015 10:53 pm
by Bruce Meagher
Beautiful pics, and a very impressive setup!

If your leak rate is truly linear then you have a real leak. You might consider running a longer test to capture data over a larger pressure range. This might help isolate the problem. As I’m sure you’re fully aware, the size of the leak is the change in pressure * volume / time. From your graph it appears to be around 9x10-4 sccm * vacuum chamber volume in liters (assuming I did the math correctly). If the pressure rise is linear then you can rule permeation, diffusion, vaporization, etc. of the o-rings. I believe these processes are all a function of pressure and therefore would not show a linear response.

How are your four ion sources plumbed to the chamber, and how is the gas connected to the ion sources? Do you have a separate larger pumping port on the ion sources for pumpdown? One area I'd think could cause some vacuum trouble would be in the gas supply system (e.g. a leaky needle valve, laser orifice not valved off, etc.).

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Tue Sep 08, 2015 11:51 pm
by Andrew Seltzman
If it looks like a duck and quacks like a duck...
I did give it a very through leak checking(with a He leak checker) including all welds, o-rings, and fittings, even inside the grid feed through box and inside the ion gauge. I must have missed a spot, somewhere... You would think that with such a large apparent leak it would be easy to find.

Dennis: How does the dust off work, you would think with a heavier molecule the leak rate would slow when blowing tetrafluoroethane over it.

John: Do you l-grease your o-rings?

Richard: I definitely want down to 1e-4 when operating with the ion sources to study what fraction of the ions hit the grid.

Bruce: The ion sources are mounted directly on the chamber nipples, they currently have the gas feeds capped off(VCR cap with new Cu o-ring) ... -index.htm

Re: Mark3 operation 3e-4 torr / quad ion source @ -40kV, 1mA

Posted: Wed Sep 09, 2015 11:18 am
by Dennis P Brown
The "Dust Off" spray works fine for finding even very slow leaks; I have detected very tiny leaks. As for "greasing" O-rings that only works if the leak is small - ie very fine cracks or scratches. First, as I pointed out, look for a scratch or a piece of debris that is causing a leak across the O-ring. However, to "grease" the O-ring, take a very small amount of high quality vacuum grease and rub it around the O-ring. Then use a clean dust free wipe and wipe ALL the grease off. That will leave more than enough to seal tiny cracks on the O-ring or micro-scratches on the metal surface. Remember that for a while, the grease will "out-gas" a good bit so give this method some time before thinking the system is still leaking. Also, use as little as possible - large amounts do more harm than good since a large leak stopped by grease will most likely start very soon and then a large amount of "grease" vapor will deposit all over your formerly clean vacuum system ... . Only do "grease" if all other methods fail - checking for scratches or debris and re-mounting all O-rings/re-tightening bolts to make sure they they are uniform and not over tightened.

Still, if you HE-leak tested I would think you didn't miss a large leak. Still, it happens so use either a "Dust Off" can or alcohol (70% or higher.) when the ion gauge is running. The gauge will "jump" when the area leaking is subjected to the test. While 10-4 torr is good for fusion, a large leak can introduce water vapor and as I understand, this can causes issues in achieving fusion. Whether that really is an issue or not, still, best to get rid of any major leaks.