Ready for assembly!

[fpg]

All the parts have arrived and it is now ready for assembly. When it is finished it will look like pic 2.

In the right upper corner in pic 1 you can see the mating flange for my diffusion pump. Some guys in a steel workshop made it for me for a quite reasonable price of 25 dollars. Sadly they said that they didn´t have the tools to assemble the reactor so i have to find another workshop that can assemble it.

To cut cost i bought 2 full nipple flanges instead of 4 half nipple once.

Any advices before I start putting it all together is appreciated

Thanks in advance
Fredrik Parnefjord Gustafsson

[Linda Haile]

It may well pay to polish all internal surfaces at this point, however it is essential that all residue of polishing compounds etc are removed prior to welding.

[Carl Willis]

Hi Fredrik,

Looks like a good start. You should consider a throttle valve between the diffusion pump and the chamber, since a high pumping speed during operation will only waste gas. I hope that requirement doesn't obviate the quite-reasonable $25 coupling the shop made for you.

Good luck with the rest of the fabrication.

-Carl

[dbrown]

Trival but don't forget to add a forced air fan for the DF air heat exchange coils - will save a lot of back flow.

Speaking of which, welding in two flat partial plates to act as "non-line of sight" inline baffles for the DF pump/chamber would also improve ultimate pressure (but not pumping rate) and reduce oil vapor from getting into the chamber from the DF.

[Richard Hull]

Fredrik,

While polishing might be pretty, it is absolutely not required and outside of just being pretty, it has no beneficial effect on influence on fusion. The time spent on a pretty polish might be better directed toward actual construction.

Richard Hull

[fpg]

Carl Willis

I don’t think that the high pumping speed should be much of a problem since I use a very weak roughing pump (2CFM). But it is probably a good idea because I will probably buy a better pump in the future.

Dennis Brown

I have several computer fans lying around so I will probably use them for cooling the DF pump. The "non line of sight" inline baffles you mentioned is an easy fix that I have thought of before and I will weld the two plates inside the half nipple flange as soon as possible.


Richard hull

I’m perfectly aware that the polishing won’t effect my fusion results but I thought that I might as well make it pretty since I had 20 min to spare.

- Fredrik PG

[Carl Willis]

Hi Fredrik,

Your diffusion pump has an effective speed somewhere in the 50-100 l / sec ballpark. To sustain a typical operating pressure in the fusor of 10 mtorr with this pump wide open on the system would require a very high gas throughput:

50 l/s * 10 mtorr = 0.5 torr l / s (~40 sccm)

I ran at about 1 sccm on all my projects, and my knowledge of others' systems suggests this is where they tend to be as well. I use a simple QF-25 manual shutoff valve as a throttle on a 50-lps diffusion pump, reducing the effective speed to an estimated ~1 l/s. 1 sccm is an easy value to achieve with capillaries, laser orifices, mass flow controllers, and metering needle valves, and will get you about 1000 hours of operating time from a deuterium lecture bottle. Too much lower throughput, and flow becomes difficult to control. Too much higher, and you'll blow through expensive gas at an uncomfortable rate.

The 2-CFM forepump (roughing pump) seems just fine to me for backing this diffusion pump.

-Carl

[lutzhoffman]

So far so good, all the advice has been spot on. Have you decided on what fluids to run in both of your pumps yet? There are two schools of thought on this, some prefer to run cheap but effective diffoil 20 in both pumps for simplicity, and others would say to use expensive Santovac 5, since it eliminates the need for a trap, and because your DP uses so little oil anyway.

Since you have the actual chamber design done now, you may want to put some serious thought into optimizing the vacuum system at this point. I would suggest at minimum also adding a copper wool filled trap between the RP and the DP. You can buy cheap copper scrubbing pads, and degrease them with soapy water, followed by rinsing with water, drying, and then a final solvent rinse followed by a high temp 300 deg F oven dry. This will absorb the back streaming RP oil, and prevent it from getting into your DP, and chamber. When these get dirty, then you just repeat the above cleaning. I adapted a fine mesh water filter unit from the hardware store, by unscrewing the cup and filling it with the copper wool. The transparent cup is great because you can see the oil when it gets dirty.

What ever you decide to do, searching the threads on vacuum stuff etc. in this forum will pay off big time in the long run. There is so much good information here, enjoy it. For D gas I have tried several ways to use D2O as a source, the best way for me was to pass D2O vapor through hot magnesium turnings in a quartz tube, with fiberglass plugs in the ends to hold the Mg in place. Hot Mg will also absorb nitrogen as well, if a small amount of air remains in the system. It is however best to flush the line with argon first. You can even use the chamber vacuum to suck the D2O vapor through the hot Mg, and aid in the evaporation of the D2O from the sealed flask, as it passes through the hot Mg, only D gas will remain.

Some folks add a dryer stage to D gas generated by various chemical and electrolytic methods, but some studies have show slightly increased fusion rates, with traces of remaining D2O, to the point where they even add a controlled amount. With the Mg method it tends to be pretty dry, when compared to electrolysis for example. You can experiment with H2O and Mg very cheaply to get familiar with the process, and then decide if you want to drop so much cash for the lecture bottle, and regulator etc. Take care and good luck, very impressive so far : )

[fpg]

Carl Willis,

Good to hear that my pump will do I was actually a bit worried about it being too weak.

Lutzhoffman,

I have not thought so much about what oil to use but I will probably go for diffoil 20 as you mentioned (I´ve heard it´s good for starters). The santovac 5 is a little expensive for my taste . Is it absolutely nessesary to use an oil trap when using diffoil 20 or is it enought (for starters) with the "non line of sight" inline baffles as Dennis Brown mentioned? (correct me if I´m wrong I have never used a diffusion pump before but I will sertainly read the vacuum threads to learn more).

I will do as you said and experiment with the Mg H20 method and familurize myself with the method before I test the real deal, but first I need to weld it all together .


Thanks for your advices,
Fredrik Parnefjord Gustafsson

[nemesistech]

Here is a link showing how I built my chamber, maybe it may help you.
http://motionpicturetransportation.com/chamber1.html

[richnormand]

WOW. Nice work.
No I have to clean my keyboard from drool....

[Daniel]

Hey, Nice start of the chamber. I'm going to do a chamber similar to yours. Why do you use the ratable conflats? Whats the difference from them to the non-rotatable?

[Richard Hester]

For diff pump oil, I would use silicone oils like DC704 and DC705 (or their equivalents). They much more affordable than Santovac and allow very respectable ultimate pressure without a cooled baffle. They also withstand oxidation much better than the cheap Octoil and equivalents, so an occasional up-to-air accident won't turn your oil charge into caramel. The mass spectrometer folks hate the silicone oils because of the extra ionic species they introduce (they use Santovac), but this should not be an issue with the fusor.
If you're using a huge 10" (or larger) pump, Octoil starts to look attractive because of the sheer cost of the oil charge, but for a little pump requiring only 50cc or so of oil, silicone is a good bet.

[lutzhoffman]

Hello:

Ref Fluids:

DC 705, and DC 704 are nice, but they have one problem, which may not be a problem with fusor application, but this IS a problem with electron microscopes, potential drop accelerators etc. The problem is that if you get any backsteaming, under discharge, ion / electron beam conditions. These silicon fluids can form insulating layers of SiO2 (Quartz) on the inside surfaces in the vacuum tank, or tube etc. This can affect the potential distribution, and HV properties, and even ruin a tube. Non silicon fluids tend to deposit carbon if anything, which is easy to deal with in comparison. Silicon fluids also have some solubility issues with many common solvents that are often used for cleaning, so you have to select the right solvent, if you need to clean your DP, or something else in your system. Maybe someone else has fusor specific knowlege as to if SiO2 deposits can be problematic in a fusor? I just do not know the answer to this question.

I ran DC 704 for years, for a coating chamber, and it worked very well, but now I use Santovac after getting 500ml on ebay for cheap, I would never go back to DC- 704 silicon fluid: No more cold traps between the DP and chamber, just a butterfly valve, and a cold cap which is built into my HS 2 DP. Plus if I were to loose vacuum again, due to breakage etc, its no big deal with Santovac, it tends to decompose into H2O and CO2 gas if somehow you get it that hot, which is not likely. I have lost vacuum at full temp more than once, due to breakage of glass parts etc, each time I first freaked out, and sweated it, until I checked the DP fluid, and all was fine.

My friend uses Diffoil 20 only, in both his roughing pump, and in his DP. His system works very well, and he has not had any major problems. He has a homemade 2 stage thermo-electric cooled trap between his DP, and his chamber. Very creative setup, he claims to get down to about -5C in the inside of the trap. The 2 stage thermo-electric cooler / fan is on the outside, and a foam insulated thick copper HV vac. feedthrough sucks the heat from the vacuum side copper baffle plate, to the external TC cooling setup,

He still has a cheap copper wool trap still between the DP, and the RP. His reason is: Even though he uses the same fluid in the DP, and in the RP, the lighter fraction breakdown products from the DP, and contaminants etc. will build up in the RP oil, he does not want these to migrate back into the DP. I have no idea if this is valid, but he is a "smart cookie", being a retired Xerox corp. engineer, with much partical experience, so I just accept his reasoning. His original cold trap was a dry ice acetone alchohol setup, which was much colder, but he is getting away with the -5C system now.

Octoil is toxic, and a suspected mutagen, with few advantages, which is why hardly no one uses it anymore, unless they have, or find a bottle for cheap somewhere. So in the end you have to decied what is best for you, and maybe for your health. Octoil cannot be that bad I suppose, because the subacate ester is used to keep C-4 plastic explosive pliable, and after my military years in the late 80's, my kids all look normal, well almost, but that's not genetic : )

No one else can tell you what is best for you, only knowlege gained can help you do this, and to select the correct fluid. Doug who contibutes to this column a lot, runs diffoil20 in his DP, and RP. He is of course not always correct, but he has actually converted my thinking on several issues, so I see him as one of the "bright bulbs" along with Carl, Richard, John and so many others in this column. These folks are people who actually build, and use the stuff that they are posting about in real life, which in my mind counts for something : )

Personally I like the fill and forget features of Santovac, and the elimination of the trap between my DP, and my chamber. I mention this only because when you look at prices, its more complicated than just the oil cost, because for example how much does a trap cost to buy, and then to fill? Yes you can probably eliminate the trap with DC 705 also, at a lower cost, but are you willing to accept the other aspects of silicon fluid? I cannot because I like DC potential drop accelerators, otherwise my pump could have DC-705 in it, jut the same. I am sure that with a little bit of reading, of past threads on this subject, you will figure it out. You have my respect already since you are a builder, so I will look forward to the day where you help me solve some problem, take care....

PS: Put a big fan on that DP with some sheet metal, plastic etc. to direct the air, a PC fan is to anemic, but a big 6" + fan from some other big IT item will work.

[Richard Hull]

I will be using pentavac/santovac in fusor V which will be underway after the HEAS conference should I decide to use my old diff pump instead of my 10" Turbo pump.

I picked up a thermo-electric cooler mounted on a 2" Veeco diff pump that is air cooled last weekend for $40.00 at a Teslathon. It came with a bottle of 704. I may use my old diff pump with the coller added of move over to the Veeco. We will see.

The turbo is water cooled and I don't like that. I have had the thing for years, but just am resisting the flowing water bit.

Richard Hull