A homebrew vacuum conductance valve for in tank use

For posts specifically relating to fusor design, construction, and operation.
Post Reply
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

I recently have been working in regimes where fine control of both gas pressure and net flow have become important, and just spinning the turbo up and down to control pumping rate has not gotten me all I want (and is a very slow thing to control), so I built this valve to put in the pump throat area, as I don't have room for one outside the tank anyway. I haven't tried this out yet, but expect it to do a nice job, and let me run the turbo in normal mode while keeping the gas flow low to save gas and keep some of the desireable reaction products around better.

I have posted a bit more on this in my website, at:

http://www.coultersmithing.com/AuxCP/CV.html

Yes, my bench is messy beyond belief, and yes I like making things out of copper and baling wire. The motor was a cheapo gear head from Marlin Jones, all metal gears and I took the oil out and replaced it with vacuum pump oil in very stingy amounts. The butterfly other pivot is, yes, baling wire in a tribute to kludgers everywhere, and so it won't vacuum weld to the copper piece.

The little motor is rated at 150 rpm at 12v, but even 6 volts is too fast. I will make a little power supply that has a lot of series R to slow it down, with a medium large electrolytic on the output to give an initial kick to overcome vacuum sticktion. If this works out well (should) I will then control it off one of the microprocessors on the system to regulate the flow on power supply current (sensitive) or a Pfeiffer gas gage (pkr251) output.

This was what, a couple thousand $ cheaper than buying one new? Can I have my scrounger credentials back now?

Maybe not in the exact spirit of some of the super cheap and easy project George S has been posting, but in the same league -- This was a one day thing with stuff lying around.
Attachments
CVOpen.jpg
Why guess when you can know? Measure!
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by Chris Bradley »

I hope that seals up OK for you, Doug! I thought I was pushing my luck with tapping engineering plastics....!!....
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

I don't think I need a great seal -- it's going to be a matter of degree. I plan to just set this in the pump hole at the tank bottom, so it will leak around the flange too. If it's too leaky, I can better fit the butterfly, add lips for it to rest on at closed, and generally tighten it up. Right now there's maybe a half square inch of leak area, and most of that is labyrinth so maybe it's good as is. I hope, but since calculating that is a bear, I'm just going to shove it in and try it. Even choking the 6" pump to 4" with the thing open, I think I'll have plenty for quick pumpdowns, this system is overpumped if anything. It normally gets to about 1e-6 mbar at the point the turbo is fully spun up -- fast. I get about 3 more orders magnitude better if I run overnight on the pump. Could be worse!

To get low gas flows at fusor pressures, we have been spinning the turbo down the max allowed for this design (from 833 rps to 165), but then have to run the forepump all the time to get vacuum enough, and there seems to be no middle ground on that we are happy with. Our normal high vacuum work we let the turbo spin full speed, and control the forepump by looking at the turbo current draw (this is automatic in the pump controller if you set it up and have a SS relay for the forepump). This doesn't work well at the lower speeds, because there's not much draw no matter what, and at lower speeds the backing pump cycling changes the net gas flow too much as the compression ratio of the turbo is very low at that point.

So, this is actually a convenience and *power saving* device to contend with my solar power system staying happy. The backing pump draws a ton of power and makes noise and vibration too -- avoiding running that so much is a good thing for completely other reasons.

I'll post on whether this works. I want to control at very low flows and keep the T and 3He in the tank if possible during long runs. Once the tank is basically outgassed (and I keep it that way most all the time) then I should be able to run really low flow rates and keep decent fuel purity. We'll see, the mass spectrometer won't lie.
Why guess when you can know? Measure!
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by Chris Bradley »

Oh - I get it, you've just wedged the parts together, no [spot] brazing or anything. You don't fancy any screws of rivets to hold it together?
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

No need so far, so why bother? Yes, it's all done by tight fit, which is not real hard to do with copper.
Sticky stuff.

I cut the butterfly and the outer ring from the same piece (with a cold chisel)! True cheapness. So they both got a little curl on the edges which makes the tight fit thing work fine -- the curl accommodates the pipe wall thickness and stiffens things. I may add a screw to hold the baling wire pivot/axle as I don't really trust the solder that much. I've had it melt, for example in my pinhole camera which did get fairly close to the action in there. This will be sitting down on the tank bottom though and should be "safe" as well as yet another thing to keep loose bits from falling into the pump easily.

If I need to tighten it up for less "vacuum leak at idle" then I will braze the ring, and make the butterfly fit better, there's spare metal there I can get just by making it straighter and then filing for tight fit. I can also add ledges it can rest on in the closed position if needed, but they make it harder to put the butterfly in there, so I didn't do it yet. I avoided hitting any of this with high temps so as to be able to skip the step of cleaning all the resulting oxides off it -- the two things I have that work for that are dilute nitric and concentrated ammonia -- neither are that much fun to work with.

I am really hoping this will give enough extra control without all that stuff, or having to screw the ring/flange to the tank -- that'd be a job of work indeed (tapping holes on that ship inside the bottle!). I can most likely spin the turbo down *some* and still have my nice auto-forepump thing working if needed. But the main idea is to skip that step -- it takes a loooonnnngg time for a new turbo to spin down to "standby" if there's not much gas in there. Much longer than it takes to spin up.

The point is, of course, to get that line on the mass spectrometer at "6" to show better after a long run.
We do actually see one, and it only happens after a long run....and it can't be much else than T2.
We always see a ton of "3" as we have D and there's always H to make HD, so that line doesn't say much and is real big as you'd expect -- you'd never see any much change from the very scarce reaction products. But this 6 line is very intriguing indeed -- even at the ~e-13 partial pressure level we see it at. In a long run (say half an hour), we start to see clutter down there where T and various combos of other H isotopes would be as molecules, so we think we've got something fairly cool going on (or hot to be more literal). About time to make that exhaust pipe for the forepump gas to go outdoors. Or at least pass it over a geiger during runs. We don't see anything at those mass lines when not proximate to a real neutron making run, by the way. Just 1, 2, 3 and only the 3 when there's been D in there recently. We've never seen 6 before, and we do see carbon normally from C) and CO2, and some plain C, but never 6 (which could be double ionized C, but evidently isn't). Now we are seeing both 6 and 4 sometimes, interesting indeed, the normal case (no recent run) is that those are both at the noise floor at about e-14 partial pressure on the mass spec. Maybe I should have sprung for the fancier mass spec head with the channeltron detector in it.
Why guess when you can know? Measure!
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by Chris Bradley »

Well, if you're getting a million fusion neutrons per sec, then you're surely getting a million tritons per sec, no question!! Just run the maths and see if your ms is that sensitive and if it ties up with your neutron counts. I'd have thought many fusion-energy-product tritons would tend to embed themselves in the chamber walls, though. Maybe also check for outgassing contaminants, if you can actually pull down to -9 following a run.

Is it possible to generate an H2O(3+) ion?
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

Yeah, I'll have to run the math. The tiny numbers on that 6 line make it believable at any rate.

A lot of bizarre things can happen in a high vacuum (molecular vs viscous flow) around species that would not be stable after even one collision with another atom of any sort -- we see all sorts of oddball things on a mass spectrum that have to be stuff like that (and literature backs that up). We've not seen anything like H2O+++ yet, but we think we've seen some other oddballs like H3O+. Takes awhile to learn how to interpret these things. Much tougher than a gamma ray spectrum.

Weird combos with single ionization are a lot more common than multiple ionized things. Almost as though a single ionized normal thing happens to pick up another atom from somewhere due to it being kind of attractive. That seems to be the norm. But that 6 line -- that's completely new, have only seen it after the recent very high output runs, never in any other situation. So, what it is has got to be of interest. Could be D3+ perhaps, but we've not seen that before in the more normal runs that more or less duplicate the other results posted here. Just in this new super high Q mode.

I'll have to run the math, but it's kind of daunting at this point because......we can't turn on the mass spec at fusor pressures (well, we can but you don't see squat in that mode, and it's hard on the filament), we have to do that during the pumpdown after, so it's at best going to be a pretty relative thing indeed -- much has been pumped out already, and the pressure is changing fast (maybe the new valve will help there).

And the readings are going to be quite skewed by the fact that the turbo pumps light gas far better than heavy. So even I'm a bit skeptical -- but it's tantalizing indeed!
Why guess when you can know? Measure!
Dustinit
Posts: 257
Joined: Tue Mar 18, 2008 8:02 pm
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by Dustinit »

Doug, the turbo pumps heavy gas much better than light gasses.
You could have a long hose between the turbo and the backing pump which will reduce the backing pump duty cycle but at the expense of longer pumpdown times.
ie : the turbo current will increase more slowly as it has a larger volume to pump into.

The mass spec could run but would require a separate vacuum system and bleed from the chamber which puts that out of the running I suppose.
It s a good thing you have nothing on mass 6 so make sure you keep all lithium away from it. That stuff contaminates like crazy and is almost impossible to clean out. I ended up tossing the cones and lenses from our system in ourder to try to get back to background noise levels.

I'm interested to see how your throttle goes, I need something like that. I just have a kluged switch on the turbo controller I toggle to try maintain pressure.
lutzhoffman
Posts: 188
Joined: Mon Dec 07, 2009 8:59 pm
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by lutzhoffman »

Very Impressive! It should work, and the copper is so soft that it should deform in the seal zone just fine.
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

Dustin, I thought I said that the turbo did that, if I misspoke, sorry. I think differential pumps also preferentially pump heavier gases. I can't notice a difference on my fore-pumps though.
This is of course a wonderful thing for those of us wanting to work with relatively pure light gasses, we have a purifier built in for us.

Interesting on the lithium as I do plan to do some work with it at some point. I do have several systems...I will take that to heart. I would think something so reactive wouldn't be hard to clean out of stuff, but maybe it makes compounds that degrade easy or kind of diffuses into other metals to come back out slowly? If you leave it around in air it gets gone pretty quick, into LiN and LiOH, finally LiCO3. I keep having to renew the argon in the bottle of it I have.

I have tried various foreline reservoirs (and in fact I do have a long hose there now, 1 meter kf 40) but the real problem is that when the turbo is "spun down" low, the forepump line pressure starts to dominate the overall pumping speed, so when it cycles, the tank pressure goes nuts. It's also "oversized" for this system in practice, if not in the catalog. I get from STP to e-6 mbar in something like 3 min total. I tried adding an expensive gate type valve in that line, but that didn't work out well at all. At those presures (evidently they are still high with the turbo at slow speed) the valve may as well be an on-off switch, can't "tease" it to the near closed condition and get any throttling to speak of there.

The mass spec *was* on another system at one point, and I have that one still, but it's more a situation of having space to do that thing (I have several little labs vs one big one), and then losing that other system for its other uses. I have dinged it once (and Pfeiffer was nice about fixing it cheap) in moves -- so that's not real high on my list, though I know *now* how to move the hyper fragile head without hurting it....I think. It turns out that they don't work as well if not nude, and they need a special large ID 2.75 CF extender to work in (larger than what Lesker lists as "large bore". So now it's happy where it is...One reason I don't want to put it on another system just now is that I want to do ion source testing with the spectrometer ion source turned off -- so I need a real straight shot into the main tank to see pre existing ions before they recombine on tank walls.

Richard mentioned once that anyone running a turbo kind of lives in fear -- not here, we have a couple and no problems. But having a new mass spec that costs even more and which HAS been broken by a light tap at the wrong time -- now, that's scary. I have it tucked away where lab traffic won't bump the thing now.

Toggling the turbo might be good, the controller doesn't make that real easy -- I never did get the external input to go back and forth with standby mode to work right. At least the inertia there would make it more controllable. I'd been toggling the forepump, but in that case, it takes a couple seconds to spin up (they used a slow start circuit from a conveyer belt on the XtraDry) and then is much too fast -- about the time you hear it move at all you had to get back off the switch, and more often than not would overshoot the target.
Why guess when you can know? Measure!
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

Lutz, yes, that's the idea. The real trouble is that you can pressure-weld copper even at STP if you're not careful, it has a big tendency to stick and gall. I don't *think* that's going to mess me up, but I plan to get it in there this weekend and we'll know soon enough!
Why guess when you can know? Measure!
MSimon
Posts: 248
Joined: Thu Mar 22, 2007 11:27 pm
Real name:
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by MSimon »

You might consider PWMing your motor supply and keeping the voltage at 12V. You get the torque without the speed.
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

Yes, and the microprocessor family I use and like has built in PWM for things like this. I will when I solve any other problems. For now, what I lashed up was pretty simple too -- I used a 1k R in series with a 12v wall wart, charging a 47 uf capacitor, which is just enough to "jog" it about 1/16" initially, then run it very slowly for finer control with nothing fancier than a DPDT switch with center off (to reverse polarity to the motor). Works like a champ on the bench. So, tomorrow, or this weekend, we get the big test as I have to get into my tank anyway for other reasons, and this will go in there too for real testing.

I did a nice PWM control in a similar situation for the X axis drive on my milling machine....I am familiar with the technique.
Why guess when you can know? Measure!
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

OK, nothing beats a guess like a test, and now it's tested. The mechanics came through with flying colors. I used a very simple drive circuit, which was a wall wart with about 12v DC output, driving a series R of 1K, 1W, with a 47 uF cap to ground after it. This provided just enough full voltage energy to ensure the motor would kick over a revolution or two, then slow down so it was more controllable. I used a simple 3 position, center off, DPDT switch to control this and provide either polarity. Works like a champ, no failures cold or hot. The mechanics did not seem to add to the outgassing significantly, though choking the pump down to 4" from 6" and adding that much pipe length does seem to raise my base pressure slightly. I had it in there during some fusor runs that heated the tank up a good bit, perhaps to 150f, and there were no problems with that.

At pressures solidly into "high vacuum" or more accurately, molecular flow, this gives about a 10::1 control ratio for a given input gas flow. That would be for pressures up to the e-4 millibar range or so. At higher pressures -- where we want to run a fusor, it doesn't do as much. As we approach viscous flow, the gas finds ways to push past the small gaps anyway, which means I still have to spin down the turbine to do fusion at low flows.

I don't want that, because a slower moving turbine means the forepump cycling does affect the pumping rate a lot more than with it at full speed, and this makes gas handling harder to do.

So much for the words "it has to be short, fat, and shiny or you'll never get a decent vacuum". Depends on what you mean by "decent", it seems. At e-3 to e-2 millibar, this makes very little difference at all (with the full range of turbo and forepump speeds having been tested).

I have minor leaks both around the butterfly, and around the ring base, which had to be bent to fit into the tank, then re-straightened. My guess is the the former is the worse of the problems.

Therefore, out of the tan it comes again to better fit the butterfly, perhaps add lips for it to rest on when closed, and braze on the ring, not because that is a major leak, but because it's a leak at all and pretty easy to improve upon over the mostly grip fit (with gaps).

So the s tory isn't over, just the first chapter. More to come, I refuse to have this not work.
Why guess when you can know? Measure!
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

Now it works much better. I brazed the tiny gap between the "flange" and the pipe, and added a lip for the vane to rest on in closed position, and now I have fine control over flow -- it's all dandy, it was just a little too cowboy to work at first. It now has significant control even up at fusor pressures (getting into viscous flow pressures at 2e-2 mbar indicated on D) and I can leave the turbo running full speed and let the forepump cycle on and off, and the tank pressure stays completely constant while that is happening.

This is joyful, finally the thing can run pretty much hands-off...whew.

Doug
Why guess when you can know? Measure!
lutzhoffman
Posts: 188
Joined: Mon Dec 07, 2009 8:59 pm
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by lutzhoffman »

Hello:

I love home made, do not get me wrong, and Doug you simply are a genius, but while on the subject check out the commercial version on ebay for only $48.00, another like it has a starting bid of $28.00. Take Care...Item #230435222603
User avatar
Mike Beauford
Posts: 419
Joined: Wed Apr 25, 2007 2:24 pm
Real name:
Location: Morton Grove, IL

Re: A homebrew vacuum conductance valve for in tank use

Post by Mike Beauford »

Hi Doug,

Nice work! Nothing like doing it yourself to make it taste that much sweeter!

Mike Beauford
Mike Beauford
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: A homebrew vacuum conductance valve for in tank use

Post by Doug Coulter »

Lutz, I get my srcounger creds questioned because I have no problem buying things that fix any problems around here. In this case, I didn't have room for the commercial version between pump and tank -- it's a jungle in there, and at that time, when I had needed the thing too long to stand anymore, this wasn't up there to be had.

Here's a pic of the underbelly of my main fusor. The turbo almost hits the shelf beneath (and if the shelf wasn't there it would almost hit the top of the forepump. In the pic it's under vacuum and the bellows is max compressed (I had to add springs across that to make it stiffer so it wouldn't have other interference troubles with the flanges on the tank bottom for other feedthroughs).
You can see the mass spec in this picture too. I hung it there to keep it out of harm's way.
The PC is the one I built for data aq and control, which is getting along. The other thing in front of it is an Allen-Bradley process control computer with tons of A/D and D/A and other stuff that's going on there soon. The beaker is a refil for the water resistor I use sometimes. There's just no space for another thing in that chain, so this had to fit in the tank -- it's more or less inside the bellows.

And this *should* have been easier, but hey, it's done, it works, and now the fusor can be almost hands free in all the happy modes -- ran a bunch last night and it's all much more stable (and learned a few things I'll report if/when they replicate here).

But yes, having made it myself does make it taste sweeter, Mike! And for more than one reason. Another is that I have and can get parts for more, I really can maintain this and not depend on uncertain sources.

I maybe swung my personal pendulum a bit over on this class of thing. For many years I designed things (hard and software) but all of them were in places where I couldn't point to them and say "I did that" -- in fact if you noticed my stuff at all, it meant trouble of some sort -- the phone or PA doesn't work to infinite 9's, or this medical prosthetic or piece of a spy satellite or military gear of some other sort.

Now I can design stuff, make it the same day -- including the tools to make it as needed, and point to it and say "I did that" which is saweeet indeed. Broke critical mass on all that and it's great!
So of course, any excuse...
Attachments
Underbelly.jpg
Why guess when you can know? Measure!
lutzhoffman
Posts: 188
Joined: Mon Dec 07, 2009 8:59 pm
Real name:

Re: A homebrew vacuum conductance valve for in tank use

Post by lutzhoffman »

Just to clear it up I was not questioning, I was just pointing out the cheap valves on ebay just in case that someone needed one to try to be helpul. I am in your camp as far as homebrew goes. The other cool thing about homebrew is that it can be a lot of fun! To engage the mind, and then to come up with a simple solution from available parts is always something to behold : )
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”