Page 1 of 1

Anode Layer Ion Source Mod?

Posted: Mon Jan 23, 2017 3:32 pm
by David Kunkle
hjuklghjk 002.JPG
One of Andrew's ion sources arrived here Friday. Got to thinking exactly how I was going to mount this. What if it was mounted/welded directly to the front of a CF or NW nipple? In the photo, the weld would fill the gap between the ion source and the inside of the CF nipple. This would force all input gas thru the ion source and eliminate piping from the gas input connection to the source. It would just be an open space behind the source for the gas to be fed into. It would also eliminate the standoff. It might make connecting the HV spring to the source a little trickier to install. It would have to line up and keep contact with the source as the back flange is being mated and sealed. Easily checked for continuity after assembly with a DMM.

Re: Anode Layer Ion Source Mod?

Posted: Mon Jan 23, 2017 8:34 pm
by Andrew Seltzman
I would recommend going with a non-welded method. Consider using a viton o-ring between the source and the nipple, that would effectively seal the source, and hold it in place due to the elastic nature of the o-ring. A gas feed going through the source is not required for operation, it will ionize the neutral background gas and form a good beam between ~15mTorr down into the 0.1mTorr range(this is how I use them on my fusor)

Consider using a hermetically sealed bulkhead BNC of MHV connector as well(bulkhead mount with o-ring, not the welded type)
Not sure how high in voltage this will go in a vacuum, but it is pretty cheap
http://www.ebay.com/itm/Pasternack-PE46 ... Sw14xWOsEK

Also something like this could be drilled/tapped for the standoff:
http://www.ebay.com/itm/2-75-4-Pin-Plas ... SwHMJYLxs2

Re: Anode Layer Ion Source Mod?

Posted: Tue Jan 24, 2017 3:08 pm
by David Kunkle
I thought an o-ring would get too hot there. I'll have to find the right size and try that first. Lot less permanent than welding it. In light of that, I already have an extra CF blank that I'll use for the base. Took it to the weld shop this a.m.

I noticed a long time ago it looked like it would simply recirculate the gas from the chamber the way it was made. The o-ring should do the trick. There are a few times I will want to have the vacuum lower than the gas pressure to the source.

Thanks for the tip on the bulkhead BNC and the plastic feedthrough. Anybody know what kind of plastic they use for those? Mostly curious, but hey, you never know. Wouldn't be hard to fab something out of a sheet of that.

Is the alumina gas feed just loose between the input fitting and the source?
Wondering why there is one hole left in the anode?

Thanks.

Re: Anode Layer Ion Source Mod?

Posted: Tue Jan 24, 2017 5:40 pm
by Andrew Seltzman
If you choose to go with welding, the casing material is 410 stainless in case they ask. Viton o-rings can get pretty hot without problem.

I run my injectors ionizing/recirculating the ambient deuterium. This will typically will work well unless you have a large impurity source/outgassing.

The plastic on that base plate on ebay looks like PEEK, though I can't verify that

The alumina tube in the injectors with baseplates is loose(not o-ring sealed) in the 4th hole in the ion source base, but mechanically retained with the snap rings, the conductance through the tube center is much higher then the conductance around the ~0.001" clearance between the tube and injector so gas would flow thought the tube into the injector.

The remaining hole in the anode was part off the original design of the injectors to accommodate welding a 4th threaded rod to hold the anode ring, however this was found to be unnecessary as the 3 threaded rods in the current design hold the ring in position accurately.

Re: Anode Layer Ion Source Mod?

Posted: Wed Jan 25, 2017 10:36 am
by David Kunkle
Got it all. Thanks.

May have given the wrong impression. Took the flange to the weld shop to have the holes drilled and tapped- no welding. Not a big fan of working on SS. Main problem is probably cheap drill bits and lack of enough lubrication in the past.

Ordered a few different size Viton O-rings from o-ring store online. One should fit well. Looks like Viton is good to about 400F.

Grainger has everything from rod to sheet stock in PEEK in case anybody cares. Must be one of the better plastics for vacuum use? Good to 480F. Now that I look closer at the prices, SS is looking pretty cheap. A measly 6" X 6" X 1/2" thick piece is $350! Anyone for a 48" X 12" X 2" piece?- a mere $16,000.00.

Got curious and went to Wikipedia: "Because of its robustness, PEEK is used to fabricate items used in demanding applications, including bearings, piston parts, pumps, HPLC columns, compressor plate valves, and cable insulation. It is one of the few plastics compatible with ultra-high vacuum applications. PEEK is considered an advanced biomaterial used in medical implants. It is finding increased use in spinal fusion devices and reinforcing rods. It is extensively used in the aerospace, automotive, and chemical process industries.[6] PEEK is often considered a high-end engineering plastic, such as delrin, PTFE or nylon. "

I see the snap ring on the gas tube now in the one photo on your website.

Any point in filling in that 4th hole? - or should the gas coming out of there get ionized as well?

Andrew, I can't find it now, but did you post some info on power usage and % ionization for these?

Thanks.

Re: Anode Layer Ion Source Mod?

Posted: Wed Jan 25, 2017 11:48 am
by Andrew Seltzman
Any chance the weld shop could TIG welds a MHV connector into the flange for you? There is no need for an external gas feed so just having them weld in the MHV connector would give you a first rate base plate for the injector.
You can get them for $38 on lesker
http://www.lesker.com/newweb/feedthroug ... ?pgid=weld

There is no need to fill the 4th hole, it does not affect operation.

About 50% of the current into the injector goes into the beam
viewtopic.php?f=12&t=5011&start=20

Re: Anode Layer Ion Source Mod?

Posted: Mon Jan 30, 2017 10:21 pm
by David Kunkle
The shop had already drilled the base plate, etc., so I stuck with my original plan.
Now that it's been fired up, the 4th hole doesn't appear to affect it like you said.
My source is at about 1kv, 5 ma, and 5 mTorr in the pics. Still playing with settings. No Viton o-ring installed yet. Can't feel any heat generated after several minutes, so o-ring should survive just fine.
Camera's not great- best I could get.
dxufhjig 005.JPG
ghikmh 001.JPG

Re: Anode Layer Ion Source Mod?

Posted: Tue Jan 31, 2017 1:11 pm
by John Futter
David
no at that drive it will only get warn 1kV x 5mA = 5 watts
Its when you start accelerating the ions so same source but 60kV and 20mA that things start getting vey hot very quick in this case 120 watts and you get a minute or two before the magnet overheats and the source stops all by itself. We use a kerosene cooled copper block inside the anode layer source to keep the magnets cool which allows operation up to 100kV @15 -20 mA ie 1.5 -2 kW of heat

Re: Anode Layer Ion Source Mod?

Posted: Thu Feb 02, 2017 12:52 pm
by David Kunkle
5W- hadn't thought of it in those terms yet. About the equivalent of a turn signal bulb in a Buick.
If I recall, I started getting some intermittent arcing when I tried to go over 2kv. Don't think the little alumina insulators would care for 60kv let alone 100! ;)

Re: Anode Layer Ion Source Mod?

Posted: Thu Feb 02, 2017 8:06 pm
by John Futter
David
no the ion source only has 1 -2 kv across it the rest is 58kv of acceleration
I use self bias of the source to get the 1-2kV which is stolen from the terminal voltage applied in today's case 60kV
and I'm running the source at between 3 and 12mA so thats 180 to 720 watts and yes the samples being hit by ions get very hot. The ion source has 4 gallons per minute of kerosine cooling that is heat exchanged with our chilled water circuit I use a petrol bowser pump to shift the kero around the cooling circuit

Re: Anode Layer Ion Source Mod?

Posted: Thu Feb 02, 2017 10:13 pm
by Dan Knapp
Why would the full power of the accelerating voltage x the ion current be dissipated totally in the source? I would think only a fraction of it in the source.

Re: Anode Layer Ion Source Mod?

Posted: Fri Feb 03, 2017 1:12 am
by John Futter
Because of the electrons returning to the ion source.
They go from earth in this case to terminal which is 60kV.
+ ions out of source equal number or more (due to secondaries) electrons going back the other way.
Simion, Comsol model this quite accurately and has worked in practice with the intercoolers heat transfer between kero / water that we use to cool the source magnets.
We use Alfa Laval plate heatexchangers as intercoolers. Our biggest anode layer ion source consumes 5kW (120kV 40mA )and stopping it wrecking a fortune of Neo magnets is not a trivial task

Re: Anode Layer Ion Source Mod?

Posted: Fri Feb 03, 2017 10:19 pm
by Dan Knapp
The energetic ions flowing out of the source are carrying energy which is deposited into whatever they hit. This energy is derived from the ion source power. Thus some of the power is being transferred to whatever the ions ultimately hit.

Re: Anode Layer Ion Source Mod?

Posted: Fri Feb 03, 2017 10:50 pm
by Rich Feldman
In the example of a 40 mA 120 kV ion beam, we are putting in 4800 watts of electricity. At first glance, we'd expect the associated heating to be partitioned between the "ion source" unit and the target, in the same proportion as electric potential drops.

As John said earlier: even 40 watts of power dissipation in the source unit is enough to overheat magnets in short order, without liquid cooling.

But as John said more recently, we can't ignore the fraction of 40 mA which is carried by electrons knocked out of the target and accelerated through tens of kV on the way back to e-bombard the ion source. Have I got that more or less straight?

Re: Anode Layer Ion Source Mod?

Posted: Sat Feb 04, 2017 12:48 am
by John Futter
Rich
Yes this is also confirmed by the x-ray flux coming off the ion source. In the 4800 watt case 600 milli sieverts/Hr @ 2 meters form the ion source hence the unit is shielded with a concrete wall 8" thick.
To get x-ray s from the ion source itself you need electrons- lots of them
My gut feel is that for this system we get 3 secondary electrons per arriving ion with one of those going back up the acceleration column all the way to the ion source.
Yes the samples dejour do get hot and this is mitigated with a rotating stage. I have burnt my self quite a few times by bringing the systerm back up to atmosphere (back filled with N2
) and getting them out too soon.

Something I have not got a handle on is the fact
When electrons leave, the cathode cools while the anode heats with the arriving electrons --this I read somewhere or was taught many years ago

Here is a link to the system some time ago
A new ion source since then but you will get the drift I'm presently using the purple door with the hundreds of resistors setting he acceleration plate voltages down the chain from the ion source
http://www.coultersmithing.com/forums/v ... 7&start=10