Page 1 of 3

Langmuir Probe Creation and Design

Posted: Fri Apr 02, 2010 4:00 pm
by bk8509a
In advance: This is going to be a long post, but pretty much a intro on how to build a Langmuir probe. As I do the rest, it will follow.

I browsed around on the site to see if anyone here has produced a Langmuir probe. There were a few 'don't try it wont works' and thats about it. To my knowledge, I don't think anyone has done this yet here. Its simple and cheap, and that's what we are all about right?

I started reading a few papers on the theory of Langmuir probes. Then I figured that I wanted instantaneous measurement, so I got a paper " Sin-Li Chen and T. Sekiguchi, J. Applied Phys. 36, 2363 (1965)" which is on the Wikipedia site for Langmuir probes. For anyone interested in doing this, start there.

To the meat:
The design is based around a 4 BNC feed through which is attached to a 6" to 2.75" zero length reducer flange. Each BNC tip has a beryllium copper wire grabber on it. Three of these be attached to the three wires from my three probe tips.

The probe itself is made out of .187" diameter alumina tube that is 5" long. The tube has 4 bores in it which are each .047" in diameter. Down three of the bores are .047" diameter pure tungsten welding rods. They are attached to the butt end of the alumina with high vacuum epoxy.

The probe is attached to the zero length reducing flange via a home made aluminum shield which will protect the connecting wires and also the exposed conductors from creating a false reading. The aluminum housing holds the probe with a set screw and is attached to the zero length reducing flange via holes we've tapped into the flange.

The design of the probe allows a melted/cracked/destroyed probe tip to be unscrewed and replaced for around 10 dollars.

I hope that the probe will be in by today and taking measurements of the ebeam in the end of jet mode next week. I will post my data from those runs.

If you are interested, don't hesitate to ask. Criticism is also encouraged, as that is how we learn.

Re: Langmuir Probe Creation and Design

Posted: Fri Apr 02, 2010 5:37 pm
by Chris Bradley
I had rather presumed some had tried this, but that there's not much to measure in a fusor.

I have tried out similarly with my ion magnetron "crimson doughnuts" and I have obtained some degree of consistency in the results, though I think with any experiments like this there is the question of whether Langmuir probe theory applies. For mine, possible non-neutrality and the magnetic field are going to throw up plenty of question marks over any results. In a fusor I don't really see what you would be measuring, maybe the central plasmoid or the general distribution of the background. The beams themselves are all [likely to be, in any simple analysis] rather tenuous to say the least, and I can't see accelerating electrons already headed for the shell (/ions towards the centre) to be hanging around a Langmuir probe for long enough to slow them up and measure a saturation current.

But we'll see!! Experimentation is experimentation, after all. Sometimes the theory just has to tag along *behind* the experiment to explain what is observed. This is true of plasma science more than most! We look forward to seeing your results!


Re: Langmuir Probe Creation and Design

Posted: Sat Apr 03, 2010 2:54 pm
by Chris Bradley
I'm currently experimenting with ways to make robust Langmuir probes, myself, and I have a choice of tip materials. I was tending to stick with stainless steel as it has good all round properties at an insignificant price. I have pure tungsten (not welding rods) and was considering using this, but I'd have thought it more likely to create an oxide layer on the tip, compromising the results? Or would a low-sputtering option like aluminium be preferred, which surely will form an oxide layer but that will be 'blasted off' quickly?

I'm gonna stick with SS, unless anyone has good, knowledgeable advice on this.

Re: Langmuir Probe Creation and Design

Posted: Sat Apr 03, 2010 3:26 pm
by bk8509a
Wouldn't the oxide layer immediately get sputtered off? All the literature I've read has the probe with a Tungsten tip.

Re: Langmuir Probe Creation and Design

Posted: Mon Apr 05, 2010 7:00 am
by DaveC
You should have some interesting experiences with your measurements.

My only thoughts on the process are first one needs to keep the mass (physical size and area) small if you are going to poke the probe into the ray.

More important to the answer you may find, is what voltage you allow the sensor to reach.. If the probe is a ground potential, (same as the shell) then it would more less strongly affect the electric field geometries depending on its position.

If the probe is at other potential and polarities, ( plus attracts electrons, and negative will collect ions.) it will collect currents of whatever is flowing. I have seen rather small probe voltages cause quite large changes in pickup current.

[Edit: Above to replace electrons with ions.]

You can of course, use this to construct an "energy spectrum" of the particles in the rays.

Since the rays are glowing, one would expect ions and electrons. So appropriate voltages on the probe should collect current of the appropriate polarities.

If you wish to poke around in close to the grid, the probe will need to be biased. Some times it may be that probe bias voltage can be controlled to give zero current pickup, thus identifying the exact voltage in the ray at that point. A closed loop controlled low power voltage multiplier is probably the simplest way to go there.

It will be interesting to see your results.

Dave Cooper

Re: Langmuir Probe Creation and Design

Posted: Mon Apr 05, 2010 1:58 pm
by Richard Hull
This is an ancient endeavor with the fusor and goes back to the original ITT effort in 1960-68.

The spherical focus nature of the device sort of obviates this approach in crticial data gathering as the introduction of a probe distorts that which is being measured. Other methods of probing and data collection in zones of the fusor have been cleverly tackled by Miley's group, Urbanna and others at Univ. of Wisconsin.

A check of the literature and search of the past URL links on this material will speak to these issues.

Richard Hull

Re: Langmuir Probe Creation and Design

Posted: Wed Apr 07, 2010 10:52 am
by Hector
I tend to agree with Richard that the probe will interfere with the fusor plasma process and thus give less than true measurement of the plasma environment. This is why you never see the University teams used them on fusors.

Don't get me wrong, you never really know until you try, but you might want to consider using Laser Induced Fluorescence (LIF) instead, because it's far less disruptive to the plasma process of the fusor than any other probing method.

I have a Thesis done by a researcher in the 90's that used this method with good success to map the potential wells inside a fusor. I believe Dr. Millie and team at Wisconsin have done the same.

Just a suggestions, like I said you never really know until you try.

Good luck.

Re: Langmuir Probe Creation and Design

Posted: Fri Apr 23, 2010 5:25 am
by bk8509a
My first series of data is in. I've taken measurements with a probe an inch outside of my outer, grounded grid. I wasn't sure of what voltages I was going to find or if the probe was going to get too hot, so I put it in the safest place in the fusor. The measurements are interesting, to say the least.

The most surprising thing for me is that electron temperature goes down as the voltage increases, until the x rays start hitting the tungsten and creating a voltage that way. As of now, I believe that the temperature decreases due to the fact that the probe is outside of the grounded grid, but i'm really not sure why this is happening. I have to do a few more tests. For all I know, the probe is not set up properly, or I'm not following the theory from the triple probe 100%, but I'm almost sure its working due to the x-ray excitation of electrons at 10kV (and the fact that its a super simple system).

Another disheartening problem is that as I increase "V d3", the curve should stay the same. It doesn't, which I'm blaming to the non-uniformity of the electrons coming out of the center of the fusor. As the Vd3 term goes up, i'm guessing the probe can pick up more electrons gaining a better average. I'm suspecting that the curve will converge around Vd3=50 V as it will be able to pull a nice average set of electrons.

A few other weird points. Check out that little bump? Where the hell is that coming from?! It happens around 17 mTorr (5KV) each time. That's right when my e beam jet is fading.

Thoughts as right now: the electron temperature increases RAPIDLY in the lower voltages, which I thought it would, but then it declines? WHY? Remember, this is not in the poissor, but outside of the grounded outer grid.

All theories accepted.

BTW, the triple probe paper I went off was

Chen, Sin-Li, and T. Sekiguchi. ``Instantaneous Direct-Display System of Plasma Parameters by Means of Triple Probe.'' Journal of Applied Physics 36.8 (1965): 2363-375. Print.

Would post it, but its too big.

Re: Langmuir Probe Creation and Design

Posted: Fri Apr 23, 2010 6:34 am
by Dustinit
The clue lies in your last plot.
The voltage on the probe determines what range of energy electrons can have that it will detect. As the voltage on your probe is quite low it can only detect low energy electrons. Fast electrons will whizz past without noticing the small field of your probe. As your fusor volts increase so does the average energy of the electrons so you detect less. It only starts to increase again because electrons hit your outer grid releasing lower energy secondary electrons which your probe will detect. In order to measure the higher energy electrons you have to increase the voltage on your probe.
Your probe will only detect electrons of lower energy than the probe voltage.

Re: Langmuir Probe Creation and Design

Posted: Fri Apr 23, 2010 6:57 am
by Chris Bradley
I'd be interested to see the raw data of your probe measurements. Can you post up a typical set of measurements from your runs (in linear(current) and ln(current))? Did you get to a steady ion saturation current?