ICP coil first test

[Carl Willis]

Chris:

Since you're in a sanctimonious mood, feel free to append a liability disclaimer onto your own posts.

Suggested text:

"You are assumed to be ten years old until proven otherwise (although for all you know, I am ten years old). You are assumed to know less than I know about what you are doing (even though I have demonstrated no experience in your subject area and have a hilarious record of inaccuracy and noisemaking in general). You are assumed to be pursuing the most dangerous course of action possible within your means unless you explicitly state otherwise (and even if you did, it's likely I'm ignoring that). You must submit a deposition attesting to your comprehensive safety understanding before I will even attempt to engage you, and my response either way is extremely relevant and must be dignified with your attention or I will bellyache. Abstinence is the only 100%-effective method of preventing adverse consequences. People who tell you otherwise do not have your interests at heart."

Boiled heads, going blind, death? You guys can take comfort in the fact that you've covered your own asses (with a six-foot-thick nappy), but don't get idea that you're making a relevant or factually-based contribution.

That's all folks!

-Carl

[Linda Haile]

No need to act like you are ten years old Carl. Grow up.

[Scott Fusare]

If using an auto match you may also find it beneficial to ignite the plasma by some means other than the ICP coil. Finding network values that allow both plasma ignition and final matching can be quite challenging depending on the operating regime.

[Linda Haile]

My own research on this subject suggests that Scott is absolutely correct.

[John Futter]

Scott lyn Steve
I presume that steve is doing this at partial pressure and therefore you shouldn't need anything to light the plasma.
Indeed if this was to be done at atmospheric pressure then a small tesla coil or similar would certainly help getting the plasma lit.
All the invacui plasmas i have managed to get going have been lit by playing with the gas pressure ie give a bit more gas to get it lit then throttle back smoothly so the automatch keeps up and doesn't shut the generator down through high VSWR.
If Steve hasn't got good gas control then a small Neon sign tranny would be all that is needed but that will require an extra HV feedthrough

[steve_rb]

As John said I am relying on igniting by pressure change. Working pressure will be about 10-3 torr but first more gas can let in for plasma to ignite and the slowly decrease gas flow for auto match to keep up. Also planning to put one extra feedthrough with a sharp tip in close distance with ion source steel wall. By connecting this to a few thousands arc can be generate to ignite plasma. Also neon sign tranny can be used.

[Scott Fusare]

"I presume that steve is doing this at partial pressure and therefore you shouldn't need anything to light the plasma."

The partial pressure of what gas? It's the total pressure he needs to worry about...

At any rate, I was just trying to share my experiences with auto-matched ICP systems that operated in the 1 to 1000 mT range with a variety of gas mixes. Inductively coupled systems can be notoriously difficult to ignite. Your mileage may vary.

By all means do whatever works for you. Do let us know if and when you actually get it to work.

[steve_rb]

I will use argon and H2 gas for initial test and D2 gas later for fusion. Pressure inside the ICP chamber will be about mTorr. Chamber is on the process of building. One easy way I have seen some people use is a tungusten filament (220 Volt lamps filament for example). I can ask for an extra feed through for this purpose before chamber finishes. Strategy I will take is as follows:
With chamber pressure around 10^-3 I will turn on the filament and then turn on the RF (adjust around 500 W with about zero reflection) and then let gas pressure to increase gradually and I am expecting ignition at some pressure and then decrease pressure if ignitione happened and match impedance again to decrease Rf power and match accordingly. If no ignition happens I will increase Rf power to 600 W and repeat the process again. I will do this with higher watts untill I get ignition at some point to find the working condition. Tangustun Filament will be on all the time. The reason I am insisting on having a preior knowledge of this because I want to install everything I need on the chamber before it finishes. Also take all saftey issues not to hurt myself or damage RF unit.

[Linda Haile]

If this is a concern, Steve, It may be worth making provision for a CCP igniter within the ion source.

This will lead to some contamination though.

[Dustinit]

I think you will not need any igniter source and 500W is over the top for that pressure. I would think 10 to 50W would be more than enough to start.
IMO.
Dustin

[Linda Haile]

Not sure who you are addressing Dustin. If me, then we plan some experiments where we hope to utilise skin effect to advantage. We want plenty of power in reserve to test/verify our expected results.

This post may not make a lot of sense, but our experiment requires extra power in order to prove that the anticipated results are not linear.

I hope all will become clear in due course.

[Dustinit]

I have waged my own wars with skin effect. In this picture is my oscillator partially constructed with silver plated copper straps coupling the tank capacitors to the work coil (gold). The strap is 1.6mm thick and 14mm wide and is water cooled at both ends but the centre of the strap still gets hot. The skin depth is such that all of the current runs in the silver plating and the current is about 80A peak. There is a small non linear effect from essentially thermal runaway in the silver but is self limiting due to negative feedback effects. I fail to see how this could be used to any advantage. The plasma generated at right is argon atmospheric with yttrium sample injected at 1.2Kw.

[Linda Haile]

Dustin, I was referring to skin effect in the plasma itself. Sorry, I assumed you'd realise that.

I could elucidate further if you wish, but a google search may prove more enlightening.

(BTW, skin effect produces effects that look very like Chris Bradley's Crimzon Donuts.)

[Dustinit]

The skin effect within the plasma is not obvious but observable in an OES plasma instrument by scanning the plasma. At 27Mhz or even 40Mhz its barely visible at 1.5Kw. 2.4 Ghz it is easily distinguishable but still not very delineated. This seems to be because the plasma has resistance, the current flowing through the resistance is spacially diffuse giving a more diffuse magnetic field which gives a diffuse skin. Trying to enhance this is difficult especially at 2.4Ghz because as you pump more power in the plasma becomes more conductive which reduces the skin depth but at the same time the lowering in impedance reflects more of the energy you are trying to pump into it (effectively increasing VSWR). The same effects occur at the lower frequencies. The inductance and resistance of the plasma is what you are coupling to, pump in more power, resistance drops making it harder to dessipate more power. Its a bit catch 22 and is limited by the RF source impedance.
I'm interested to see if you get anywhere and what results you get.
Dustin

[Linda Haile]

Are we going off topic Dustin? Should we start another thread? I don't want to pollute Steve's thread.

Maybe now you can see why I want plenty of power in reserve, due to the increased resistance and other factors.

[Dustinit]

Sorry steve, I've been digressing on your thread.
Lyn, While interesting, and obvious that you don't want to share your ideas just yet,
I'm skeptical but content to sit and wait for anything you may later share.
Good luck.
Dustin.

[steve_rb]

John Futter wrote:
> Steve
> My guess is that your coil is lowZ the j notation = 2 x Pi x L (in Henries) x F in Hertz
>
> There are free coil calculators on the net ie turns diameter and wire diameter will give a fairly good approx in uHenries then use the above to get j Ohms.
>
> by off tuning I meant to start with the tuning cap offset from the tune point. One way will be much faster to get tune.
>

1- Following is the matchbox circuit along with the way I will connect my antenna coil. Currently by factory default tune coil tap is on point "C". For off tuning should I start with point A or B? Which one will automatch faster?

2- For calculating inductance my coil is not circle. My coil is rectangular shape with 26.5 cm long and 4.5 cm width and about 2 cm thickness. I have used effective diameter of 10 cm (=(26.5+4.5)/3.14) . I have calculated L for two following set:

turns diameter lenght
2 10 cm 2cm
1.5 10 cm 2cm

For case 1 I have got 0.6 microhenry and for case 2 I have got 0.34 microhenry. Both case according to the automatcher manual is Low Z as you mantioned but I am not sure about my choosing effective diameter for coil and validity of the calculator for my type of coil. Calculator is here:

http://www.66pacific.com/calculators/coil_calc.aspx.

I would like to have some opinion and advice from people here about the above 2 items?

[Linda Haile]

One would normally start at the mid-point, C. You don't know if it will be higher or lower, so start in the middle.

(I'm assuming the AMU will 'match' the system, and that your coil plus load will be within the operating range of the unit.)

Others are (hopefully) better able to advise regarding modelling the coil.

[morgoth31]

is your output coil matched to your supply?
the output capacitance of the gen is normally made for a specific inductance given the turns and gas used along with the size of the torch. (diameter)
this could cause the match to not be able to properly seek and get good match.
if you have the information it might help you get a better match.

im not trying to tell ya what to do but when i built and tested units they were created for specific torches so that the match would work. creating a stable plasma is hard enough but to light one with out a spark can be difficult at best normally you start at a higher power to light it and find the lower limit by turning it down till it destablizes and back up till it normalizes.

on the coil the easiest and simplest way is to use copper tube that is water cooled if your concerned about water conductivity a simple di filter will keep it clean if you start with distilled water. my best advice on filters tho is use a feed to drain bypass filter arrangement not an inline that could clog and cause flow issues.