first I would make a better gas delivery system, I think that just putting the tube there is not sufficient (maybe put the whole experiment in a box). Then, do you have a ballast resistor in your circuit, which limits the current?

How is the argon flow controlled?

Dear Ahmed,

The I-V characteristic of the gas discharge tubes is in form of an S-curve. To set an operating point which is stable you must use constant current sources rather than voltage sources. So, You have to stabilize the current in tube at the required value.

The best operation is achieved by a constant current sources. Constant current sources can be built by transistor with the collector or drain current passing through the gas discharge tube and the current is controlled by the voltage on the gate or the base current according to the transistor type.

It will work ISA.

Best wishes

@Dr, Johannes, thanks for your comment

• The gas is delivered from a Cylinder, there is a gauge that shows the flow rate, and the tube is pointed to the electrodes as shown in the photo.
• A box is not applicable due to moving objects.
• We've a normal resistor in series with the discharge.

@Dr. Abdelhalim, thanks for your comment

• I'll try the current source and come back to you.

@Ahmed Magdy Abdul-Wahed

so, you just measure the flow, but you do not stabilise it? - that might be a problem too.

If you have already a resistor in series with the discharge, it's not so likely that the current and/or voltage run away. So I am not sure, if a better power supply will lead to further improvement (but try it to be sure). How large is your serial resistor?

If you have moving parts, you can just use a bigger box ;)

@Dr Johannes, thanks for the reply

In fact, today I've used a box :), shown in the figure, the plasma was more stable (to some extent) and it was maintained at larger gaps between electrodes.

• I used commercial copper wire, is there any way to make the discharge starts from the tip of the electrode (not the side) ? 
• In addition, do you have a clue how can we prevent the copper wire from extra heat generated due to current?
• Finally, plasma don't strike until both electrodes are in contact, then the copper electrode is lifted up, if I applied the voltage while the electrodes are separated, the gas doesn't break down, why ? How to generate plasma without contact? (See the video)

• I would actually expect the discharge to start from the tip due to the tip effect of the electric field - maybe you got some dielectric dirt on the tip that increases the resistance there (try to clean the wire as good as possible).
• To prevent the heat is impossible (except if you are lowering the voltage and current due to ohmic heating). You could try to use a thicker wire with are sharp tip as the higher cross section area decreases the specific resistance. Another point is heating due to the plasma itself, as it can be in thermal equilibrium, which means that it is very hot and thus evaporating the surface of your metal electrode.
• This is due to the Paschen curve (you can google that). The breakdown voltage is proportional to the product from pressure and electrode distance (as soon as the plasma burns, the maintaining voltage becomes much smaller). So, if you want to have the plasma without the contact, you have to increase the voltage, but that will again increase the heatload on your wire.

Paschen curve says that for argon at atmospheric pressure, you will need less than 500V and I'm applying more than that.

I had experience in argon gas with different gas pressures . Plasma was formed very stable plasma. Experiment has been  done in a closed reactor . Information can be found at:http:

http://iopscience.iop.org/1742-6596/495/1/012041/pdf/1742-6596_495_1_012041.pdf

@Ahmed Magdy Abdul-Wahed:

The Paschen curve vor Ar says 500 V at 0.3 [Torr x cm]. This means at p = 760 Torr (atmospheric pressure), 500 V are sufficient for a gap space of ~ 3.95 x 10-4 cm. That's why your electrodes have to touch in order to ignite the plasma.

Thanks guys, I'll keep working and inform you about the updates :)

Dear Ahmed,

I would like to thank you and Johannes for this persisting dialog. 

But i would like to remember that the breakdown phenomena in materials including gases are difficult to initialize and sustain on large cross sections of the device.

It is so that the electric field must be equal or greater certain electric field to affect the breakdown.

The fluctuations in the material properties from a location to the other causes  fluctuations in the critical field.  These fluctuations are statistical leading to a  narrow break down channel. The situation will be more pronounced with increasing the gas pressure.

In order to define the path of the discharge you have to shape the electrodes and to align them well.  If you want to enlarge the cress section of the discharge you have multi tips electrodes such that they form a two dimensional array of tips.

Naturally, you must provide the suitable power supply for the discharge. It seems me that you passed very rapid on my tips.

The gas discharge tubes have an IV characteristics in the shape of S- Curve. The best supply is the current source. If you increase the current of such supply from zero the voltage on the tube increases reaching a maximum which is the break over point at the start of the breakdown . By increasing the current further, the voltage on the discharge tube decreases reaching the so called sustaining discharge state.

So, your current source must be able to deliver the turn over voltage and the sustaining current.

You must be aware that the gas discharge tubes are highly stressed and the electrodes will get hot. So, you may cool the electrode.

wish you success

Thanks Dr Zekry for these valuable info,

 will try the current source and come back to you 

Merhaba Ahmed,

• I think you need a MFC (Mass flow controller)
• A serial resistor with cooler
• A modulated power supply can be better than a dc PS to control the glow
• Grounded Faraday cage will be suitable for glow tip
• Cooling is very important for system and you can apply a few solution like water with little pump, peltier or fan.

Aluminum mesh or steel mesh - make certain it is as flat as possible. Al will also passivate whereas steel will flake with time if in an oxidizing atm. Can be reduced to some extent in pure Ar. If your electrode is a point or cylinder the discharge seeks a fresh site (since its not thermonic ) it tends to rotate looking for lower resistance. Please note: I am not at SJSU my association with them ended in 2005. Glow btw occurs at low pressure low current. 500VDC seems low i would up convert voltage using whatever I your supply has available. A PVM500 unit is relatively cheap and offers wide range of frequencies adjustable voltage so your options increased. It handles resonant and resistive loads. Google PVM500 It also has a voltage doubler built in. Great for sputtering depending on electrode & distance. IMHO