Hi Dr. Sahu, 

Pretty straight forward, actually.

You have a matchbox between the rf source and the powered electrode, and there is a blocking capacitor (the series capacitor) as part of the matchbox.  Thus, in the classic setup, there is no net current (no DC current) on the electrode and the electrode takes on a DC bias voltage so that, averaged over an rf cycle, there is zero net current.

What you need to add is a low pass filter between your DC power source and the electrode, on the plasma side of the blocking capacitor. This needs to handle the full rf-voltage you have on the electrode.  This would look like a high-power rf inductor with one end connected to the electrode, the other end to a vacuum capacitor connected to ground. The higher the value capacitor the better -- you could use high-voltage ceramic-chip-capacitors here too (see ATC 3600V range...)   For the inductor I would suggest a 20-turn, 80-100mm diameter coil, with 2.5mm diameter wire and 5mm turn spacing.   This is like the series inductors in 2MHz matchbox. You should be able to attach your DC power source onto the inductor/capacitor junction directly.   Note that if your DC voltage is less-negative than your rf-dc-self-bias would have been, then you could get *lots* of current. Start by assembling the rf components (without the dc power supply connected) and check to see what rf voltage you still have on the capacitor-inductor junction.  If this is too high (too high depends on your dc power supply) you might need a second low-pass filter.

I expect to be in SKKU in May/June, I'll check in and see how you are getting on!

       -Bert

Hi Dr. Sahu, 

Pretty straight forward, actually.

You have a matchbox between the rf source and the powered electrode, and there is a blocking capacitor (the series capacitor) as part of the matchbox.  Thus, in the classic setup, there is no net current (no DC current) on the electrode and the electrode takes on a DC bias voltage so that, averaged over an rf cycle, there is zero net current.

What you need to add is a low pass filter between your DC power source and the electrode, on the plasma side of the blocking capacitor. This needs to handle the full rf-voltage you have on the electrode.  This would look like a high-power rf inductor with one end connected to the electrode, the other end to a vacuum capacitor connected to ground. The higher the value capacitor the better -- you could use high-voltage ceramic-chip-capacitors here too (see ATC 3600V range...)   For the inductor I would suggest a 20-turn, 80-100mm diameter coil, with 2.5mm diameter wire and 5mm turn spacing.   This is like the series inductors in 2MHz matchbox. You should be able to attach your DC power source onto the inductor/capacitor junction directly.   Note that if your DC voltage is less-negative than your rf-dc-self-bias would have been, then you could get *lots* of current. Start by assembling the rf components (without the dc power supply connected) and check to see what rf voltage you still have on the capacitor-inductor junction.  If this is too high (too high depends on your dc power supply) you might need a second low-pass filter.

I expect to be in SKKU in May/June, I'll check in and see how you are getting on!

       -Bert

Dear Prof. Ellingboe,

Thank you for your kind advice. After series of iteration we have solved this problem yesterday using the lowpass filter circuit.

Indeed, self-bias voltage matters a lot. Surely, we will meet whenever you will be in SKKU.

@ Bibhuti B Sahu

Dear Bibhuti B Sahu, may I ask you for a help?

I am not a person with electrical engineering background. But I am dealing with plasma etching recently. Would please mind helping to tell me why do we need a grounded electrode and what is the purpose of blocking capacitor in the powered electrode?

Thank you very much.