Quartz torch firing, first RF experience

For posts specifically relating to fusor design, construction, and operation.
Post Reply
steve_rb
Posts: 145
Joined: Sun Oct 18, 2009 4:18 am
Real name: Steve Robinson
Location: Tehran University

Quartz torch firing, first RF experience

Post by steve_rb »

As my first experience after a few times burning my finer finally I could turn a quartz torch on using about 1.5 Kw 13.56 Rf. Attached is a photo of the fired torch . But flame didn't came out as it should be. Instead of a nice heart shape flame I see a cylindrical shape flame with turning lines. I used three gas as follows:

1- center gas: argon (1 L/min)
2- midle gas : argon (2 L/min)
3- Outer cooling gas: argon (10 L/min)
4- all three gases were running during firing

I can see only outer gas is ionized and turned on . I was expecting outer gas only cools the quartz glass by separating flame from the glass. I know something is wrong but couldn't find what. I tried changing coil position and playing with Rf power and gas fllows but not much change in the flame shape. I want flame to be stretched out of the quartz.

I also used a spark type arc igniter by letting two cupper wire enter inside the quartz torch a little bit. Without igniter plasma does not starts. After starting igniter for a short time torch fires and then I release igniter power botton and plasma stays stable. Arc lenght is about 1-2 cm and I can see gas flow is stronge enough that can bend the arc a bit before applying RF. I had no previous experience with this and it took me only two days for setting things up and firing the torch. How should I get the flame shape to a proper shape (a nice heart like shape) ?
Attachments
quartz torch.JPG
steve_rb
Posts: 145
Joined: Sun Oct 18, 2009 4:18 am
Real name: Steve Robinson
Location: Tehran University

Re: Quartz torch firing, first RF experience

Post by steve_rb »

Giving some though I guss cooling gas should be off at first when RF is turned on. After plasma ignition start cooling gas should be turned on gradually before quartz gets hot.

Also seems spark igniter should have a short arc and should be right in the midle for the midle gas to be ionized.

Run out of gas and will test this when fresh cylinder came in.
morgoth31
Posts: 31
Joined: Tue Jun 22, 2010 9:01 pm
Real name:

Re: Quartz torch firing, first RF experience

Post by morgoth31 »

ok i worked for seren ips that makes open air rf torches.
the tubes they use pass the gas around the outside as cooling then thru a tube in the center this cools the glass while preheating the gas at the end.

if properly tuned the sparker is not needed only sufficent power to ignite the gas this is very dependent on the tuning of the output. the capacitance at the output is normally adjusted using what they call a matching network which consists of 2 variable caps one air var cap for rough adjust and one vac cap for fine tune. doing so by hand is not easy since its the reflected power you need to know to make this adjustment. i have a friend at RFVI (called rf 7 ) that is an engineer of power supplies. this company sells used equipment at parts such as the torch coils and other pieces you should look into them they might have a relatively cheap torch you can buy to make this easier.

the shape of the torch and the flow of gas also affect the plasma torch in how it burns and how stable it is. also frequency of the torch has affect also most plasma torches are 13.5Mhz to 24Mhz 13.5 being more average these are nomally used in mass specs and gas chromatagraphs.

hope this helps and feel free to let me know if you need to know more as i can talk to my friend to get more precise details since it has been a few years since i did this actively.
steve_rb
Posts: 145
Joined: Sun Oct 18, 2009 4:18 am
Real name: Steve Robinson
Location: Tehran University

Re: Quartz torch firing, first RF experience

Post by steve_rb »

I looked at RF 7 site and realised they have a few 13.56 MHZ refurbished RF units and they also take order for matchings too. If it is possible please send me his email address. I would like to talk to him and see if I can purchase a RF from them.

Also I need to know how to prevent cooling gas from turning on. At the moment instead of the main gas cooling gas turns on. I tried different tricks but seems need more experniece from a hands on person. I have too check all parameters with him. Since it is non relevant to the forum I can't continue discussion here.
morgoth31
Posts: 31
Joined: Tue Jun 22, 2010 9:01 pm
Real name:

Re: Quartz torch firing, first RF experience

Post by morgoth31 »

the web addy is www.rfvii.com

they sell just about anything and have thier own line now too

you prolly have to call them to get quotes hes and engineer not a sales guy.
steve_rb
Posts: 145
Joined: Sun Oct 18, 2009 4:18 am
Real name: Steve Robinson
Location: Tehran University

Re: Quartz torch firing, first RF experience

Post by steve_rb »

morgothian

One thing I realised I have about 2000 V peak to peak which with about 2KV RF power I used shows current is about 1 Amps. This creates a very weak magnetic filed. I don't know why despited zero reflection I have still very low current in the coil. I was expecting 10 Amps with about 200V peak to peak. Low current is the reason for poor plasma creation. Low current may caused by the poor connections I have. I have connected the ground side of the coil to chassis with similar connection method can be seen on the above picture which shows how the othr end of the coil is connected to the RF output. I have to prepare some special fittings to fix the connections and see if current goes up?
morgoth31
Posts: 31
Joined: Tue Jun 22, 2010 9:01 pm
Real name:

Re: Quartz torch firing, first RF experience

Post by morgoth31 »

if you had a bidirectional coupler you could activly monitor your reflected if your getting an output of 1 amp that might mean you have 9a reflected. so your only getting 10% of total power to your plasma this is normally a matching issue. remember the output is an LC circuit and its resonate freq needs to match the operation freq. this is complicated by the dynamic effect of the gas and plasma on the inductance of the plasma coil hence the matching network.

remember amps are directional you can use 2 with thier directions opposed and get a rough estimate of the reflected. since they cant be nulled activly you can use a 50ohm load to figure out how close they are to nulled and use that ratio to determine it. you can get 2kw oil cooled loads these help with figureing out your null.

but just to keep it simple set your current loop to read amps going toward your PS and you will know how bad the reflected is this can also help you tune the matching network.
steve_rb
Posts: 145
Joined: Sun Oct 18, 2009 4:18 am
Real name: Steve Robinson
Location: Tehran University

Re: Quartz torch firing, first RF experience

Post by steve_rb »

morgothian

For the moment I have almost zero refleted power and I am not worry about it. What I have found lately is the importance of the ignitor. They way I have applied my ignitor is wrong (two copper wire inserted inside the torch from front having a short arc) . For torch to turn on the way it should be I have to have a long arc (about 8 cm long) axially for a short time to generate a whole lot of energetic electrons along the axis of the torch. I don't have any means to creat such an arc but I am thinking about using a tesla coil for this purpose. Do you have any experience and info about quartz torch ignitors?
morgoth31
Posts: 31
Joined: Tue Jun 22, 2010 9:01 pm
Real name:

Re: Quartz torch firing, first RF experience

Post by morgoth31 »

sorry all units i worked on were self igniting.
with proper flow and match (coil inductance to output capacitance - coil being mildly variable given the flow and and such varying the inductance vs air)

most auto matches have a set for light point and seek from there using a bidirectional coupler to find lowest point of reflected power as the perfect match. (most newer rf supplies have a voltage match output on thier dsub meant to be connected to the matching network this normally comes from thier Bicoupler.) this light point is mostly because the best light point is not always the best match point for reflected. (VISWR voltage impedance standing wave ratio)
wiki has a nice write up (http://en.wikipedia.org/wiki/Standing_wave_ratio) vswr is for transmit type and viswr is more for continuous wave types.

i will ask my friend if he has any insight on this given he designs rf power supplies for this type of application.

im sure ill get some more info from him on this.
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”