design on coaxial line - RF Cafe Forums

 Post subject: design on coaxial line

Posted: Fri Aug 03, 2007 8:43 am 

Dear all,

I am trying to design a coaxial line that fits the geometry of the power coupling port of a cavity. RF power to be fed is at 300MHz to 330MHz, pulsed at 2ms at 50pulses per sec(pps) of peak RF power 1MW. Coaxial line in air, cavity is in vacuum.

First thing i thought was, to find out what industrial coaxial lines are rated:
(http://www2.rfsworld.com/RFS_Edition4/p ... _55-69.pdf)
From this catalogue, i suppose none of them is possible for my application... because:
if you look at the tables on the right of each coaxial line specs, at 300MHz to 400MHz, the max. average power that can be withstood is 300kW, i.e. peak of 600kW even using the 9" coax line.

OR, am i looking at the wrong specs please??? what about on the table on the left where is mentions the peak power rating is 1.560MW for the 4-1/8" coax.....

what should i look at please? i am confused.....

i look forward to hear from anyone regarding this matter... Thank you for your kind attention.

Regards,
tmntRF

 Post subject:

Posted: Mon Aug 06, 2007 11:02 am 

Hi
Still playing with that high power then?
Good to know you are still alive
Data I have unearthed from the technical vaults chez moi suggests a 4.125 inch line can transmit 25kW mean power at 350MHz at an ambient of 40C.
Looking at the de-rating curve, this means 37 kW at 20C.
Assuming the termination is perfectly matched there will be about 45 amps of circulating current in the coaxial line. This means any joint with a resistance of 0.1 ohm will dissipate 200 watts ie it will get hot.
The average power you wish to transmit into the cavity is simply the peak power multiplied by the duty ratio, in your case 1MW * 0.1 which is 100kW. This exceeds the 4 inch line rating by a lot.
A 6 inch coaxial line would just about cope with this. If the 4 inch line were to be pressurized it would improve the power rating, but I would instrument every joint to watch the temperature rise.
The other unknow is what VSWR you might actually achieve. A worse than 1:1 VSWR will seriously reduce the power rating but I have no graphs for this situation.

 Post subject: regarding High power coaxial line loop coupling

Posted: Fri Aug 10, 2007 6:35 am 

Hi nubbage, thanks so much for your reply and input. I guess, i most likely just have to divide the 100kW peak power into a few more coaxial lines to feed the RF power into the cavity.

may i also further ask, designing a loop coupler to feed high RF power into a cavity,
1) does the plane where the loop terminates from the coaxial line be at the 'detuned-short' position from the cavity please?

2) does the length of the coaxial line before the loop terminates have any significant play from transmission line theory please?

i look forward to hear of any comments anyone has. Thank you.

regards,
tmntRF

 Post subject:

Posted: Fri Aug 10, 2007 10:34 am 

The loop should be positioned where the magnetic lines/field are maximum and the electric field a minimum in the cavity, and if this is a multi-mode cavity there are several choices.
If the coupling is efficient into the magnetic field of the cavity, then the match to the line is usually good enough in low power systems, and can be improved by a screw shunt capacitance. However in the case of your rig, it is anything but high power, so I do not advise any screw inserts to improve the match. As a first pass try to estimate the inductance of the loop from its length at the operating frequency, and see how this compares with 50 ohms. Then if you need to, try to estimate a correcting capacitive reactance of the same value but opposite sign, and position it a half wavelength away from the loop. However, most constructs I can envisage will deteriorate the power handling rating of the coaxial line, so this needs careful thought.
Regarding the power level in the cpax, I think you are right: separate coax feeds into the cavity magnetic field max points. Problem then is you need to worry about differential phase shift in the lines. Oh Boy.

 Post subject: temperature and pressure effects on coaxial line

Posted: Mon Aug 13, 2007 6:36 am 

Dear nubbage,

thank you again for your prompt and reply and help.

May i ask if, in general, can i explain that
1) the higher the temperature of the environment of the coaxial line, the lower the coax's power handling will be, because electric field breakdown occurs easily at high temperatures???

2) at lower pressure environment, coaxial line power handling is higher, because lower pressure means lower temperature, hence electric field breakdown is more difficult to occur???

I look forward to hear of your comments. Thank you.

regards,
tmntRF

 Post subject:

Posted: Tue Aug 14, 2007 10:59 am 

Hi tmntRF
Negative: with ambient temperature the issue is not the voltage breakdown but the current heating effect in the inner (mainly) and outer conductors. It is the I^2*R dissipated power and the thermal capacity that leads to excessive temperature rise in the copper. The temperature rise is the I^2*R term divided by the product of mass and specific heat of copper. The resistivity of the surface layer (the layer where 99% of the current flows) has a positive temperature coefficient, thus a kind of run-away occurs, where dissipated power increases as temperature increases, and this increases the temperature some more. For a given current and physical geometery (diameter) there is an equilibrium threshold beyond which the temperature is unstable.
The break-down potential (voltage) of air is proportional to pressure and moisture content. For dry air it is just the pressure.

 Post subject:

Posted: Tue Aug 14, 2007 11:06 am 

Thanks nubbage. really helpful.

 Post subject:

Posted: Thu Aug 23, 2007 3:53 pm 

Hi,

beginning to get really interested Is this for any physical experiments? How do you get the power? IOTs, klystrons or? You must send us some pictures!

We have one station here transmitting 3x40kW analog TV channels and 5 x 3kW DVB-T into two 5" cables. Works nice but the antennas are very well matched!
I have been in contact with http://www.maxlab.lu.se/ and they have some cavities were they put a lot of power - resulting sometimes in flashovers...
I think you should go for only one input, several will be difficult to get together.

Be careful!
Jens