RF radiation out of closed steel container

[litup]

Suppose you have an RF generator feeding a vacuum system with argon ionized by that RF. With the tuner you see the VSWR is very good.

That is version A, no ground or poor ground.

Version B is the same thing, the container is stainless steel half inch thick to hold the vacuum.

But in this case there is a very good close ground, where the frequency is one of the medical frequencies, in this case 13.56 Mhz.

Power, 500 to 900 watts.

If condition B has a ground within 1/4 wavelength, about 4 odd meters to a literal ground rod 8 feet deep in damp soil. Would there be less RF radiation from the stainless steel case vs one where the case is poorly grounded?

Is that the point of grounding, to prevent RF from being radiated by a steel case or copper or whatever?

I know it is a safety issue also, if not grounded, high voltages can appear but what about radiation? Does grounding help lower RF being radiated by the case of the vacuum system?

 

[tech99]

Suppose you have an RF generator feeding a vacuum system with argon ionized by that RF. With the tuner you see the VSWR is very good.

That is version A, no ground or poor ground.

Version B is the same thing, the container is stainless steel half inch thick to hold the vacuum.

But in this case there is a very good close ground, where the frequency is one of the medical frequencies, in this case 13.56 Mhz.

Power, 500 to 900 watts.

If condition B has a ground within 1/4 wavelength, about 4 odd meters to a literal ground rod 8 feet deep in damp soil. Would there be less RF radiation from the stainless steel case vs one where the case is poorly grounded?

Is that the point of grounding, to prevent RF from being radiated by a steel case or copper or whatever?

I know it is a safety issue also, if not grounded, high voltages can appear but what about radiation? Does grounding help lower RF being radiated by the case of the vacuum system?

Assuming that the RF is fed by a coaxial line into the tank, with the shield connected to the tank at the point of entry, there should be no radiation and the Earth electrode will have no effect. If for some reason the shielding is imperfect, then the ground wire will be expected to radiated some energy.

 

[Baluncore]

The presence of an Earth or ground will not influence radiation from the container.
You can expect radiation to leak through optical windows and from the seal where the container is closed.

Details of the container closure and external windows / connections will make it easier to predict leakage.
The orientation of the internal RF excitation and the windows / seal may be important.

 

[litup]

Well that sucks:) Ok, I get it. I would have sworn grounding the puppy would send the skin radiation to ground where it belongs. Doesn't at least SOME of the RF head to ground, assuming it is a nice low impedance path to Earth ground? So you are saying the skin, the metal of the casing does not radiate, even if un-grounded? You would have Argon ionized by the RF internally so I would think the case would have to be one of the connections to the RF.

How can the case NOT transmit? I know about rf shielding for doors but not sure how I could to that with a big fat oring about 8 feet in circumference around the main door. You would have to use the braided RF shielding away from the O ring grooves, that would be tricky to keep the two separated. In microwave ovens it's not such a big deal since both sides of the door are at atmosphere but when one has to be at a decent high vacuum level it is another story.

We do have sight glasses but they are only about 50 mm diameter, 4 of them. I would think at 13 Mhz there wouldn't be much radiation through those windows since 13 Megs is something like our 20 meter ham band so the window at something like 0.0025 % of the bandwidth of that 20 meters would preclude much radiation from the windows.

There are other little ports that could contribute to RF radiation but they are much smaller yet, say 8 mm across and there is the Cryo pump but it is like a skin around a skin. There is the roughing manifold but it goes to a valve which should block radiation since it is also a metal seal. But they have O rings so there could be leakage round those but usually the two skins are touching and thus part of the outer case, electrically speaking.

I know there is some radiation escaping the RF generator/case/whatever else is involved because I have a portable short wave receiver (Tecsun 660) with a signal strength meter and you can see a strong signal anywhere within 20 feet of the machine, not so much outside that distance, but if you extend the antenna you can find that 13 megs for quite a lot further distance.

The next question, does it matter much? I saw one time, using a 1/2 watt walkie talkie at a job site, we were talking between me on the outside of the cleanroom and my buddy on the inside, observing effects of me tweeking and so forth. I saw some boards with LED's light up whenever I transmitted and was within a couple of feet of physical distance to the circuit board in question so I know for a fact RF can effect circuitry but I don't have a signal strength meter except for the radio which is only a relative number.

So I have not seen a definite correlation between other equipment in the area and RF from the sputtering tool so maybe I am just tilting at windmills.

 

[Baluncore]

I know about rf shielding for doors but not sure how I could to that with a big fat oring about 8 feet in circumference around the main door.

That explains how the RF is getting out. The door seal will be critical. Maybe you could design it with two mating surfaces, one cylindrical with the O-ring in a groove, the other a disk with the RF gasket crushed by the vacuum.

If you ground the outside of the container then it should radiate better since any RF current flowing in the ground lead will radiate better, like an antenna.

 

[litup]

That explains how the RF is getting out. The door seal will be critical. Maybe you could design it with two mating surfaces, one cylindrical with the O-ring in a groove, the other a disk with the RF gasket crushed by the vacuum.

If you ground the outside of the container then it should radiate better since any RF current flowing in the ground lead will radiate better, like an antenna.

I thought you would not radiate if the ground wires was 1/4 wavelength or less. Right now I imagine the 'ground' is a nice radiator since it is clearly more than 1/4 wavelength, probably more like 3 or more wavelengths. I keep pushing management to let me drill a hole in the floor (concrete about 6 inches thick) to allow me to pound a real copper ground rod into Earth which is about 4 feet below the concrete. In that case, suppose I keep the ground length say 3 or 4 meters, less than 1/4 wavelength, it should not radiate I thought, the ground wire, that is.

 

[tech99]

I thought you would not radiate if the ground wires was 1/4 wavelength or less. Right now I imagine the 'ground' is a nice radiator since it is clearly more than 1/4 wavelength, probably more like 3 or more wavelengths. I keep pushing management to let me drill a hole in the floor (concrete about 6 inches thick) to allow me to pound a real copper ground rod into Earth which is about 4 feet below the concrete. In that case, suppose I keep the ground length say 3 or 4 meters, less than 1/4 wavelength, it should not radiate I thought, the ground wire, that is.

I think we need to know more information. To suppress radiation from a 1kW transmitter to the point where it is undetectable is not realistic. So do you have a problem with radiation; is it causing interference, or are you concerned about regulatory compliance? If you need to suppress radiation further, it would be useful to have a diagram of the set up and maybe a photo of the type of transmitter/tuner etc. It can be quite a skilled task to resolve the sort of problem you describe.

 

[Baluncore]

Grounding will not help reduce radiation. A ground lead shorter than a quarter wavelength will not stop radiation, it will reduce it, but so long as current flows there will be a significant dipole radiation from the “Ground Lead”.

The major source of radiation will be from a slot that has a peripheral circulating current, probably the end door of the container. Radiation from an external RF generator, power cables or internal sensor cables is also possible. You may need some really good LP filters or ferrite toroids on your cables.

 

[litup]

Ah, ferrites. Good idea, it is only 13 megs so that should reduce at least that part of it. Better RF shields on the doors would help. So far I don't find interference with nearby electronics but just want to reduce the possibility of trouble.

 

[tech99]

Ah, ferrites. Good idea, it is only 13 megs so that should reduce at least that part of it. Better RF shields on the doors would help. So far I don't find interference with nearby electronics but just want to reduce the possibility of trouble.

I would guess that common mode currents on leads are the main source of radiation. In particular, put ferrites on the AC mains lead to the equipment, on any data and control leads and on coax cables, both ends. The hatch cover is much smaller than half a wavelength and should not radiate very much, but it would be an idea to make sure that at least it is connected to the chamber at some point.

 

[litup]

One problem is water cooling of the RF tubing. We use 5/16th inch silver covered tubing with water flow inside for cooling and we use a corrugated sleeve as the outer conductor going to the object needing the RF. This sleeve is about 2 inches in diameter with teflon spacing internally to keep the tubing in the center. When the VSWR gets to about 1.3:1 that corrugated sleeve starts getting hot. Like 60 to 70 degrees C which to me is saying there is a lot of outer sleeve ground current and that might be an offender in the RF emission. It would take a rather large ferrite to go around that stuff. You know of an ferrite that can go around a 2 to 3 inch sleeve?

 

[tech99]

One problem is water cooling of the RF tubing. We use 5/16th inch silver covered tubing with water flow inside for cooling and we use a corrugated sleeve as the outer conductor going to the object needing the RF. This sleeve is about 2 inches in diameter with teflon spacing internally to keep the tubing in the center. When the VSWR gets to about 1.3:1 that corrugated sleeve starts getting hot. Like 60 to 70 degrees C which to me is saying there is a lot of outer sleeve ground current and that might be an offender in the RF emission. It would take a rather large ferrite to go around that stuff. You know of an ferrite that can go around a 2 to 3 inch sleeve?

Why such a big, water cooled transmission line for just 1kW of power? I am sure quite small cables could handle the power easily.

 

[litup]

Of course a cable such as RG58 or better yet RG8 can handle 1Kw, the problem in our case is the switching to 4 assemblies, was done with a very expensive device called a 'Superswitch' which is operated by a matrix of solenoid valves, able to switch in a DC supply for some apps and an RF supply for others and either of those to one of the 4 loads. There is a newer technology that uses vacuum switches to get rid of the dreaded Superswitch but either switching technique has to have cooling, proven to die if getting too hot. So that requires silver coated tubing, 1/4 inch and 5/16th inch tubing with water flowing in them. That in turn means there is a lower resistance limit to the cooling water, quite a bit better than tap water so we use distilled water. We could use DI water but water molecules are polar and it leads corrosion of metal parts coming in contact with the DI water and that eliminates DI water from being used inside metal tubing. So it is a matter of heat control.

 

[Tom_K]

One problem is water cooling of the RF tubing. We use 5/16th inch silver covered tubing with water flow inside for cooling and we use a corrugated sleeve as the outer conductor going to the object needing the RF. This sleeve is about 2 inches in diameter with teflon spacing internally to keep the tubing in the center. When the VSWR gets to about 1.3:1 that corrugated sleeve starts getting hot. Like 60 to 70 degrees C which to me is saying there is a lot of outer sleeve ground current and that might be an offender in the RF emission. It would take a rather large ferrite to go around that stuff. You know of an ferrite that can go around a 2 to 3 inch sleeve?

Something doesn’t sound right. If your VSWR is 1.3, there should only be about 17 Watts reflected back for 1 KW forward power. Sounds to me like you have a lot more than that coming back if the transmission line is getting that hot especially with water cooling. If the mismatch is that bad, it may be causing generation of harmonics that can be the source of the RF radiation problem

At what point in the system is the VSWR measurement taken? If it is at the load, after the matching box, then it won’t be indicating what is reflected back from the interface between line and match box, upstream of the VSWR measurement point.

Here is a very good Pdf that may have a few ideas that will be of some use to you.