I can't give a direct answer except to point out that inflatable radomes are around.  I have seen a number, a couple quite large, like 70ft dia.

At least it can be treated as a thin wall design which simplifies things but I would try the radome people like Norton.

Thanks Brian, but the inflatable is probably not an option, since weight and portability are the main concerns with this system.  The entire system is packed into two cases, and hand-carried or dropped from a helicopter to the operations area.
See this site:
http://www.syrres.com/stc/products_lcmr.htm

How can a blow-up radome not fit your criteria of weight/portability?  It takes minimal space when deflated, and weight is maybe a few pounds...

Dan - Owner
http://www.Hi-TecDesigns.com

For one thing, on battery operation, the power to run the air pump along with the radar electronics is limited, unless you pack larger, and you guessed it, heavier batteries.  The radome structure would need to be sealed to minimize leakage, which can be a challange on sand, requiring more stuff to be packed into the hand portable cases.  Did you look at the picture of the system?  It is pretty simple, and it is hard to get lighter than the nylon shroud presently employed.  We are looking into a cotton canvas radome material, which is kind of in the middle of the triboelectric series and not nearly so apt to accumulate surface charge, but mechanically it has limitations.

Have you considered embedding a loose wire grid in the nylon?  

Or the same with something like Tyvek?

TTFN

Make the top and bottom plates and the baggie, I mean radome, a bit bigger so that the radome doesn't get so close to the antenna array. If necessary, add bag stiffeners made from RF-tranparent plastic.

I'd also investigate why the DC-grounded antennas aren't protecting the LNA inputs from the static. I would have thought that they would...

VE,
We thought the same thing.  The inputs are pretty well protected by grounded stub input filters and limiter diodes, but the outputs of the LNAs are vulnerable to the overvoltage of the discharge.  These antenna columns have approximately 5' of cabling between them and the primary power supplies, so there is some impedance between the local ground and system ground, thus allowing ground plane charging.  I have had success putting 1N5907s across the bias supply filter caps for the LNAs, but would like to find a non-electronic solution until we can re-spin the board.

IR,
I do not understand the loose wire grid in the cover material.  Would that still be radiolucent?

The length of the cabling shouldn't matter. You only need to address the peak voltage across the LNA device in question (in other words, local ground). Putting it another way, the protection scheme across the devices should be designed to work irrespective of external grounding.

If the static discharge is getting into the LNAs through the power supply (as you've indicated), then fixing it should be fairly easy.

If you're not perfectly clear on the exact failure mode, then it might be well worth doing a detailed analysis to see how the 'sparks' are getting in.

If it is getting into the device via the power supply, then you can fix that (easily), then you don't have to worry about the radome any more...

Depends on the spacing and resistance, so something with 2-3 inch gaps should be reasonably transmissive.  And the wires, or carbon fibers, can be quite thin, like 40+ AWG.

You could try buying an ESD labcoat with embedded wiring just to test it out:
http://www.esdproducts.biz/ESD_Grounding/Garment_Table/Lab_Coat/lab_coat.html
http://www.vidaro.com/StaticControlESD/index.html
http://www.sullivanuniforms.com/Merchant2/merchant.mvc?store_code=SU&screen=PROD&product_code=KK18

TTFN

VE,
The static is definitely coming from the radome.  The reason it causes the ground-to-power bus voltage to exceed the absolute maximum for the LNA device is that the impedance of the twisted pair back to the power supply system ground has some finite inductance which resists the dissipation of the many coulombs of charge injected onto the antenna elements (local ground plane) for several hundred nanoseconds while it flows through the output pin lead wire of the LNAs.  I can short the power supply connector and inject charge onto the ground plane and the charge dissipation time is reduced to several nanoseconds.  We first encountered this phenomenon out at Fort Huachuca Az during MIL 810F sand tests,which are performed in 40 MPH winds.  The local humidity was a whopping 7% that day, and if it had been dark, I'll bet we could have seen the little lightening bolts going off inside the radome.

Voltage protection should be local and directly (as close as possible) across the device being protected. Once this is done, then the instantaneous common mode voltage relative to some distant ground should be a non-issue.

But, you might also want to think about the 'RF Noise' issue of all these static discharges. Will your system still work with all this static occuring right at the input?

It seems like you've got the worst possible radome material when it comes to static. How about tossing it into the dryer with a sheet of 'Bounce' or similar consumer static control product? It seems like anything would help.

Can you go with a hybrid inflatable system?  Maybe use a standard material, but the supports themselves are inflatable.  That would require a very small air pump that eats a minimal amount of energy compared to blowing an entire structure up.  Or maybe use an air canister that is refilled when the unit is at base, or a larger pump to refill the canister only when the unit is at home base and power concerns are at a minimum.

Nylon just seems like the worst way to go in a dry environment with a lot of ionising-particulate circulation.

Dan - Owner
http://www.Hi-TecDesigns.com

That's an excellent idea Dan.  We were kicking around a polystyrene leaf spring structure which would mount between each of the 24 columns to hold the radome away from the antenna elements, but an inflatable balloon might be even more effective, if we can figure out how to keep it from getting punctured by the surface mount stuff on the boards.  Another idea we are looking at is to tie bungee cords axially around the structure between each of the columns to pull the radome into intimate contact with the antenna elements, as we have found out the hard way that the antenna columns facing into the prevailing winds do not fail because the radome tends to stay discharged by the ground plane of the column boards.  It is the columns in the +/- 90 degree quadrants which tend to flap in the breeze that build up enough charge to do damage.

Another option might be to use a few semi-rigid horizontal rings stacked vertically every couple of inches, like a plastic dryer vent hose.  The rings themselves should present a minimal radar return issue, especially if you choose the correct material, but provide enough stiffness to the radome to keep the correct overall shape.  A couple of stiff vertical sticks to tie each ring to should also prevent the middle rings from pushing off center and towards the antenna elements during heavy winds.

Dan - Owner
http://www.Hi-TecDesigns.com

Beech aircraft adds/paints antistatic coating on their radomes. The system we used was 2-18 ghz and it had minimal effect on performance. This coating was on the outside of the radome on their aircraft. It's a simple paint coating.

This is the prime reason why I suggest always using grounded center conductor antennas. You could switch to a folded dipole antenna which has typically wider bandwidth and the two arms are connected, i.e. center conductor to ground.

also, the famous quarter wave short = open circuit.
If your antenna is a printed circuit, you could add a quarter wave short on the distribution systems to short out the center conductor.

kch

Hig,
As I mentioned in the original post, the elements are an integral part of the ground plane.  The problem is that the small impedance between the local antenna system ground and the main system ground causes ground charging when massive charge dump onto the plane occurs.  We have actually observed this phenomon an oscilloscope attached between the power and ground pins during the sand tests.

"The problem is..."

...lack of voltage protection ACROSS the victim device (locally). Distant grounds, and the associated inductance to get there, are not 'the problem' - they're just a real world given. It's a red herring to even mention it.

Once the radome is sorted, then you'll probably find that human static discharge during set-up and take-down handling will be the next problem... You might as well respin the PCB to add the protection diodes. And just be glad that the susceptibility is - apparently - just on the power supply side and not in the RF side where adding protection might be much more difficult.

 

1) that has decent surface charge mobility in an insulating flexible dielectric? Sounds contradictory.

If your antenna is not circularly polarized, and is linear polarization, then following IRstuff's suggestion  you can metallize your radome in the opposite linear polarization. It's possible to metallize about 75% of the radome surface and still have great rf transmission properties in your primary linear polarization. Normally you'd think to add thin wires as suggested by IRstuff, oriented cross polarized covering 1-5% of the radome surface, but a recent analysis by a co-worker showed that having much more metal, or just the right thickness metal to non metal ratio can provide minimal thru loss over a pretty wide bandwidth. I was very surprised at this result, but I believe it.
Is your antenna linear polarization? If so, you could try a metal tape experiment, or wire experiment with very little effort.

Actually, what is your plan? Is this a very high priority item with funding requiring detailed analysis and a 9 month project with modifications of 462 pieces of hardware? Just trying out my psychic thoughts.

kch
PS: Has anyone noticed that the word your is often typed you, must be a human factors natural oops, I do it alot.

If the metal tape idea works out, you might be able to combine it with my dryer vent tube radome format to great success.

P.S. I often type "woudl" instead of "would"... no matter how carefully I try to type, I always default back to 'dl'.  Very frustrating, especially when other words like "told", "sold", and "should" seem to come out OK the first time around. <shrug>

Dan - Owner
http://www.Hi-TecDesigns.com

Hig,
This system is vertical pol in the orientation of the array cylinder.  If you follow the link in the 3rd post, you will see a picture of the system.  If you go further and follow the Army link in that piece, you will see a picture of the system without the radome, and if you look carefully, you can see the outline of the individual dipole elements through the PC board material.  That is an interesting experiment with the cross-poled wire.  We may try that ourselves.
Our next test is scheduled in early August, and we are going to try to "bungee" the radome axially between the individual columns, and if old Dan Bernoulli is kind, hopefully we can maintain intimate contact between the sides of the radome and the antenna elements, keeping the charging to a minimum.  Just in case, however, I am taking along an extra set of columns with 1N5907's back-soldered across all the LNA bias filter capacitors.

Static can usually be dissipated with surface resistivity values of as high as single-digit Mohms per square. The values that would affect RF are probably (a guess) orders of magnitude less. I think that the workable range is a wide target.

lighterup,
I saw the photo's. The wires should work unless dissipated static is a localized point source and not globally connected, which is probably is since current doesn't move on an insulator. Hence, I'd guess that individual sections of your radome make static that's isolated from the other areas of your radome. Maybe you'll need this anti-static paint http://www.all-spec.com/1/viewitem/4000-1/ALLSPEC/prodinfo/w3path=cat (try it on a piece of dielectric and measure S21 change). Seems simple to spray or brush it on and be done with it. Maybe call Beech aircraft in Wichita KS and ask what they use.

 Slight concern using wires that you keep them crosspole if your radome material droops or can change orientation. I'm curious about the rain and it's affects with a wet radome. Normally, you'd want the rain to ball up and roll down the surface, seems like it would wick down and cause problems. A smooth surface can have hydrophobic coatings added to minimize loss due to sheeting. At 2-3 ghz, expect 5-10 dB loss due to a sheet of water, plus pointing error accuracy degradation.

Raytheon Raymarine makes a marine radome 19" diam. x 19" high, although your unit looks larger than that. There are other larger sized and low cost lightweight radomes from Andrew Corp available.

kchiggins

Hig,
I think I will order a sample of the staticide and try it.  The problem I see with it is if it stiffens the material when it dries.  This system has to be taken down and moved frequently, so the radome gets folded up into a small volume and crunched into a carrying case.  If the paint tended to crack, I can forsee problems.  Perhaps a thin coat applied in the field in arid locations would work.

Worth a try,
I haven't used the material shown, just found it on the web. I'd do an S21 from antenna to antenna with a sample to make sure it's not too lossy.

I realize from your statement why the array looks so odd, cuz it folds small. Interesting mechanical arrangement.

Maybe a secondary anti-static bag or cloth painted with anti-static coating fitted over the array first may be friendlier than spraying your radome material? At least you won't have to tell your co-workers that you ruined a radome with the anti-static paint until you know it works.

Maybe if you add a plastic disk to the top surface that hangs over 3 inches or so, it might keep the radome from touching the antenna surface and keep the rain off better. The bird poop would miss the radome too if the top disk is large enough. Maybe funnel water collected down thru pipes between your elements. Lots of useless ideas to consider I know.

Let us know how your experiments come out.
I hope your system can be attached to an audio warning system for inbound projectiles to give people a few seconds to take cover. Add some pits or sandbags for more protection. Good luck.

kchiggins
www.toyon.com

Thanks for all the food for thought, guys.  My mechanical guy tells me that we found a way to cinch the cover really snug without crushing the columns by using a corset cinch with nylon strings and plastic eyes to take up excess material in the radial plane, kind of like a shoe lace concept.  This seems to keep the material right up against the antennas in a 30 MPH air stream, which is all we can generate here.  MIL-810F calls for 40 MPH, but we will probably go with the radial cinch for our next test attempt.  Also, it keeps the surface charge below 500V as seen through our little ACL static locator meter, compared to 1500V with the radome just flapping in the breeze.  I think that's going to be more acceptable to the users than lugging around all those bungee cords.

Also think about washing the nylon in an anti-static liquid... doesn't Woolite do something like that?  Or maybe that was the Bounce sheets I used ;)

Dan - Owner
http://www.Hi-TecDesigns.com

Yes, Mac.  But like your household washing, after a couple of weeks out in the warm, dry breezes of beautiful Iraq or Afganistan the effects of the antistatic coating wears off and you are left with the same old problem...and no laundromats within 1000 kilometers.

Add it to the ATP (or whatever it's called for this application).

"Radome must be washed on spin cycle using Woolite at least once a week during normal use."

I've seen army use procedure manuals that actually listed things like "Do not handle board with taco sauce in vicinity." (Gotta love our government).  A board was fried because some moron tech handled the boar while eating a taco... sauce spilled on it, shorted it out, magically creating a new caution line in the manual.  When the board came back, the taco sauce had dried in place, the tech didn't even worry about cleaning up the evidence.

Dan - Owner
http://www.Hi-TecDesigns.com

I am sure that troops in the field will be able to wash their radome.  

Another option is a antistatic spray:
http://www.mystaticguard.com/products.cfm

which, apparently, you can also make yourself:
http://www.recipegoldmine.com/house/house17.html

TTFN

lighterup,
just saw this in the Plastics solutions newsletter and thought of vous. Maybe they have a thin sheet you could use to dissipate your static.

Mike Oliveto,
Quadrant Engineering Plastics Products
Statically dissipative plastics reduce losses during the handling of electro-sensitive devices and improve safety and reliability of packaging equipment.

Kevin

Thanks Hig.  That material is a machineable stock similar to delrin, except it is semi conductive.  I think I may try it in an ESD sensitive gluing process for one of the press fixtures used there where we currently use a delrin platten.

In the search for antistatic material don't lose sight of the overriding requirements of mechanical properties, RF loss tangent and water absorbtion.  The material must not be hygroscopic and must have a low tan-delta.

The fabric listed in this url might be what you want:

http://www.fabricatedsystems.saint-gobain.com/Data/Element/Node/Category/category_edit.asp?ele_ch_id=C0000000000000001282

Good point Brian, fix one problem, create another. Accidental job security?
kch

Thanks Brian.  I have contacted them to see if they know where their fabric is in the triboelectric series, shown in this URL:
http://www.school-for-champions.com/science/static_materials.htm
We are also looking for a combination polyester-nylon fabric, which should have pretty good anti-static properties, since these materials have complementary work functions.  So far, we are trying to buy COTS without getting involved in a big R&D effort with a supplier, but have not had any great success as yet.

Thanks to everyone who has contributed to this thread.  I think Brian has come upon the answer, although it is not within the specific microwave materials that Saint-Gobain markets.  They also make a Kevlar composite material used for chem-bio and vapor protection gear which is rated very good for resistance to surface static charge, a very important thing if you are wandering around in a building full of volatile chemical vapors.  I have ordered a sample of it to test for transmission loss in our chamber, and if it works out, I think this may be the answer to my original question.  
Thanks, Brian.