200Kg door on anechoic chamber.

[waveboy]

Dear Engineers,

I would be very grateful if anyone had an answer for why the door on the anechoic chamber i just assembled weighed 200Kg. (This door cost thousands of pounds).
The chamber is for testing antennas at 1.8 to 2.2GHz. The antennas are not particularly high power ones. (-Just like a mobile phone antenna). The chamber is in an area which is a "quiet" RF area (i.e. there are no big RF transmitters in the vicinity of the chamber).
The chamber was about 5 metres cubed and walls and ceiling were covered with RF absorber spikes (these "spikes" were about 50cm long).
The whole structure was completely lined in metal as all the wooden panels that constituted floor, walls and ceiling were covered in metal sheet. With regard to this, can anybody explain why the floor had to be metal lined?

 

[VEBill]

To make accurate RF measurements, you need to eliminate stray RF from local sources (like TV and FM stations). Thus the chamber needs to be a Faraday Cage to keep out external RF.

The door is probably heavy because of the brass (?) frame. The main panel in the door could be plywood lined with sheet metal. The frame is made that way so that it will last a long time.

The floor is metal because the shield needs to be complete. The room designers don't know where it is going to be installed, so they need to provide a shielded floor too.

Bring in an FM radio and close the door. Any signals?

 

[Comcokid]

You have a large chamber.

1) The door has to hold good tolerance over time and use. Otherwise, the fingers which line the door edge may not make adequate contact.
2) Additionally, each finger may have a small spring force, but multiply that times hundreds of fingers and try to hold door tolerance without much deflection.
3) The rest of the chamber is mechanically static. It doesn't move, it just sits there. The door gets used all the time. It is the one thing that could determine the useful lifetime of the chamber as a whole. Other than the foam, the door is the one thing subject to daily abuse.

These issues result in the need a door with a really strong frame.

You cannot have a linear gap at the seams ofthe door in your chamber - otherwise you create a slot antenna.

 

[waveboy]

Thanks for your replies..and for Comcokid for supplying the answer that i suspected the heavy door with the "fingers" was designed to address....

"You cannot have a linear gap at the seams of the door in your chamber - otherwise you create a slot antenna."

However, having tried to design decent slot antennas with little success in the past, i am doubtful that one would "accidentally" make a decent slot antenna with a "normal", lighter door that had a slim slot in the gap!! (even if the door "deformed/warped" and produced a half decent slot....-remember that it isnt as if someone's firing an RF signal toward the "slot" at the door frame edges anyway. -so i still think that the 200kg door is going to extremes.....In any case, whilst i was inside the chamber with the door wide OPEN, my mobile phone FAILED to find any signal (it did find a signal just outside the chamber). I also think that the "fingers" in the "socket" which receives the "ridge" round the door when it closes are unecessary and provide an insignificant effect.

I have worked in antenna design companies where perfectly satisfactory antenna patterns have been obtained outside in the garden, with no chamber at all!! -these antennas were of all types -cavity backed planar spirals from 0.5 to 18GHz, Dipole arrays for 6 to 18GHz and many many others. -some of our antennas (dishes) had such long near fields that you wouldnt have got the test set-up inside a chamber of practically buildable size. -All of these antennas have been sold to customers who have no complaints, -not to mention that the customers came to witness our field trials and had no problem with the lack of a chamber.

I wondered if (hopefully) any engineers can refute my claims?

The major problem with outdoor antenna pattern testing (if any), is that due to multipath -and 200kg doors will do nothing to alleviate this problem.

Regarding chambers, they are generally poorly lit and obviously have no windows and scant ventilation...they are a poor, unpleasant working environment, and generally, antenna testers will very possibly be angling to get their test time down to a minimum whilst cooped up in a claustrophobic chamber....testers may well devote unsufficient test and re-test time to antenna pattern testing in a chamber. -and if you ask me, since Electromagnetic simulation progammes never provide accurate "real-world" results, i say that long test and re-test (and re-test again!)time is essential for antenna design.

 

[IRstuff]

"accidentally" make a decent slot antenna

That's not the issue. You don't need a "decent" antenna to screw up critical measurements

TTFN

 

[Comcokid]

I'be been to testing companies and used their chambers. At at least two different places when doing sweeps, low level spurs would show up in the 88 to 108 MHz band. The tech would say "ignore it, it's just the local FM station". Good chambers are no guarentee that you still have perfect shielding. Unwanted signals can leak-in/out through numerous paths. In fact, at one company with a 10 meter chamber, they had a small chamber attached to the big one that was just for the analyzers and the test personnel. That way they could keep external signals from leaking directly into the equipment itself.

Never-the-less, you take all precautions you can. I'm sure that the chamber company doesn't necessarly have different types of doors depending upon if you are testing a radar, doing suceptability sweeps at 100V/m or just testing for low level emissions. Hence, one massive door for all applications - including your antenna pattern testing.

Presently, the cost of copper is skyhigh, hence the high cost of brass.

 

[waveboy]

When you mention "spurs" in your antenna patterns -i too have had great problems with these -I believe they are due to perhaps problems with the detector, or due to the RF source suddenly varying in output, -or even some mechanism where a reflected ray comes back in phase after being re-radiated from the antenna under test and then coming back from the transmit antenna. -In our case, these spurs were not due to any RF transmitter, (since we were so remotely located that there were no other transmitters) -not only that, but we were working above 3 (even 18) GHz and there was definetely nothing transmitting round there.

These spurs have haunted me ever since. They mostly occur (but not always) near boresite and seem to be more prevalent with low gain antennas (whether the low gain is due to low directivity or poor efficiency or both).

Strangely, the "spurs" (i called them spikes) in the patterns were intermittent -you'd take a re-scan and they'd be gone.

So i doubt that your chamber is at fault for these spurs.
-Perhaps another antenna pointing at the door could confirm that there was nothing coming through during the scan. 100MHz has a wavelength of 3 metres and i can't see it getting significantly into a chamber via any small gap.

 

[Comcokid]

In my case, the "spur" was external. When checking emissions, you use a spectrum analyzer. You could center on them, narrow your sweep and bandwidth, and see the FM modulated signal typical of a radio station. In the situation, the test facility and the local FM radio station coexisted for years, but then the station got new owners and a new license and went up 50x in power - it was only a mile away. I happened to be running tests in the oldest and somewhat battered anechoic chamber they had.

In your case where you are checking antennas, you are probably using a level or normalized swept signal generator and detector. Since detectors are broadband, and unless you have a band-selective filter (or waveguide) somewhere a strong signal outside even outside of the band you are checking could be rectfied by the detector and cause a spike. Depends upon details of your test setup, and I'm no expert on antenna sweeping.

 

[waveboy]

Hi, Regarding the "spurs" in my antenna patterns, they occurred when using an RF source transmitting at particular selected test frequencies...the antenna_under_test was receiving, and fed into a spectrum analyser, limited to the frequency of interest, so it is highly improbable that the spur was caused by an out of band signal. -And at 3 to 18GHz, the spur really could not have been caused by an in-band signal as no-one transmits that high GHz -especially in out remote location.

 

[VEBill]

For waveboy's 'spurs':

If these 'spurs' are increases in amplitude (and you mentioned the boresight), then it probably isn't the generator. I assume you're using something close to maximum output, thus a +10dB increase (for example) is unlikely.

Likely just a mundane equipment glitch. Try slowing down the frequency sweep speed a bit, or using smaller frequency steps, or both. If you suspect the detector, try knocking 3dB off the output just for laughs.