RF test chambers and ranges | Hacker News Reader

RF test chambers and ranges

planetanalog.com

49 pointsdev_tty0113 comments

My first job was finite element analysis for mirrors at an antenna test site. The larger mirror was about 10 m by 12 m, if I remember correctly, the other one not much smaller.

The mirrors where placed in huge RF test chamber. Being inside of it was mighty impressive. The cones must have been made of a conductive material, but I think they were also some kind of foam, which made the whole thing act like an anechoic chamber as well. Very eery experience and the complete opposite of what I experienced in the other building at the same site I had the pleasure to visit.

The environmental lab (with environmental meaning the harsh conditions space vehicles encounter). It was were the shaker equipment was operated. Basically gigantic magnetic speakers, large enough to bolt satellite parts on and subject them to the vibrations they were to encounter during the rocket launch. Solid metal parts wobbling like pudding at a noise level that'd make AC/DC blush. Not eery, but the other kind of frightening.

EDIT:

Found an image of the chamber where I worked. The reflectors we were designing were for a different, a bit bigger but otherwise similar site I unfortunately never visited.

https://www.alamy.de/aggregator-api/download/?url=https%3A%2...

> Solid metal parts wobbling like pudding at a noise level that'd make AC/DC blush. Not eery, but the other kind of frightening.

Reminds me of something my dad snarked on decades after happened. He was on a team doing FEA to verifying satellite structures. They told one team that a cross member on their design was going to break when subjected to launch vibrations. Other team had a big hissy fit and told them to buzz off. And it broke on the shake table.

Here's some info on how the absorbing cones in these chambers are constructed:

https://en.wikipedia.org/wiki/Radiation-absorbent_material

"To be sufficiently lossy, RAM can be neither a good electrical conductor nor a good electrical insulator as neither type actually absorbs any power. Typically pyramidal RAM will comprise a rubberized foam material impregnated with controlled mixtures of carbon and iron. The length from base to tip of the pyramid structure is chosen based on the lowest expected frequency and the amount of absorption required."

Bit of advice: If you are ever moving RAM, wear gloves. Looks just like anechoic foam, but has the capacity to carry very sharp bits and cut you up.

At first glance at the pic, my impression was it wasn't that impressive compared to other shots I had seen on the interwebs. Then, I saw the person for scale, and the meh went right out the window =)

I haven't been inside an RF protected room, but the eerie feeling reminds me of being in a professional sound proof recording booth. Once the door closes, you can't hear anything while hearing everything at the same time. It is surreal the first few times, and then it just becomes one of those things of going to work and you no longer notice it until you stop to think about it again.

Well RF chamber is already luxury. Cheapest thing is RF tent. Like TekBox from Tektronix. A tent is perfect solution to get testing quickly and cheaply. RF chamber needs some commitment - dedicated office space and big chamber needs in addition some time for installation.

A while back we had a skunkworks project that needed to pass CISPR 11 but also had ~0 budget. Initial testing in a rented chamber revealed a grossly-failing amount of energy emitted that had a clear peak but covered a fair bit of spectrum. The solution was to head out to a nearby field and set up a Yagi antenna at one end and an easy-up and generator to power the device at the other. This allowed a few months testing for ~free, and with a fast loop of "iterate design" -> "retest" -> "repeat". It wasn't great testing, but suitable enough to tell if the frequency we cared about was improving and by how many dB.

At the cheap / DIY end, you might be surprised just how effective a metal toolbox is, with the paint filed off the seams around the lid, and some aluminum foil to help make connections / cover the gaps to the lid etc.

We had this problem with a test station for Bluetooth Low Energy (BLE) devices in our factory. With over 250 visible BLE devices visible, the local Bluetooth stack would crash. With a USB Bluetooth module within the toolbox, and the device under test also in the toolbox, we could reliably get it to pick up just the one device, so long as the lid was closed and latched firmly. And then confirm with RSSI that you’re only seeing a very-nearby transmitter. :)

If you think about it, it is a kind of ridiculous idea that you would need super-sensitive equipment to determine if some cheap device would interfere with the workings of some other relatively cheap device.

Do you need sensitive equipment to see if something is totally destroying performance? No. But arguably by definition, you need the ability to reliably measure at or below the noise floor in order to determine if some emitter is putting out interference that will raise the noise floor (thus degrade performance).

Note also that not all measurements are about interference. Antenna patterns and such you just want a more accurate measurement.

You still basically need only the performance of the original devices perhaps plus a little more. This also holds for the antenna application.

Aren't there tons of situations where you'd see >250 BLE devices? Seems to me that fixing the software to not crash should also be a very, very high priority...

But yeah, chambers can be real cheap. My standard is a trash-picked microwave oven, with the magnetron and trafo removed to make it a lot lighter weight. The door is perfect, the isolation is already great at 2.4GHz, there's even a window so you can see status lights.

My first job was as a test engineer for a microwave outfit that did TEMPEST testing.

Antenna physics is a seriously hairy vocation, and the antenna geeks at my company were really smart people.

Reply on news.ycombinator.com