Forums

Slot Machine Music, Vol. 1 & 2: Field Recordings from Middle American Casinos

One of the last products I worked on was a "beam former". These beasts are used in 5G type communication systems. They typically work with 4 signals simultaneously, which is how you get so much data with 5G. A simplified block diagram doesn't even do it justice. Mounting these devices onto a board is a bear too. Each solder connection has to work at all 8 RF ports across the whole frequency band over the entire temp range. Not as easy as it sounds. Board assembly was a huge part of my days at Analog Devices, as getting these parts soldered down perfectly every time was ultra-important to our customers. The last one of these I worked on came in a bumped package i.e. little solder balls in the bottom. The biggest problem with them was that they curl a bit under solder heat temps (~260 degrees C) and sometimes, the little solder balls don't make good connections if the curl is too great.

When I decided to retire, I took a few pictures of my last work areas. I was an applications engineer for Analog Devices' RF products group. I tested parts from MHz to 100 GHz. Here was one of my last test benches, but I don't think the product was mine. By this time, I was responsible for beam forming products. This looks like a GaN ampllifer - you can tell by the monster capacitor on the test board. Those things needed to be pulsed and the drew a ton of current, so you needed a massive cap to be able to supply that current quickly. It was probably this 1-22 GHz amp: https://www.analog.com/en/products/adpa1112.html The labs I worked in typically had 50 or 60 test benches like this. Here is one "chase" with about 10 test benches/racks in it. The floor is conductive (as are  the chairs) for ESD purposes. You can see a temp chamber on the left. We used to have to characterize our parts from -40 degrees to +85 degrees Celsius. The plexiglass box/cover is there because we had to pump in Nitrogen to prevent frost/freezing. This was one of our workhorses in the test lab. As you can see if you can zoom in, this box could test from 2 Hz to 110 GHz with extraordinary dynamic range. Cabling was extremely important with these test boxes. Some of our cables cost upwards of $12,000 each. My understanding of using cables capable of cleanly transmitting high frequency signals is why is am a bit biased against a lot of the audio cable debates... audio (Hz to maybe 16 kHz if you're lucky) is a piece of cake to conduct compared to the signals I used to have to worry about.   BTW - I took this picture to show to the Keysight rep. Those calibration stickers have to be redone every year in a special cal lab.

No way of knowing. It could be anyone.