That’s about it for commercial stuff. The trick is really managing it pre-compliance which amounts to renting a signal analyser / SA / measurement receiver and going at it with a near field probe to avoid having to pay for it more than once. Rigol do some adequate kit for less than your mugging from a leasing company now. As for the lab guys they don’t have a tight loop with the design engineers when you are going for compliance testing.
We used stuff in defence sector I can’t talk about even today past saying they have specially designed facilities that rival the commercial sector and fairly capable high end 3D field mapping equipment that cost more than my house did. It was based on commercial kit from Agilent with their own software and hardware.
As for antennas, they’re not really that voodoo. I’ve built quite a few even up to 2.4GHz. I played with 10GHz as well but not successfully yet. Same for radiation. PCB traces are antennas and transmission lines. Impedance control is fun IMHO.
There are design patterns and crib sheets at most large companies that avoid such pitfalls though. Spinning another board is expensive so avoiding this sort of stuff is where your design engineers should be leveraged. Some stuff is indeed tweaking but that’s getting less these days with some of the CAD software around. But it costs real money.
I'll clarify my comment on the antenna end: If you do a standard pattern (a yagi, a quarter-length dipole, or whatnot), it's well-understood. What's deep voodoo are the radiation patterns from complex shapes. I've seen research projects where this is basically just done numerically -- a computer tries a bunch of shapes, computes their radiation patterns, and you get bizarre antennas.
EMC looks a lot more like the latter than the former. Our design had a PCB with several daughterboards, and a whole bunch of cables going in and out. That was voodoo. We'd add shielding. Emissions would go up. Or down. There was little rhyme or reason.
But in the end, we didn't break the bank. We went to the test facility twice, I think. We didn't have any sensible kit in-house, but we did have a scope with an FFT function, and we could make loops our of wire. We couldn't measure in real units, but we got a sense of when things went up and when they went down. And, I think, we had a healthy dose of luck.
I totally believe your equipment was superior than commercial. For defense, you want things like stealth and rad-hard. For us, we just want basic compliance.
We used stuff in defence sector I can’t talk about even today past saying they have specially designed facilities that rival the commercial sector and fairly capable high end 3D field mapping equipment that cost more than my house did. It was based on commercial kit from Agilent with their own software and hardware.
As for antennas, they’re not really that voodoo. I’ve built quite a few even up to 2.4GHz. I played with 10GHz as well but not successfully yet. Same for radiation. PCB traces are antennas and transmission lines. Impedance control is fun IMHO.
There are design patterns and crib sheets at most large companies that avoid such pitfalls though. Spinning another board is expensive so avoiding this sort of stuff is where your design engineers should be leveraged. Some stuff is indeed tweaking but that’s getting less these days with some of the CAD software around. But it costs real money.