The_Roadie said:
Solved a very difficult tech problem at a customer clean room site in San Jose at midnight last night. So I get to fly home tonight.
For the curious, my equipment had problems calibrating in the highest sensitivity range, around 10mv full scale. Challenge was doing it in 100us, so no time for traditional averaging over a few thousand measurements. Worked perfectly in my lab, of course. Root cause was the customer had a low voltage lamp a foot away from my circuit, and the wall plug power supply for this cheap - ass lamp generated far more noise than the CE regulations allowed. Chinese crap.
Thus, the importance of "clean" power. I remember a big mention of this in the RapidSort manuals, the power supply and certain other wires for communication running to the RapidSort Controller box were to be run separately from everything else, and preferably with its own dedicated power, to minimize the possibility of this type of crap. The RapidSort box was also taking in information from 10 speed sensors, nearly 20 photo-eyes, 2 high-speed scanners, reporting barcodes to and receiving destinations from the server, running 3 monitor stations and a crappy old dot-matrix printer, and operating 58 divert boxes (it would send a "fire" signal and hold it for a length of time, the divert's circuitry and solenoids had their own independent power).
So, there was a dedicated conduit just for the power, one for the communications, and a big old cable tray where everything else was wired up. And in a sealed and ventilated box all to itself. All for the sake of clean communication and minimizing the possible effect of EM on anything, even though I'd venture to guess it would have survived in something less than a practically-clean environment.
I guess the old 'picture them in their underwear' idea wouldn't really help here, since that would be more of a distraction, especially if it's an attractive female 
im guessin that was meant in a good way lol.
