On 8/18/2023 8:05 PM, john@kk9a.com wrote:
I am wondering why nearly everything is made this way. Is connecting to the
chassis too costly?
Yes. For 30-40 years, standard practice has been to build boards with
connectors mounted to the board and shove them into an enclosure with
holes for connector to fit through. It costs more build so that the
shield contact makes contact with the enclosure (if it's conductive at
all).
The problem exists in some products when intended contact with the
enclosure is insulated by paint. Every Astron PSU I opened around 2010
had that fault -- the power line green wire was soldered to the mounting
lug of an old-fashioned terminal strip that was insulated from the
chassis by paint, AND V-, which was properly floated on the circuit
board, was, by default bonded to that lug. I'd seen reports that these
supplies could be unstable; now, I knew why! I published this on my
website at that time in this app note, which W4TV, W8JI, and I discussed
at the time. For the products he sold that needed 12VDC, he urged users
to run them from a 12V wall wart rather than the DC buss.
http://k9yc.com/PowerSupplyBondingAndAudioDistortion.pdf
And there's ignorance of the issue, in large part due to "Balkanization"
of electronics education by specialization. So much of the industry is
dominated by those with digital and computer training; only when Johnson
and Graham published two important books analyzing microstrip and
stripline as transmission lines did that classic discipline enter their
universe. Henry Ott introduced me to "High Speed Digital Design" and
"High Speed Propagation," sub-titled "A Handbook of Black Magic" and
"Advanced Black Magic" in a 3-day EMC class I took from him around 2003.
I may have been the only analog guy in a class of 20-25 engineers, and
at 62, probably the oldest.
Pro audio is the only part of the industry that I'm aware of that
cleaned up their act, based on an important EMC papers session at an AES
convention in Los Angeles in 1994, all of them published in the June '95
Journal of the Audio Engineering Society. It became the most re-printed
edition of the Journal, and manufacturers all adopted the practices
raised by the 6-8 presenters. One of them, an engineer at Rane Corp, a
manufacturer based in the Pacific Northwest, showed the problem in their
gear. A few years later they reported that their support calls for hum,
buzz, and RFI were reduced by a factor of 25 or more after they made
running production changes throughout their product line. When Neil
Muncy published his paper on Pin One in 1994, every product on the very
large exhibition floor had Pin One Problems; ten years later, almost
none did.
73, Jim K9YC
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|