Feeding DC through RF shield

[mtwieg]

I'm going to have to feed a DC supply (48V at 40A max) through a filter panel into a shielded laboratory (working frequency in the 100-150MHz range), and am looking for a robust solution that doesn't compromise the shielding. Panel mount feedthrough capacitors are the basic option, but I am worried about their physical robustness. I've used the smaller ones with pins in the past and found that they don't tolerate stress too much before failing. There are also larger ones with stud terminals, but their datasheets don't say much about what they can handle. I would probably have to mount some sort of strain relief onto the panels, which would take up a lot of space.

Are there any off the shelf solutions for feeding high power through a shield? Using an N type coax feedthrough also occurred to me, but we don't have tooling for that, and don't want to drop a huge amount of money on such a thing.

 

[chuckey]

Neither of your capacitors will handle 40A current. I would use an "Anderson 2P connector (various ratings up to 120A) and load the cables inside the chassis with ferrite beads. If required you could strip the braid of a piece of coax and thread your cables through it to decouple them.
Frank

 

[biff44]

you do not need them to be robust. Have big beefy terminals to hook your cables to, and then run short flexible cables, protected by a housing cover, to the filtercon feedthrus.

 

[D.A.(Tony)Stewart]

The weak spot is the cables not the connector.


For a Faraday Cage type operation, all I/O and power cables must be filtered with CM chokes and shielded.

You can use any connector you like then add shunt caps to earth at both ends of the cable inside the lab.

These are filter parts found in Line filters to prevent conducted noise ingress and also prevent it from being radiated along the cables.

When I tested Shield Rooms they usually resonated around 60MHz depending on size plus harmonics.

My preferred RF room was a 8ft cubed copper screen Faraday cage made by Lingren. surplus equip. $2k good for -100dB at 1GHz

 

[vfone]

8ft cubed copper enclosure from ETS-Lindgren cost about $16k, so paying $2k is a very good deal. Perhaps transportation and handling in/out the lab cost more than $2k.
They are also my favorites (usually the 1m cube version), because their >120dB isolation is hard to beat.
ETS-Lindgren sell various shielded/filtered connectors or just filters.

 

[D.A.(Tony)Stewart]

Ok triple that amount for inflation since 2 decades ago. Yes even $6k would be a good deal. $500 by truck from Toronto to Winnipeg. 1 day for assembly. I got one of many factory test chambers for the once famous Canadian Stanley Door Co who made remote control garage openers. Double layer copper gauze ( like bug screen) gazebo with a big metal door and beryllium copper finger stock for all gaps, which they call the Series 71 See & Hear-thru RF screen room.

 

[mtwieg]

Neither of your capacitors will handle 40A current.

Those are just for picture examples, the threaded stud types are available in >100A.

I would use an "Anderson 2P connector (various ratings up to 120A) and load the cables inside the chassis with ferrite beads. If required you could strip the braid of a piece of coax and thread your cables through it to decouple them.
Frank

Powerpole connectors are nice, but I'm specifically asking about how to feed the power through the shield, not how to make pigtail connectors inside or outside the shield. I can't just drill holes in the filter panel and run SOW through it, even if I apply external chokes and capacitors. The entry needs to tight, like what a feedthrough capacitor offers.