Thread: power supplies with minimal radio interferance?

The RS units are chassis mounted and current rating is not bad. The Mouser unit looks like they are not chassis mounted but are solder in, chassis mounted ones are $18,00 each. That's ok as the price is way less. With solder in and a diecast case you will need to drill a hole for each cap that is larger than cap, then cut a large piece of double sided copper PCB so that it covers all the openings with one piece. Drill the PCB for a snug fit when the cap is inserted and then solder each of them in (they have one solder flange). The PCB can then be screwed securely to the inside of the case which should then bond both sides of the PCB together.

The install direction is not so much electrically important but the design of the cap results in the larger part of the cap having the strongest mounting point for the cable, this should be on the outside of the case.

I could not see any current rating for the mouser caps and as they are small you may need to use 3 caps on the +5 rail and the +3v rail. Each of these rails has a number of wires in the loom so this should be easy but unfortunately costs a bit more for the extra caps of course.

Mick,
Thanks for the clarifications and information. Placed an order for 20x of the Mouser's. Should arrive mid Feb. I'll see if I can make it all work then.
Needed one size larger box from Jaycar to fit the M4-ATX, plenty of room inside it now.

Updated parts list with http://www.jaycar.com.au/productView.asp?ID=HP9515 as you've made a point of it being double sided. I'm assuming "screwed securely to the inside" = drill a couple of holes in the die cast box and bolt it?

I'm also assuming each hole I make, makes the box "leak" interferrence?


I'll probably also add a few of these http://australia.rs-online.com/web/p/products/483-8607/ (4/9 ways) to allow me to disconnect the PSU box from the rest of the car for maintenance purposes as I've yet to get a case for both the mainboard and the psu case.

It all adds up, but this should hopefully take care of my interferance issue once and for all.



JJ

Looks good. Correct, double sided PCB is screwed to the inside.

Actually, cutting a hole in the lid and mounting a fan in the lid so that the fan supply leads and motor are on the inside of the case lid does not cause any real RF leakage. You can also drill 4 or 5 1/4" holes on the side furthest from the Feed-throughs to exhaust warm air from the case. Radiation of RF is limited by the design of the board, it's RF coupling onto any and every lead that penetrates into the case that is the problem.

The feed through caps bypass RF signals to the inside of the case and keep RF currents inside the case walls. The fan opening is far enough away at these frequencies to not allow RF to bleed/radiate over and back onto the PSU cables. That's why simple toroids and/or caps placed on the PSU leads have almost no effect, RF simply radiates around and away from these and continues on it's merry way. The feed-through caps with the case as a barrier to this localised RF coupling is what stops this from happening and keeps the leads relatively free from HF-RF.


Best of luck and hope it works out for you.

The Mauser part has finally arrived (Sorry, not the most in-focus image, but if one steps back from the computer a meter or two it gets better. The part is just under 1cm long and hollow inside):




Not 100% how this is supposed to work. Actually, I have no idea what I'm doing here. and google's not much help....


Do you solder:
a) on the inside of the "short" side, and the outside of the "long" side?
b) and the PCB goes on the right side of the flange on the image?
c) nothing "touches" the brown part except the PCB itself?


JJ

Really need a clear close up pix of both ends.

Looking at the PDF it shows only one end tinned (for easy soldering). This looks kind of strange in the picture you took.

These devices are a tube with the outer casing soldered to the case around the lip that passes through the case wall. The inner tube conductor is separated from the outer with a dielectric. This forms a capacitor that runs the length of the tube.

It works by bypassing RF from the inner pass-through conductor to the case walls and keeps RF confined to the inner wall of the case therefore removing most of the RF from the lead.

No macro lense on the camera, but hopefully these are better:

Its about 9mm from one end to the other. The brown parts appear to be insulators.



JJ

2404-014-X5U0-102P-8922.pdf


Looking at this PDF I think you have to pass the wire through the device and solder to the small solder ring on the outer tip. Have not seen one like this before but I see no reason why it won't work. Be careful not to put any strain on the device when using heavy hard to bend PSU cables and don't use to much heat.

Just to clarify on installation.

If you are using an aluminium case you would cut a square section out of the case. A piece of double sided copper PCB would later be placed over this cut out. You drill small holes in the copper PCB that are fractionally larger than the caps. The cap is inserted into the hole with the flange on one side (which will be the outside of the RF shielded case when the PCB is mounted). Solder the flange to the PCB and solder around the cap shaft on the other side of the PCB. This electrically connects both sides of the PCB together, repeat for each cap.

Hi and post #229 is ok (I think), but post #228 puts both ends of the wires on the same side of the flange.
Clarification: Not both the tips, but the main body of the wires are on the same side of the flange.




Little confused as to how that works as I thought the purpose was to have each end on different sides of the PCB?




JJ