Ferrite bead on LM1084 adjustable regulator

The regulator will be used in an environment which is exposed to RFI.
I would like to put some ferrite beads on the regulator PCB. Any
comments? Should I put the ferrite bead at the 3-terminal regulator
input, which is 12V or regulator output which is 8.5V? Any criteria in
choosing ferrite bead? Max load current is 5A. Thanks.

 

Well, if you thing a bead will help, then one at the input and one at
the output leads.
Better yet, use SMD regulator and SMD capacitors close to the leads
and forget beads near the regulator; use RFI shiiielding and suppression
at widget electrical lines going thru the box.

 

Dummy wrote:
|| The regulator will be used in an environment which is exposed to RFI.
|| I would like to put some ferrite beads on the regulator PCB. Any
|| comments? Should I put the ferrite bead at the 3-terminal regulator
|| input, which is 12V or regulator output which is 8.5V? Any criteria
|| in choosing ferrite bead? Max load current is 5A. Thanks.

A good idea, but keep the small decoupling capacitors close to the
regulator, the bigger electrolytics can go behind.
.-------------.
| |
___ | Regulator | ___
o--+-----UUU--+----| |----+-----+---UUU-o
| + bead | | | | | bead
### --- | | .-. ---
--- --- | | | | ---
| 0.1u | '-------------' | | | 0.1u
| | | '-' |
=== === | | ===
GND GND +-----------+ GND
| |
| .-.
--- | |
--- | |
| '-'
| |
=== ===
GND GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

ciao Ban

 

The beads are probably best placed on the input lead, before the high
frequency bypass capacitor. Keep in mind that most beads saturate
with am anp or two DC, so you would probably be better served by some
wound coil rated for the max. current than a bead. The series
inductance will increase the regulators tendency to oscillate (at the
frequency where the coil and bypass cap resonate) so oversize the
bypass cap, to lower the resonance to where the Q is lower.

 

Any problem by sprinkling more than 1 bypass cap,
0.1uF//51pF//100pF//680pF at input and output?
My idea is to bypass the RF freq from HF to UHF. I know the caps won't
be covering a large range like that, but it worths a try right?
Then there will be multiple resonant frequencies.

 

Probably no problem, but I doubt that it is worth using more than 2
caps. A 1 uf stacked film and a .01 uf film should do it.

If you want to eliminate frequencies not handled by those, you will
probably have to do it with a feed through cap that passes through the
box that contains the supply.

 

More is *NOT* better.
Use *one* capacitor close to the regulator at the input, and *one*
capacitor close to the regulator at the output.
If you are really worried, add a bead at the wire going to the input
capacitor bypass, before the regulator (ie: *not* between the cap and
the regulator).
If the current is high enough that one turn (wire passes thru bead
*once*) may saturate it, then get a toroid ferrite of same lossy
material, or use a carbon composition resistor, or a flat film resistor,
as ordinary film resistors are inductive.

 

Stacked design, whether film or ceramic are good; this includes the
chip caps as well.
If one is to use the old flat, round ceramic cap, then wrap copper
foil around it and solder the foil to the ground lead: the foil acts
like a shorted turn and can extend the resonant frequency to near 1GHz.

 

What are the drawbacks by putting more bypass caps at the input and output?