[Amps] Line Isolators for RF feedback

[Ian White, G3SEK]

R. Measures wrote:

>>> Joe -- the only problems that I have heard of is the ferrite beads 
>>>getting so hot they cracked and dropped off.  This is hardly 
>>>surprising since the ferrite material that is typically used in bead 
>>>baluns is rated by the manufacturer (not by the seller, mind you) at 
>>>1MHz maximum,
>>
>> That "frequency rating" is only for resonant circuits, and does not 
>>apply to baluns. For balun applications you need the "wideband" rating 
>>- this *starts* at 1MHz typ and extends up to 50-100MHz according to 
>>the grade of ferrite.
>
>This is according to the seller or the manufacturer?

The major manufacturer - see www.fair-rite.com


>Permag Pacific is a manufacturer, and their catalog says  Nothing about 
>different frequency ratings for one-turn (bead-balun) applications and 
>multi-turn applications.

I haven't seen their data, but in the Fair-Rite data you will find the 
relevant information under the heading of "EMI Suppression". This is 
because the balun and interference suppression applications share the 
same objective: to create a high impedance along the wire/cable passing 
through the bead.

Figure 2 in Fair-Rite's App Note on 'How to Choose Ferrite Components
for EMI Suppression' shows how resistive loss in a typical nickel-zinc 
ferrite (#43 grade) starts to climb rapidly above 1MHz, which is why it 
is not suitable for tuned circuits above that frequency.

But it *is* still suitable for EMI suppression and balun applications at 
higher frequencies, provided that you use enough beads to create a large 
impedance.

W8JI's balun page explains the limits on heat dissipation quite clearly 
(www.w8ji.com).  There is also good background information in the 
'Choosing the correct balun' page at http://www.dxengineering.com

>>
>> Over its "wideband" region, a ferrite bead becomes increasingly 
>>resistive as well as inductive. The resistive losses in each bead will 
>>be I-squared*R and these will of course heat up the bead.
>
>... and crack, and fall off.

Obviously, not always...


RF engineering doesn't EVER reduce to simple "one-liners".

(except for that one :-)


-- 
73 from Ian G3SEK         'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek